View
2
Download
0
Category
Preview:
Citation preview
1
Increasing customer satisfaction Increasing customer satisfaction Increasing customer satisfaction Increasing customer satisfaction with public transpwith public transpwith public transpwith public transportortortort
Author: Jorrit van ’t Hart
AtAugust 2012
Author: Jorrit van ’t Hart August 2012
Master thesisMaster thesisMaster thesisMaster thesis
i
GGGGraduation thesis raduation thesis raduation thesis raduation thesis MSc TMSc TMSc TMSc Transportation, Infrastructure and Logisticsransportation, Infrastructure and Logisticsransportation, Infrastructure and Logisticsransportation, Infrastructure and Logistics TU DelftTU DelftTU DelftTU Delft
Increasing customer satisfaction Increasing customer satisfaction Increasing customer satisfaction Increasing customer satisfaction withwithwithwith public transport public transport public transport public transport
Author: Jorrit van ‘t Hart
Student number: 1285637
TIL5060
Graduation CommitteeGraduation CommitteeGraduation CommitteeGraduation Committee
Prof. dr. G.P. van Wee – TU Delft, TPM Dr. Eric Molin – TU Delft, TPM
Dr. ir. R. van Nes – TU Delft, CiTG
Drs. P. van Beek – Goudappel Coffeng
Date August 24th, 2012
Attribute Report Thesis Master Program TIL
Kenmerk XXB001/Ht2/0159
Resource cover picture Google Images
Increasing customer satisfaction with public transport Managementsamenvatting ii
Increasing customer satisfaction with public transport Managementsamenvatting iii
DisclaimerDisclaimerDisclaimerDisclaimer
Dit document is geschreven in het kader van een mede door Goudappel Coffeng BV gefaciliteerd
onderzoeksproject, geschreven op persoonlijke titel van de auteur. Goudappel Coffeng BV hecht grote waarde
aan de wetenschappelijke vooruitgang binnen haar vakgebied en de brug naar de praktijk voor onderwijs en
onderzoek. Vanuit die betrokkenheid is inbreng van ons zichtbaar in dit document. Dit document is een
persoonlijke proeve van bekwaamheid van de auteur, waarbij het kan zijn dat de in dit document verwoorde
opvattingen niet overeenkomen met die van Goudappel Coffeng BV. Mocht u vragen daarover hebben, dan
kunt u het beste contact opnemen met de auteur of de betrokken begeleider vanuit Goudappel Coffeng BV.
Increasing customer satisfaction with public transport Managementsamenvatting iv
Increasing customer satisfaction with public transport Managementsamenvatting v
PrefacePrefacePrefacePreface
You are reading my master thesis for the study Transport, Infrastructure and Logistics (Delft University of
Technology). Public transport has always been one of my interests. For my master thesis I integrated public
transport and customer satisfaction. Customer satisfaction is a really interesting topic, because hard and soft
aspects of public transport services are facing each other. Much research is done regarding customer
satisfaction with public transport. However, much more research can be done the next years. Nevertheless,
research does not need to be a goal in itself. Therefore, this thesis focused on policy implications that are based
on customer satisfaction research. I am happy to say that some cost efficient improvements are found.
I would like to thank all people who shared their visions, ideas and theories with me. You helped me to find
the focus of my thesis and with the development of an expert vision. I like to thank my supervisors from the TU
Delft: Eric Molin, Rob van Nes and Bert van Wee. Your criticism, positive feedback and meetings helped me to
write and structure the report. I would also like to thank Paul van Beek for his supervision from Goudappel
Coffeng. You really helped me to find the focus and to interpret the results even better by meetings with other
experts. I especially found the implications from the field interesting. Also, I really appreciate the support to
finish the thesis.
I would also like to thank all my colleagues of Goudappel Coffeng for their visions, inspiration and fun. I
especially would like to thank my public transport colleagues Erik Oerlemans, Hendrik Bouwknegt and Niels van
Oort for focusing on the aspects that are interesting for public transport operators and authorities. I thank the
psychologists Matthijs Dicke-Ogenia (Goudappel), Wilco van Dijk (University of Leiden) and travel satisfaction
researcher Dick Ettema (University of Utrecht) for their convergent vision regarding how travelers give marks
and the data analyses techniques. I thank Joep ten Brink and Ronald Coelman (HTM) for sharing our visions
regarding service improvements based on the results of my data analyses. I also thank Mark van Hagen (Dutch
Railways) for sharing his vision regarding improving the positive experience for train travelers. The last expert I
like to thank is Gerard van Kesteren (KPVV) for sharing his vision regarding the results and the methodology of
the KPVV OV-Klantenbarometer. You all really helped me my by keeping me motivated and completing the
thesis.
I would especially like to thank my co readers and grammar improvers Zoé Tresch, Rosalie de Niet, Mirjam
Hoeffelman and Hugo Fonteyn. The text was tough to read in the beginning, but many improvements are made
with your help and feedback. Finally, I would like to thank my family and friends for the fun moments in
between and all their support.
Jorrit van ‘t Hart
Increasing customer satisfaction with public transport Managementsamenvatting vi
Increasing customer satisfaction with public transport Managementsamenvatting vii
ManagementManagementManagementManagementsamenvatting: Vergroot klanttevredenheid in het openbaar vervoersamenvatting: Vergroot klanttevredenheid in het openbaar vervoersamenvatting: Vergroot klanttevredenheid in het openbaar vervoersamenvatting: Vergroot klanttevredenheid in het openbaar vervoer
Het is belangrijk dat OV reizigers tevreden zijn. Tevreden klanten reizen meer en verkopen het product via
mond-tot-mond-reclame. Tevreden klanten zijn een basis voor het voortbestaan van het openbaar vervoer
systeem. Daarnaast kunnen vervoerders een bonus verdienen als de klanten tevreden zijn. De bonus kan
gemakkelijker worden verdiend door incidentele reizigers, oudere reizigers en vrouwen aan te trekken in
minder drukke periodes.
Maatregelen kunnen het beste worden gericht op de meest ontevreden reizigers: • Frequent reizende scholieren, studenten en woon-werk reizigers tot 40 jaar.
Klanttevredenheid kan worden vergroot met de volgende maatregelen: ■ Voorkom drukte in het voertuig
- Vergroot de betrouwbaarheid van de diensten � Voorkom voortijdig vertrek � Houd de afstand tussen voertuigen gelijk als de frequentie hoog is, ook bij een
kleine vertraging, dit kan met real time bijsturing � Pas de dienstregeling aan aan de actuele rijtijden (35 % percentiel) � Zet meerdere lijnen in op een druk deeltraject � Gebruik wachthaltes of splits lijnen bij een duidelijk breekpunt van
reizigerspatronen - Verleid spitsreizigers naar andere periodes met kortingen en marketing campagnes - Zet de voertuigen vooral in op drukke lijnen
■ Zorg dat klanten de meest recente informatie van de vervoerder ontvangen, bijvoorbeeld via een smart phone app
■ Bied reële verwachtingen ■ Wees klantvriendelijk en klantgericht ■ Rijd netjes ■ Houd de voertuigen netjes ■ Ga een betrokken relatie aan met de klant ■ Stem de eigenschappen van het personeel af op de reizigersgroep van de service bij de
personeelsplanning - Onderzoek reizigerssegmenten en hun behoeftes
■ Positive branding - Vergroot de positieve reisbelevenis - Benadruk het comfort en de reisbelevenis met campagnes - Richt de campagne op een doelgroep, niet op een concessiegebied.
■ De vervoerder en de OV-authoriteit zullen de kwaliteit van de dienst en de klanttevredenheid goed moeten blijven monitoren.
Hulp bij het prioriteren van verbeterpunten is mogelijk. De meningen van reizigers in de KPVV
Klantenbarometer maken waardevolle informatie inzichtelijk voor vervoerders en OV-authoriteiten: ■ De oordelen van reizigers op deelaspecten ■ De gewenste verbeteringen door reizigers ■ De geschatte belangrijkheid van de deelaspecten ■ Met het empirisch klanttevredenheids model kunnen verbeterpunten worden geprioriteerd per concessie.
Het model geeft antwoord op de vraag hoe belangrijk het voorkomen van drukte, de focus op frequente reizigers en het verbeteren van deelaspecten zoals betrouwbaarheid, reistijd, klantvriendelijkheid en rijstijl is binnen een concessie.
Increasing customer satisfaction with public transport Summary viii
Increasing customer satisfaction with public transport Summary ix
Summary ‘Increasing customer satisfaction with Summary ‘Increasing customer satisfaction with Summary ‘Increasing customer satisfaction with Summary ‘Increasing customer satisfaction with public transporpublic transporpublic transporpublic transport’t’t’t’
Main function public transport and importance customer satisfactionMain function public transport and importance customer satisfactionMain function public transport and importance customer satisfactionMain function public transport and importance customer satisfaction The main function of public transport is to provide accessibility. Customer satisfaction is very important to
transport services. Customer satisfaction with public transport is defined as the degree to which an individual
positively evaluates the overall quality of a public transport service. Satisfied customers could become loyal
customers. Literature shows that customer loyalty is important for the operator, because loyal customers use
the service more frequently and recommend the service to others by telling them about their positive
experiences. Therefore, demand and revenues of transport services are highly dependent of customer
satisfaction.
Research goal: cost efficResearch goal: cost efficResearch goal: cost efficResearch goal: cost efficient measures to increase customer satisfactionient measures to increase customer satisfactionient measures to increase customer satisfactionient measures to increase customer satisfaction In the literature, a lot of research has been done regarding costs and travel time, but less about person
characteristics and image aspects influencing customer satisfaction with public transport. The goal of public
transport operators and authorities is increasing customer satisfaction cost efficiency. However, there is a lack
of knowledge on how to do this. Therefore, the research goal is to determine what variables contribute to
customer satisfaction of public transport services in the Netherlands. Result can be used by operators and
authorities to increase customer satisfaction cost efficiently, based on the KPVV Klantenbarometer. The main
research question is: what variables contribute to customer satisfaction of public transport services in the
Netherlands and to what extent?
Research scopeResearch scopeResearch scopeResearch scope
The research focuses on the contribution of supply (regions and modalities), travel and traveller’s
characteristics, regions and transport service types to customer satisfaction. The methods used are literature
study and linear regression analyses using KPVV Klantenbarometer data. The relations to investigate are shown
in the research framework, see Figure S.1. Independent variables are supply, travel and traveller’s
characteristics. Dependent variables are customer satisfaction and the appreciation of four traveller’s needs:
safety, speed, ease and comfort. No data are available regarding the fifth need: experience.
Figure S.1 Research framework
Customer satisfactioCustomer satisfactioCustomer satisfactioCustomer satisfaction: evaluation of supply and expectationn: evaluation of supply and expectationn: evaluation of supply and expectationn: evaluation of supply and expectation
Literature is combined to provide a theoretical framework. Customer satisfaction with a public transport journey
is the result of an evaluation of offered supply by a public transport operator and authority compared to
traveller’s expectations, see Figure S.2. Traveller’s expectations are influenced by their needs. The needs safety,
speed and ease are dissatisfiers. When dissatisfiers are evaluated insufficiently, travellers with alternative
travel options may quit using the public transport service. The needs comfort and experience are satisfiers. New
travellers can mainly be attracted with comfort and pleasant experienced aspects. In addition, expectations
Increasing customer satisfaction with public transport Summary x
regarding traveller’s needs are influenced by promised supply. Moreover, customer’s satisfaction of a journey is
influenced by the experience and satisfaction of previous journeys. Furthermore, customer satisfaction could be
influenced both positively as well as negatively by media attention.
Figure S.2 Journey satisfaction model
Relationships according to the literatureRelationships according to the literatureRelationships according to the literatureRelationships according to the literature
Concluded from many customer satisfaction researches can be that the most important supply sub aspects are
travel time and punctuality. Many other variables make low contributions to customer satisfaction individually;
however, together, many appreciations of sub aspects properly explain customer satisfaction. This research
focuses on the influences and contributions of offered supply, travel characteristics and traveller characteristics.
First, supply characteristics are transport service types. According to the literature, metro is appreciated more
than tram; tram more than bus. Second, travel characteristics are the amount of passengers, day part and trip
purpose. According to the literature, crowdedness contributes negatively to comfort and customer satisfaction;
commuters and students have lower customer satisfaction compared to social-recreational travellers. Third,
traveller’s characteristics are car availability, travel frequency, age and gender. According to the literature, car
availability may contribute positively to customer satisfaction; travel frequency may contribute negatively or
positively to customer satisfaction; age may contribute positively; women feel less safe compared to men.
Customer satisfaction measurements Customer satisfaction measurements Customer satisfaction measurements Customer satisfaction measurements Most customer satisfaction researches are descriptive; no relations are investigated. The KPVV
Klantenbarometer measures overall satisfaction with the trip, appreciations of sub aspects, travel and traveller’s
characteristics in vehicles, in all urban, agglomerational and regional concession areas yearly.
Methodology and classification in empirical customer satisfaction modelMethodology and classification in empirical customer satisfaction modelMethodology and classification in empirical customer satisfaction modelMethodology and classification in empirical customer satisfaction model
The model is built in a way that it analyses the contributions of supply, travel and traveller’s characteristics to
customer satisfaction and the four needs. The model is prepared by clustering of needs, supply characteristics,
trip purposes and day parts. First, the needs are clustered by sub aspects based on a factor analysis in the
literature. In order of importance, the most important factors are speed, comfort, ease and safety. Second, a
spatial distinction of customer satisfaction is found between the four big cities and other areas of the
Netherlands, see Figure S.3. Due to the relationship between spatial area and transport service, both are
combined to transport service type. Therefore, nine transport service types are analysed: city bus within G4, city
bus without G4, agglomorational bus, regional bus, urban tram, agglomorational tram, metro, regional train and
ferry. Third, a pre test showed that trip purposes could be divided into commuting, education, visiting and other
trip purposes. Fourth, no day parts other than morning peak, evening peak and off-peak differ significantly from
each other.
Increasing customer satisfaction with public transport Summary xi
Figure S.3 Results of significant difference between areas controlled for other variables (left: tested areas, right: outcome significant spatial difference)
Results empirical customer satisfaction model: travel and traveller’s contributionsResults empirical customer satisfaction model: travel and traveller’s contributionsResults empirical customer satisfaction model: travel and traveller’s contributionsResults empirical customer satisfaction model: travel and traveller’s contributions
Results for customer satisfaction visualised for the input characteristics
6
7
8
City b
us w
ithin
G4
City b
us o
uts
ide G
4
Agglo
mora
tional bus
Regio
nal bus
Tra
m w
ithin
G4
Tra
m o
uts
ide G
4
Metr
o
Regio
nal tr
ain
Ferr
y
0-2
5 p
assengers
26-5
0 p
assengers
50-1
00 p
assengers
100-5
05 p
assengers
Morn
ing p
eak
Evenin
g p
eak
Off
-peak
Com
muting
Education
Vis
itin
g
Oth
er
trip
purp
ose
Car
availa
ble
No c
ar
availa
ble
0-1
trips p
er
week
2-4
trips a
week
5 o
r m
ore
trips a
week
12-1
8 y
ear
18-2
7 y
ear
28-4
0 y
ear
41-6
4 y
ear
65-8
9 y
ear
Man
Wom
an
Input characteristics
Cu
sto
mer
sati
sfa
cti
on
(m
ark
1-1
0)
Figure S.4 Contributions of supply, travel and traveller’s characteristics to customer satisfaction
The contribution of transport service types, travel and travel characteristics is weak: 5 % variance is explained.
Travel and traveller’s characteristics contribute more to customer satisfaction compared to transport service
types, see Figure S.4. Main travel and traveller’s characteristics contributing to appreciation and satisfaction are
the number of passengers, trip frequency and age. First, the more passengers in the vehicle, the lower
customer satisfaction, controlled for the other input variables. Second, the more one travels, the less one is
satisfied. Third, age influences customer satisfaction and appreciation of speed, ease and comfort positively
significantly. The number of passengers can influence customer satisfaction up to 1.0 points; age up to 0.8
points and trip frequency up to 0.35 points. In addition, women are slightly more satisfied than men with the
overall trip, speed, ease and comfort appreciation. However, women appreciated safety aspects much lower
compared to men. Day parts, trip purposes and car availability contribute weak to customer satisfaction and its
four factors.
Results empirical customer satisfaction model: transport serResults empirical customer satisfaction model: transport serResults empirical customer satisfaction model: transport serResults empirical customer satisfaction model: transport service types contributionsvice types contributionsvice types contributionsvice types contributions No high differences in appreciations of transport service types are found, considering customer satisfaction as
overall trip appreciation. On the national level, transport services are appreciated in the following order:
agglomerational bus, agglomerational tram, regional bus, ferry, city bus outside G4, regional train, metro, tram
Increasing customer satisfaction with public transport Summary xii
urban within, city bus within G4. Therefore, transport services are lower appreciated in the four big cities.
Transport service type can influence at most 0.8 points of customer satisfaction. Transport service types
contribute more to the four appreciation factors compared to customer satisfaction. Therefore, most
(dis)advantages of transport service types are averaged out into customer satisfaction.
RobuRobuRobuRobust results st results st results st results The results are robust according the results of several robustness tests. First, the results are robust for several
methods of dealing with missing values. Second, the predicted values do not exceed the borders of the answer
possibilities. Third, the results of 2011 are comparable to the results of 2010 and 2009. Fourth, the suggested
model and analyses improvements do not change the conclusions.
ConclusionsConclusionsConclusionsConclusions Customer satisfaction is explained weak by supply, travel and traveller’s characteristics. Therefore, the main
conclusion is that much interpersonal bias occurs, which means that different travellers show different
satisfaction. The results show that customer satisfaction is mainly higher
1. when the offered supply is better and the appreciations are higher with respect to travel time,
frequency, punctuality, customer friendliness and driving style. Thus, the needs with respect to speed
and comfort are most important;
2. when the vehicle is less crowed;
3. for incidental travelers and
4. for elderly travelers, especially 65+.
Research implicationsResearch implicationsResearch implicationsResearch implications
Research implications are:
■ A quantitative research: why did travellers fill in the questionnaire as they did? Investigate, e.g, why
travellers of the same service give different marks. Therefore, research interpersonal variance in customer
satisfaction reports by travellers of the same line in a qualitative way; ■ Measure traveller’s occupation dynamic to investigate at what circumstances travellers judge the vehicle
as too crowded and are less satisfied;
■ Whether crowdedness, travel frequency and age are useful axes for market segmentation.
Policy implicationsPolicy implicationsPolicy implicationsPolicy implications Policy implications to increase customer satisfaction are:
■ Prevent crowding. Therefore, increase the service reliability by, e.g., preventing early departures, adjusting
time table information to the current supply, keep headways between high frequent services equal by real
time traffic management, providing multiple lines on shared track if a part of a service is much used and
hold vehicles at waiting stops or split lines if travel patronages are splitted at a stop. Crowding can also be
prevented by attracting peak travellers to the off-peak with price incentives and higher frequencies on the
busiest routes.
■ Negative situations need to be managed well when they occur: provide the right information to travellers
and keep them up to date through the operator;
■ Promise only expectations that can be delivered;
■ The focus of improvements needs to be on frequent travellers and young travellers, especially commuters
(below 40 years), students and scholars. Therefore, marketing should be focused on travellers younger
than 40 years too;
■ Provide customer-friendly behaviour of staff: focus on customer friendliness and driving style;
■ Be polite to students, despite they mostly travel without paying;
■ Frequent travellers could be satisfied more by small attentions;
■ Investigate the most important traveller’s needs and match the traveller’s needs by qualities of the staff in
the planning;
Increasing customer satisfaction with public transport Summary xiii
■ Higher marks on customer satisfaction can be reached by attracting more incidental travellers, woman and
elderly;
■ The operator and their personnel should build a committed relationship with the traveller;
■ The operator and authority should control the service and customer satisfaction by analyses by line and
concession area;
■ Make public transport a positive experience, e.g., by campaigns and vehicle design.
Improvements in customer satisfaction measurementsImprovements in customer satisfaction measurementsImprovements in customer satisfaction measurementsImprovements in customer satisfaction measurements The KPVV Klantenbarometer can be improved by several aspects. First, improvements are more easily to define
when the questionnaire focuses on satisfaction with a line or a bundle of comparable lines, instead of the
current trip. Second, the location and time can registered when the questionnaire is fulfilled on an electronic
device such as a tablet pc. This information can be coupled to crowdedness by chip card data. As a result, the
impact of crowdedness can be investigated in more detail. In fact, the question could be answered ‘what
occupancy rate influences customer satisfaction negatively’. In addition, the input of data can be checked real
time to control the interviewers. Third, the proposed questions concerning experience and loyalty should be
added. Fourth, the method of the most important question needs to be applied to investigate the satisfaction of
short distance and standing passengers. Fifth, a telephonic research among non travellers needs to investigate
how non travellers can be attracted in a concession area. Sixth, special topics can be investigated yearly such as
the main interests of travellers as the explicitly express, market changes for food and drinks at regional
transfers, relation between objective and subjective measurements, innovations, marketing success and the
research recommendations as described above. Seventh, the data and figures can help prioritizing focus points
within specific concession areas or lines.
Key words: public transport, customer satisfaction, appreciations, crowding, frequent travellers
Increasing customer satisfaction with public transport xiv
IndexIndexIndexIndex
Disclaimer iii Preface v
Managementsamenvatting: Vergroot klanttevredenheid in het openbaar vervoer vii Summary ‘Increasing customer satisfaction with public transport’ ix
1111 Introduction: Importance of customer satisfaction with pubIntroduction: Importance of customer satisfaction with pubIntroduction: Importance of customer satisfaction with pubIntroduction: Importance of customer satisfaction with public transportlic transportlic transportlic transport 1111 1.1 Societal and economic importance of customer satisfaction with public transport 2 1.2 Knowledge gaps in the market 3 1.3 Research objective 4 1.4 Research questions 4 1.5 Sub questions 4 1.6 Scope and demarcation 5 1.7 Research methodology 5 1.8 Report structure 6
2222 Conceptual framework customer satisfactionConceptual framework customer satisfactionConceptual framework customer satisfactionConceptual framework customer satisfaction 7777 2.1 Market context 9 2.2 Customer satisfaction system 12 2.3 Traveller’s needs 13 2.4 Appreciation of important supply aspects 15 2.5 Supply characteristics 16 2.6 Travel characteristics 18 2.7 Traveller’s characteristics 19 2.8 Trends in customer satisfaction and patronage 20 2.9 Policy implications 22 2.10 Research implications 22
3333 Methodology and data preparation empirical customer satisfaction modelMethodology and data preparation empirical customer satisfaction modelMethodology and data preparation empirical customer satisfaction modelMethodology and data preparation empirical customer satisfaction model 24242424 3.1 Method linear regression and factor preparation 25 3.2 Missing value analysis 28 3.3 Classifications of regions and transport service types 29
4444 Results empirical customer satisfaction modelResults empirical customer satisfaction modelResults empirical customer satisfaction modelResults empirical customer satisfaction model 32323232 4.1 Univariate descriptions 33 4.2 Bivariate relationships 36 4.3 Results empirical customer satisfaction model on customer satisfaction and factors 40 4.4 Robustness of results 45 4.5 Conclusion results empirical customer satisfaction model 46
5555 Discussion empirical customer satisfaction modelDiscussion empirical customer satisfaction modelDiscussion empirical customer satisfaction modelDiscussion empirical customer satisfaction model 47474747 5.1 Model verification 48 5.2 Discussion of classifications and possible improvements 49 5.3 Discussion of tram bonus 50 5.4 Policy implications regarding the number of passengers, travel frequency and age 51 5.5 Conclusion discussion empirical customer satisfaction model 53
Increasing customer satisfaction with public transport xv
6666 Conclusions and rConclusions and rConclusions and rConclusions and recommendations to increase customer satisfactionecommendations to increase customer satisfactionecommendations to increase customer satisfactionecommendations to increase customer satisfaction 55555555 6.1 Answers sub questions 56 6.2 Answer to the main research question: contributions to customer satisfaction 57 6.3 Research implications 58 6.4 Policy implications 59
7777 Epilogue: expert vision customer satisfactionEpilogue: expert vision customer satisfactionEpilogue: expert vision customer satisfactionEpilogue: expert vision customer satisfaction 60606060 7.1 What customer satisfaction is in public transport according to the author 61 7.2 How customer satisfaction should be measured according to the author 62 7.3 How the results should be used to improve customer satisfaction according to the author 66 7.4 Conclusion expert vision: Policy implications customer satisfaction 68
LiteratureLiteratureLiteratureLiterature 69696969
Experts 71
AppendixesAppendixesAppendixesAppendixes AAAA----1111 A. Terminology A-1 B. Klantenbarometer method A-3 C. Klantenbarometer questionnaire A-7 D. Customer satisfaction measurements and important supply aspects abroad A-9 E. List of supply aspects influencing customer satisfaction A-11 F. Cross validation results empirical customer satisfaction model 2009 and 2010 A-12 G. Methodology and results interaction test A-14 H. Potential cost efficient focus on lowly satisfied travellers A-15 I. Customer satisfaction analyses specified to HTM A-18 J. Input OV Visie Drechtsteden A-26 K. Methods to determine the priority of improvements A-29
Increasing customer satisfaction with public transport xvi
Increasing customer satisfaction with public transport 1
Importance of puImportance of puImportance of puImportance of pubbbblic transportlic transportlic transportlic transport Mobility is important to reach destinations where people can perform activities for reasons to achieve their
personal and business goals. The main function of public transport (PT) is to provide accessibility of places
where people perform activities in regional as well as urban areas.
Relevance customer Relevance customer Relevance customer Relevance customer satisfaction withsatisfaction withsatisfaction withsatisfaction with public transport public transport public transport public transport Demand for public transport is defined as the number of people who are willing to travel by public transport.
Travellers in public transport are customers. Customers of public transport experience a service product moving
them from an origin to a destination. From a customer’s point of view, public transport is seen as a transport
service. Customer satisfaction is a main issue for transport services. Customer satisfaction is important for the
demand and therefore for continuity and profitability of the suppliers of public transport.
Reading guide chapter 1Reading guide chapter 1Reading guide chapter 1Reading guide chapter 1 Main function of the first chapter is to introduce the research goal and question about customer satisfaction
with public transport. To guide from importance of public transport to the research questions different steps are
made, as shown in Figure 1.1 Structure chapter 1. Above, the importance of PT and relevance of customer
satisfaction with PT is argued. Moreover, societal and economic and importance of customer satisfaction will be
described in section 1.1. With the importance of customer satisfaction with PT in mind, section 1.2 describes
what is known about it, and especially what knowledge is lacking. The importance of customer satisfaction with
PT and the lacking knowledge lead to the research objective, as described in section 1.3. To obtain the research
objective, a main research question is formulated in section 1.4. To answer the main research question, sub
questions are derived in section 1.5. In addition, this section provides an overview of methods used to answer
the sub questions. The research is demarcated by available data; section 1.6 will demarcate the scope. To guide
the reader furthermore, section 1.7 will describe the report structure for the following chapters. At last, Section
1.8 will summarize this chapter. Definitions can be found in Appendix A Terminology.
1111
Introduction: Introduction: Introduction: Introduction: Importance of Importance of Importance of Importance of cucucucusssstomer tomer tomer tomer satisfaction satisfaction satisfaction satisfaction withwithwithwith public transport public transport public transport public transport
Increasing customer satisfaction with public transport 2
Figure 1.1 Reading guide chapter 1
1.11.11.11.1 Societal and economic importance of customer Societal and economic importance of customer Societal and economic importance of customer Societal and economic importance of customer satisfaction wsatisfaction wsatisfaction wsatisfaction withithithith public public public public transporttransporttransporttransport
Customer satisfaction with public transport is important for several reasons, which are described in the sub
sections. Customer satisfaction with public transport is defined as the degree to which an individual positively
evaluates the overall quality of a public transport service delivered by a public transport operator and authority
(Ettema et al., 2010).
1.1.11.1.11.1.11.1.1 Increase of patronage and revenuesIncrease of patronage and revenuesIncrease of patronage and revenuesIncrease of patronage and revenues by customer satisfaction by customer satisfaction by customer satisfaction by customer satisfaction
In the introduction of this chapter it is stated that customer satisfaction is important for demand, continuity and
profitability of the operator. In this paragraph, the relevance of customer satisfaction will be explained in more
detail. Satisfied customers could become loyal customers. Customer loyalty is important for the operator,
because loyal customers use the service more frequently, are less price sensitive and recommend the service to
others by telling about their positive experiences (Heskett et al., 1994: Grigoroudis & Siskos, 2001: Lai & Chen,
2011). So, patronage and revenues for operators are expected to increase customers are satisfied.
1.1.21.1.21.1.21.1.2 BonusBonusBonusBonus for customer satisfaction for customer satisfaction for customer satisfaction for customer satisfaction as incentive for operators as incentive for operators as incentive for operators as incentive for operators
Besides revenues for operators, a bonus might be an incentive for operators to increase customer satisfaction.
Customer satisfaction is important because operators can earn a bonus or malus depending on customer
satisfaction in about half of the concessions in the Netherlands.
Increasing customer satisfaction with public transport 3
1.21.21.21.2 KKKKnowledge gapsnowledge gapsnowledge gapsnowledge gaps in the market in the market in the market in the market
As described above, public transport systems need to be seen as a service, taking into account customer
satisfaction. However, little or no money is available for increasing customer satisfaction due to cost cutting
demands from authorities influencing the strategy of operators. Therefore, public transport operators and
authorities like to know what measures increase customer satisfaction cost efficiently. To argue the knowledge
gap about cost efficient measures, this section briefly gives an overview about what is known in literature
about customer satisfaction with public transport. Thereafter, knowledge gaps are stated. The knowledge
overview argues the choice of research goal and question presented in the next sections.
1.2.11.2.11.2.11.2.1 Existing knowledge in the literature and marketExisting knowledge in the literature and marketExisting knowledge in the literature and marketExisting knowledge in the literature and market
This research builds on existing knowledge. To give an overview, in the literature could be found:
■ customer satisfaction and loyalty could increase usage (e.g., Fellesson, et al., 2009);
■ customer satisfaction with public transport is related to social well-being (e.g., Ettema et al., 2010);
■ Important sub aspects are travel speed and reliability (e.g., van Oort, 2011);
■ variables which are related to customer satisfaction are spread in literature (e.g., Konijnendijk, 2003);
■ judgements about customer satisfaction and factors (defined in Table 1.1) (e.g., BEST, 2012);
■ the importance of judgements about sub aspects (defined in Table 1.1) (e.g., KPVV, 2012);
■ that the judgements and the importance of customer satisfaction and sub aspects changes through the
years and (e.g., KPVV, 2012);
■ to what extent and in which direction judgements and importance of customer satisfaction aspects
changes at Dutch Railways through the years (e.g., Brons & Rietveld, 2007).
1.2.21.2.21.2.21.2.2 Lacking knowledge in the literature and marketLacking knowledge in the literature and marketLacking knowledge in the literature and marketLacking knowledge in the literature and market
However, literature is lacking knowledge about:
■ an overview of variables influencing customer satisfaction with public transport, especially
- about supply characteristics (defined in Table 1.1),
- travel characteristics (defined in Table 1.1) and
- traveller’s characteristics (defined in Table 1.1);
■ an overview to which extent variables influence customer satisfaction, especially in urban,
agglomerational and regional public transport in the Netherlands;
■ which measures could influence customer satisfaction is unknown
■ and therefore which cost-efficient measures could influence customer satisfaction is unknown.
Table 1.1 Description of introduced terminology
TermTermTermTerm DescriptionDescriptionDescriptionDescription
Factor / appreciation factor A group of public transport travellers needs. Four factors of travellers needs are
distinguished: safety, speed, ease and comfort, based on van Hagen (2011) and van
Beek et al. (2008). Factors are grouped by sub aspects of customer satisfaction
using a factor analysis. See section 2.3 for more explanation about the distinction
of factors.
Sub aspect Sub aspects are visualized within the pyramid in Figure 1.2, e.g., customer
friendliness.
Variable General description for an aspect, input characteristic, factor or sub aspect that is
related to customer satisfaction.
Supply characteristics Modality and region
Transport service type A transport service type can be seen as a modality with a function. Therefore, bus
services are divided into urban bus, urban-regional bus and regional bus. Several
transport service types are distinguished: urban bus, urban-regional bus, regional
bus, tram, metro, regional train and ferry.
Increasing customer satisfaction with public transport 4
Travel characteristics The number of passengers, day part and trip purpose
Traveller’s characteristics Car availability, travel frequency, age and gender
The knowledge gaps reported that it is mainly unkonwn what variables contribute to customer satisfaction and
to what extent. This knowledge gap accounts especially for urban, agglomerational and regional public
transport in the Netherlands. In the next section, the research objective is described based on the knowledge
demand of operators and authorities and knowledge gaps in customer satisfaction literature.
1.31.31.31.3 Research objectiveResearch objectiveResearch objectiveResearch objective
As described above, the Dutch public transport market demand cost-efficient measures to increase customer
satisfaction. These cannot be described yet. Therefore, this question about the impact of variables needs to be
answered first. The KPVV Klantenbarometer measures customer satisfaction and provides much data for this.
The answer to the question can be used to come up with cost efficient measures by operators, decentralized
public transport authorities and consultants. Therefore, research objective of this study is summarized as:
To determine what variables contribute to customer satisfaction with public transport services in the
Netherlands, such that it can be used by operators and authorities to increase customer satisfaction cost
efficiently, based on the KPVV Klantenbarometer.
1.41.41.41.4 Research questionsResearch questionsResearch questionsResearch questions
As introduced in the previous section, an overview of variables influencing customer satisfaction is desirable. To
which extent variables influence customer satisfaction is directly relevant to come up with cost-efficient
measures. The impact of variables and traveller’s characteristics is important, because it implicates to which
extent customer satisfaction could be improved. This will answer a knowledge gap in literature and market
demand. Influence of traveller’s characteristics and external variables is also unknown in the literature.
Therefore, the main research question is:
What variables contribute to customer satisfaction with public transport services in the Netherlands and to
what extent?
1.51.51.51.5 Sub questionsSub questionsSub questionsSub questions
To answer the main research question, sub questions are needed. This section describes the sub questions and
their functions.
First, it is necessary to consider what customer satisfaction is. Moreover, it is necessary to consider how
customer satisfaction is measured and what variables contribute to customer satisfaction according to previous
researches. Second, the data of the KPVV Klantenbarometer needs to be prepared for the analyses. In fact, it is
necessary to consider how several variables should be clustered. Third, the contribution of supply, travel and
traveller’s characteristics to customer satisfaction and to the four appreciation factors needs to be investigated.
Fourth, the results need to be discussed. In fact, policy implications need to be considered. If a policy
implication could not directly be given, it is necessary to consider research recommendations in advance.
Following from above, the following sub questions are necessary to answer the main research question:
1. What is customer satisfaction with public transport and what variables contribute to customer
satisfaction of public transport services according to the literature?
Increasing customer satisfaction with public transport 5
2. How should the variables be clustered in a model?
3. What do contribution of supply, travel and traveller’s characteristics contribute to customer satisfaction
and to the four appreciation factors?
4. What implications follow from the results?
1.61.61.61.6 Scope and demarcationScope and demarcationScope and demarcationScope and demarcation
The scope and demarcation of the project are set by the literaturei and variables of available data. The data
resource is the KPVV Klantenbarometer. The KPVV Klantenbarometer methodology is described in Appendix B;
the questionnaire is published in Appendix C. The KPVV Klantenbarometer measures yearly customer
satisfaction with public transport. The scope is determined by the variables that are included in the data. A
research framework is built based on the scope, as shown in Figure 1.2. The scope of this dataset is chosen,
because other data is hardly available and linking to the files is too time consuming.
Demarcation is set by the data files:
■ The data demarcates the time span. The Klantenbarometer survey data of 2011 will be used for the
analyses, because this is the most recent data. Policy implications should be based on recent results of
customer satisfaction, because judgements and interest can change through the years, which is concluded
from section 2.8. The survey data of 2009 and 2010 will be used for cross validation of the results. Cross
validation will be done to test the robustness of the results. These years are chosen because they are
nearest by 2011 and therefore less sensitive to changes in supply compared to earlier years.
■ The focus will be on urban, agglomerational and regional public transport, because customer satisfaction of
train services of the main rail network (in Dutch: hoofdrailnet) is not available in the data.
■ Only aspects asked in the questionnaire could be analysed.
■ Grouping of concession areas to regions with the same urbanisation degree is restricted by the current
classification of concession areas.
The scope is summarized in a research framework, as shown in Figure 1.2. The aspects of the research
framework will be described in chapter 2.
Figure 1.2 Research framework
1.71.71.71.7 Research methodologyResearch methodologyResearch methodologyResearch methodology
Table 1.2 Sub questions, functions and methods
Sub questionSub questionSub questionSub question FunctionFunctionFunctionFunction MethodMethodMethodMethod
1 Defining customer Theoretical framework; Literature study
Increasing customer satisfaction with public transport 6
satisfaction;
investigating variables and
impact
Description of variables and extent known;
Determining questions for data analyses
2 Structure of model Classification of variables Model preparation / linear regression
3 Results data analyses Empirical results: the contribution of supply,
travel and traveller’s characteristics to customer
satisfaction
Linear regression
4 Implications Meaning of the results, research and policy
implications
Critical thinking, literature and expert
meetings
The main research question will be answered by literature study, expert meetings and explorative data
analyses, as shown in Table 1.2. The table provides an overview of the sub questions, functions and methods.
The structure of the report becomes clear in the next section.
1.81.81.81.8 Report structureReport structureReport structureReport structure
The report structure is based on the sequence of the
sub questions as described below, see Figure 1.3.
Before investigating variables contributing to
customer satisfaction with PT, it is necessary to
consider what customer satisfaction with PT is.
Therefore, chapter 2 describes a theoretical
framework that describes, among others, the trip
evaluation process by customers based on literature.
Furthermore, this chapter describes variables that
contribute to customer satisfaction as found in
scientific literature and several customer satisfaction
measurements. They are used to argue the variables
chosen in the empirical customer satisfaction model
and to improve the barometer questionnaire. The
methodology of the empirical customer satisfaction
model is described in chapter 3. This chapter includes
several pretest results to argue classifications of
variables. The model is build to answer the sub
questions regarding the contribution of travel,
traveller’s characteristics and transport service types.
The results of the linear regression model are
presented in chapter 4. The results and implications
following from them are discussed in chapter 5,
including recommended questionnaires. Chapter 6
will answer the main research question in the
conclusion and provide recommendations based on
the results. Finally, chapter 7 describes the
developed expert vision of the author. It describes
what customer satisfaction is, how is should be
measured and what needs to be done with the
results according to the author. In addition, policy
implications are included based on expert meetings.
Figure 1.3 Report structure
Increasing customer satisfaction with public transport 7
Figure 2.1 Reading guide chapter 2
2222 ConceptualConceptualConceptualConceptual framework framework framework framework customer saticustomer saticustomer saticustomer satisfactionsfactionsfactionsfaction
Increasing customer satisfaction with public transport 8
The aim of this chapter is to answer the first sub question: what is customer satisfaction with public transport
and what variables contribute to customer satisfaction of public transport services according to the literature?
To answer that question a theoretical framework of customer satisfaction with PT is described in this chapter.
The answer to the sub question is used to describe important variables. They are used as foundation for the
research framework as introduced in chapter 1.
Reading guide chapter 2Reading guide chapter 2Reading guide chapter 2Reading guide chapter 2 The theoretical framework is described in different steps, as visualised in Figure 2.1 Reading guide chapter 2.
First, the context of the market including main stakeholders is described in section 2.1. Using the description of
the market, the position of customer satisfaction with the public transport is described. The section shows how
customer satisfaction, revenues and employee satisfaction influence each other. A main part of this chapter is
the customer satisfaction system in section 2.2. It describes that appreciation depends on demand, e.g.,
requirements, needs, expectations, and supply, e.g., traditional PT supply characteristics and image. How these
variables lead to appreciations is described too. Traveller’s needs are described in more detail in section 2.3. As
a result, the appreciation and importance of supply aspects is described in section 2.4. Supply aspects are
aspects that could be influenced by operators and/or public transport authorities, e.g., frequency and drivng
style. Customer satisfaction measurements from abroad are used as references in this section. Sections 2.5-2.7
discuss the contributions of the input characteristics to customer satisfaction. In fact, section 2.5 focuses on
supply characteristics which are included in the data file, e.g., transport service types ad tendering. Section 2.6
discusses the contribution of travel characteristics to customer satisfaction. In addition, section 2.7 discusses the
contribution of traveller’s characteristics. The literature show different trends in customer satisfaction, which are
described in section 2.8. Moreover, the relationship between the average customer satisfaction and ridership is
discussed. The research goal is to contribute to policy implications for operators and authorities; therefore,
section 2.9 describes policy implications based on the literature. Finally, section 2.10 sums up the questions to
answer with the empiric analyses. Table 2.1 shows the sources regarding the topics in the literature study.
Table 2.1 Sources literature study
TopicTopicTopicTopic FocusFocusFocusFocus SourceSourceSourceSource Research typeResearch typeResearch typeResearch type
Theoretical
framework
Position of customer satisfaction with PT
market
Heskett et al., 2004 Empirical supported theoretical
model
Theoretical
framework
Customer satisfaction system Fellesson et al. 2009; van
Hagen, 2011;
Johansson, 1979;
Lu & Beamish, 2004
Theoretical models
Theoretical
framework
Evaluations through journey satisfaction
model
Fellesson, 2009; Zeithaml
et al., 1993
Theoretical model;
Empirical model
Theoretical
framework
Needs Maslow, 1943; van Hagen
2002; van Hagen, 2011
Theoretical models
Content output
factors
KPVV Klantenbarometer customer
satisfaction data
Van Beek, et al, 2008 Empirical model
Content output
factors
Dutch Railways customer satisfaction data Hagen, 2011 Theoretical model
Content output
factors
Asking importance of aspects Konijnendijk, 2003 Questionnaire results
Reference
projects
Urban and region customer satisfaction
measurements and results abroad; urban and
region
Fellesson et al., 2009 Method description,
questionnaire CS results
Supply
characteristics
Preferences modalities Bovy and Hoogendoorn-
Lanser, 2005
Questionnaire about preference
Supply Tram bonus Bunschoten, 2012 Literature review and empirical
Increasing customer satisfaction with public transport 9
characteristics results
Supply
characteristics
Descriptives KPVV Klantenbarometer
regarding transport service types
KPVV, 2012 Empirical results
Supply
characteristics
Tendering Engelsman et al., 2010 Empirical results
Travel
characteristics
Crowding Friman and Gärling, 2001 Evaluation of case descriptions
Travel
characteristics
Crowding Beirão and Sarsfield Cabral,
2007
Quantitative interviews
Travel
characteristics
Trip purpose van Hagen, 2011 Empirical results
Traveller’s
characteristics
Car ownership Savelberg and Harms, 2009 Empirical results
Traveller’s
characteristics
Travel frequency and seasonal tickets Woldeamanuel and
Cyganski, 2011
Empirical results
Traveller’s
characteristics
Age Jakobsson Bergstad et al.,
2009;
Ettema et al., 2010;
Jakobsson Bergstad et al.,
2010
Empiric data, developed
measurement scale, literature
review
Traveller’s
characteristics
Gender Yavuz and Welch, 2010 Empirical results
Ridership Dutch Railways Brons and Rietveld, 2007 Empirical results
Ridership Public transport in several regions Savelberg and Harms, 2009 Empirical results
Ridership Public transport in several cities Fellesson et al., 2009 Case studies based on empiric
data and patronage
Policy
implications
Positive image and marketing Savelberg and Harms,
2009;
Konijnendijk and van Beek,
2008;
van Hagen, 2011
Empirical results
Policy
implications
Influence tariff by correlated aspects
Konijnendijk and van Beek,
2008
Empirical results
Policy
implications
Specific measures needed by output factor
Konijnendijk and van Beek,
2008
Empirical results
Policy
implications
Most important aspects to improve Konijnendijk, 2003 Empirical results
2.12.12.12.1 Market contextMarket contextMarket contextMarket context
The public transport market serves as an important context to customer satisfaction with PT. Therefore, the
three main stakeholders and their relationships are discussed.
2.1.12.1.12.1.12.1.1 Customer needs and goalsCustomer needs and goalsCustomer needs and goalsCustomer needs and goals
The three main stakeholders are travellers, operator and authorities, as shown in Figure 2.2. Travellers demand
and operators offer supply. In addition, public transport authorities (PTAs) set goals for operators in concessions,
based on the needs and wishes of travellers, according to PTAs. Operators organise and operate according
these goals. It is desirable that the goals that are set by PTAs fit with customer needs.
Increasing customer satisfaction with public transport 10
Figure 2.2 Main actors in public transport market 2.1.22.1.22.1.22.1.2 Power of customersPower of customersPower of customersPower of customers
If the expectations of travellers are not met, they will complain to the operator. Several instruments of
customer power are described. First, travellers may quit their use of public transport if they are not satisfied.
This is explained as follows. Customers have requirements, if not fulfilled, they will search for alternatives.
Travellers with car and bike alternatives may quit using the public transport services, if their requirements are
not fulfilled. However, captives do not have other choice than to use public transport. Second, they may switch
to the service of a competitor if available. Third, customers have power to operators and PTAs by lobby groups,
e.g., Rover. Rover investigates the wishes and complaints from passengers. Rover participates in official organs
such as regional customer meetings called Rocov. Operators and authorities are obligated to consult Rover
regarding major changes such as new timetables and price changes. Law WP2000 obligates this. Fourth,
customers have elective power. Every four years the provincial parliament (in Dutch: Provinciale Staten) is
chosen. Provinces are public transport authorities. Therefore, travellers have an indirect influence on PTAs being
a province; however, travellers do not have elective power to PTAs that are agglomerational governments.
2.1.32.1.32.1.32.1.3 Spatial oSpatial oSpatial oSpatial organisation of supplyrganisation of supplyrganisation of supplyrganisation of supply and demand and demand and demand and demand
Figure 2.3 shows the main aspects of the public transport market. It shows how supply and demand are
organised. The public transport market is divided by spatial areas called concession areas. A concession area is
administered by a PTA. Within a concession area, each transport service type is offered by one operator.
Therefore, one or more operators could be active in an area. Within a concession area, one or more public
transport service types are offered. The services are financed by the authority by subsidies and by the customer
by tickets. Depending on the concession, revenues from ticketing are earned by the operator, authority or the
revenues are divided by both.
The services within a concession area depend on the travel demand and travel patrons. Travellers could travel
within one concession area or across different concession areas making use of multiple transport services and
multiple operators. In addition, travellers have different trip purposes, e.g., commuting, education and visiting.
Trip purpose may influence what travellers expect form the transport service. A discussion about the influence
of trip purpose based on literature will follow in section 2.6.3.
Increasing customer satisfaction with public transport 11
Figure 2.3 Organisation of demand and supply (source map: KPVV, 2012a)
2.1.42.1.42.1.42.1.4 Position of customer satisfaction in the PT market Position of customer satisfaction in the PT market Position of customer satisfaction in the PT market Position of customer satisfaction in the PT market
The position of customer satisfaction with PT is described to define the last part of the market context:
profitability. The Harvard service-profit chain clarifies the position of customer satisfaction with services
(Heskett et al., 2004; Heskett et al., 2008). Figure 2.4 shows the service-profit chain. Already described is that
customer satisfaction leads to customer loyalty, revenue growth and profitability. In addition, revenue growth,
working environment and employee rewards contribute to employee satisfaction. Employee satisfaction
influences the service value, which is defined as what employees have to offer to customers. The better the
employees meet the customer’s needs, the higher customer satisfaction is, and again the higher the revenues
will become. So, customer satisfaction is part of a loop, tha influences revenues and profitability. Concluded
from the loop can be satisfied employees serve customers better.
Figure 2.4 Harvard service-profit chain (source: Heskett et al., 2008)
Increasing customer satisfaction with public transport 12
2.22.22.22.2 CCCCustomer satisfactionustomer satisfactionustomer satisfactionustomer satisfaction system system system system
The service-profit chain positioned customer satisfaction. Now it is necessary to consider how satisfaction is
influenced and how it is evaluated.
2.2.12.2.12.2.12.2.1 Public transport servicesPublic transport servicesPublic transport servicesPublic transport services
Public transport services are a main aspects of the customer satisfaction system, because they are evaluated by
travellers. together with travellers. PT services exist of a line, vehicle and additional services, such as trip advice
and ticket services. Different types of PT service exist, e.g., urban bus services and metro. A PT service offers
supply to travellers organised as described above. Supply can be divided into a) traditional PT supply
characteristics, namely travel speed, frequency and reliability and b) images and services, e.g., factors safety,
ease, comfort and experience.
Figure 2.5 Customer satisfaction system (based on schemes above, Fellesson et al. 2009, van Hagen, 2011 and Johansson, 1979) 2.2.22.2.22.2.22.2.2 Evaluation Evaluation Evaluation Evaluation of the of the of the of the current journeycurrent journeycurrent journeycurrent journey
Appreciation is the evaluation of a factor in the pyramid of customer needs, which are described in section 2.3.
Appreciation leads to customer satisfaction: the overall appreciation of the service. Figure 2.6 shows how a
journey is evaluated by travellers. The figure is based on research of Fellesson et al. (2009). In their study,
travellers filled in a questionnaire within a public transport vehicle in Oslo. The respondents are asked to rate
their satisfaction with variables on a five point scale. Afterwards, qualitative surveys are done with the main
question why did respondents filled in as they did. Their theoretical framework for journey satisfaction, based
on the interviews and literature, will be described. Supply of the current journey, including traditional PT supply
characteristics; image and service, directly influences satisfaction. In addition, supply of the current journey is
compared to the expectations of the traveller. If the expectations are met, then the appreciation is sufficient.
Moreover, the qualitative survey showed that passengers tend to include previous travel experiences with the
Increasing customer satisfaction with public transport 13
operator. In fact, travellers have expectances of the service and could compare the current journey with
previous ones. In addition, their opinion regarding the operator influences customer satisfaction too. The
opinion is influenced by media. Both negative and positive publicity has been reflected in satisfaction scores.
Figure 2.6 Evaluation of journey satisfaction by travellers (based on Fellesson, 2009 and Zeithaml et al., 1993) EEEExpectations of current journeyxpectations of current journeyxpectations of current journeyxpectations of current journey Expectations vary from person to person and could be influenced by many variables. Zeithaml et al. (1993)
interviewed customers of various service industries to conduct a conceptual model of customer expectations of
service. Expectations are influenced by demands aspects such as requirements, needs and trip purposes,
explicit promises made by advertising, implicit promises made by price, word-of-mouth, past experience and
situational variables, e.g., bad weather.
Underestimation of customer satisfactionUnderestimation of customer satisfactionUnderestimation of customer satisfactionUnderestimation of customer satisfaction
The qualitative interview study of Fellesson et al. (2009), as described above, showed that travellers consider
experiences of previous trips when the are asked to give marks; even when the questions focus on the current
trip. In fact, travellers admitted that they give lower marks compared to their feelings about the current trip,
because of consideration of experiences of previous trips and image of the operator. Therefore, reported grades
of passengers may be underestimated. The same underestimation of customer satisfaction could account for
other questionnaires emphasizing the current trip. Thus, the actual satisfaction regarding the current trip may
be higher than reported in the barometer.
2.2.32.2.32.2.32.2.3 Relation supply and useRelation supply and useRelation supply and useRelation supply and use
Seen from a marketing perspective, supply and use are related to each other in an s-curve (Johansson, 1979; Lu
& Beamish, 2004). In fact, a small change in the supply affects the travel behaviour much if the supply is not
extremely bad or good. The curve of the relationships supply-appreciation and appreciation-use is unclear.
2.32.32.32.3 TTTTraveller’sraveller’sraveller’sraveller’s needs needs needs needs
The journey satisfaction model showed how supply is evaluated by expectations and that expectations are
based on needs. This section describes the needs of the traveller. In fact, theory behind the pyramid in the
circle of the customer satisfaction system is described.
Increasing customer satisfaction with public transport 14
2.3.12.3.12.3.12.3.1 Pyramid of human needsPyramid of human needsPyramid of human needsPyramid of human needs
Maslow (1943) ordered five human needs hierarchically: physiological needs, e.g., food and housing; safety
including health, well-being and personal and financial security; love and belonging; self-esteem including
respect; self-actualization, which is about realizing potentials. Remark on the hierarchically structure is
desirable; empirical studies have yielded little or no support for the ideas of Maslow. The identified needs
probably exist; however, the needs are not that deterministic organized as proposed on individual level. That
means that higher factors could be of importance, even if lower needs are not fulfilled as desired.
2.3.22.3.22.3.22.3.2 Pyramid of traveller’s needsPyramid of traveller’s needsPyramid of traveller’s needsPyramid of traveller’s needs
In addition to the human needs, travellers have multiple needs too. Traveller’s needs are indicated by van
Hagen (2002; 2011). Traveller’s needs are reported in a pyramid for visualisation, see Figure 2.7. The factors are
indicators for important aspects. The factors are not shown to be deterministic, therefore, higher factors in the
pyramid can be more important than lower factors for some groups of travellers, depending on, e.g.,
expectations and currently offered supply. The theory of customer needs is used by, e.g., Dutch Railways, to
improve customer satisfaction. The meaning of the factors will be described below.
Dissatisfiers: Safety, reliability, speed andDissatisfiers: Safety, reliability, speed andDissatisfiers: Safety, reliability, speed andDissatisfiers: Safety, reliability, speed and ease ease ease ease
The factors in the pyramid will be discussed. This paragraph describes the dissatisfiers safety, speed and ease,
which are generic traveller’s needs related to the traditional PT supply characteristics. First, basic needs are
safety and reliability. For passengers, safety particularly means social safety at the stop and in the vehicle. If
potential customers perceive a vehicle or stop to be unsafe, they will avoid it as long as they have an
alternative. In the pyramid theory, reliability is the degree of offered transportation facilities as promised. If the
service is not available when and where customers expect it, there is a shortcoming in reliability, which will
result in dissatisfaction. Second, speed should be seen in the perspective of the traveller: when travellers talk
about speed, it is about a short door-to-door travel-time; however, not about a number of kilometres per hour.
Third, ease is about easiness for the brain: less cognitive effort. Travel information and routing signs must be
seen as logical, not causing doubts Safety, speed and convenience are called dissatisfiers; these aspects are
rated negatively if expectations are not met. According to the theory, current passengers may quit using the
service when dissatisfiers are appreciated low (van Hagen, 2011).
Satisfiers: Comfort and experienceSatisfiers: Comfort and experienceSatisfiers: Comfort and experienceSatisfiers: Comfort and experience Satisfiers are travellers needs that pleasant travelling. The expectations and interest in the satisfiers comfort
and experience are specific and can therefore vary by traveller First, comfort is a physical need during the trip.
Comfort differs from ease; ease is regarding cognitive aspects, comfort regarding physical aspects. At train
stations, travellers expect sheltered waiting, seats, food and refreshment facilities. It is unclear which of these
sub aspects is important and to what extent, regarding urban and regional stops not being train stations. In the
vehicle, physical comfort can be arranged by the comfortable driving style and vehicles that produce little
noise. Thus, little hassle is important. Second, pleasant experience can be influenced by such sensory aspects as
station architecture, vehicle design, used materials, colours, daylight, artificial light, smell and music.
Furthermore, a pleasant waiting time is enhanced by food, drinks and other shops at the station. Experience is
enhanced in the following way: travellers see a shop, buy a consumption or something else in the shop they
experience the time as less unpleasant, so they become more satisfied with the overall travel. Besides, the
operator could profit from extra revenues from stops at the stations. Therefore, increasing customer satisfaction
and profit from revenues at station shops can go together (van Hagen, 2011).
2.3.32.3.32.3.32.3.3 Customer needs in urban, agglomerational and regional PTCustomer needs in urban, agglomerational and regional PTCustomer needs in urban, agglomerational and regional PTCustomer needs in urban, agglomerational and regional PT
The theory about traveller’s needs is also applicable to urban, agglomerational and regional public transport
services. Van Beek, van Ingen and Konijnendijk (2008) performed a factor analysis on the data of the customer
satisfaction barometer (N > 80,000). A factor analysis examines which judgments about aspects belong
together and differ from each other. The aim of a factor analysis is to investigate whether separate dimensions,
also called factors, in the appraisals exist. The found factors are: safety, travel speed, ease and comfort. No
Increasing customer satisfaction with public transport 15
questions for experience exist in this questionnaire. The variables are placed within the output factors in the
scheme as they are grouped by the factor analyses, as shown in Figure 2.7. The appraisals within one output
factor correlate high with each other, and less with appraisals in other output factors. Thus, the output factors
are well isolated.
Figure 2.7 Factors and sub aspects of KPVV Klantenbarometer (translated from van Beek, van Ingen and Konijnendijk, 2008) 2.3.42.3.42.3.42.3.4 CCCComparison schemes of van Hagen and van Beek et al. omparison schemes of van Hagen and van Beek et al. omparison schemes of van Hagen and van Beek et al. omparison schemes of van Hagen and van Beek et al.
The results of van Hagen (2002; 2011) and van Beek et al. (2009) are comparable. They show no deterministic
relations. In contrast to the pyramid of van Hagen (2011), the factor analysis showed that reliability is highly
correlated with travel speed, frequency and frequency; less correlated with safety. The results of van Beek et
al. (2008) will be used in the empirical customer satisfaction model for several reasons. First, this structure is
based on customer satisfaction of urban, agglomerational and regional public transport services. Second, the
data of these variables is available. Therefore, factors can be constructed, as found in the literature.
2.42.42.42.4 Appreciation of iAppreciation of iAppreciation of iAppreciation of important supply aspectsmportant supply aspectsmportant supply aspectsmportant supply aspects
The order of needs in this pyramid of customer needs is slightly different compared to the order in the study of
Konijnendijk (2003). In this study, a questionnaire was filled in by 3300 people leaving a vehicle and waiting at
a transfer stop in urban, agglomerational and regional PT. Travellers were asked to weigh the importance of 32
sub aspects on a five-point scale. Thus, the questions was regarding the weights how important it is that the
sub aspects are delivered well, not regarding how important it is to improve the sub aspects. Table 2.2 shows
the importance of sub aspects in relationship to the appreciation of these sub aspects, as investigated with the
barometer.
In Konijnendijk (2003), travellers ranked the variables in the following order of importance: punctuality, social
safety at stop, social safety in vehicle, information about delays, acceptable transfer time at transfer, frequency,
information about departure times and routes, the number of transfers, egress time, lightening at stop, seat
availability, friendliness driver, driving style, access time and knowing where to buy a ticket. The importance of
travel time was found to be the same in Konijnendijk (2003) and van Hagen (2002). Differences in importance
between the studies may be explained by the offered quality of the current services.
Increasing customer satisfaction with public transport 16
Table 2.2 Matrix sub aspects (based on Konijnendijk, 2003)
Low appreciationLow appreciationLow appreciationLow appreciation High appreciationHigh appreciationHigh appreciationHigh appreciation
High importance High priority improvements:High priority improvements:High priority improvements:High priority improvements:
Punctuality
Information about delays
Frequency
Transfer time
Low number of transfers
Lightening at stops
Information in vehicle
Information at stop
Social safety in vehicle
Social safety at stop
Information about departure times and
routes
Seat occupation
Friendliness driver
Access time
Egress time
Transfer time
Driving style
Low importance Cleanness
Signs to/from stop
Noise inside vehicle
Tariff
Facilities at stop
Possibility to eat and drink
Transfer walking distance
Signs at station to stop location
Lightening in vehicle
Knowing selling points
Travel time
Easiness of boarding
Noise outside vehicle
2.4.12.4.12.4.12.4.1 Learnings from customer satisfaction measurements abroadLearnings from customer satisfaction measurements abroadLearnings from customer satisfaction measurements abroadLearnings from customer satisfaction measurements abroad
Customer satisfaction measurements abroad are researched for two reasons. First, convenient aspects of these
measurements may be included in the expert vision regarding how to measure customer satisfaction in chapter
7. Second, variables will summed up that influence customer satisfaction in several cities and sub urbs.
Stockholm, Helsinki and Oslo are chosen as reference studies, because of similarity in method with KPVV
Klantenbarometer. Appendix D argues the similarities, describes the methods of the reference studies and
reports the results. The main conclusions will be reported below.
From the measurements abroad can be concluded that the KPVV Klantenbarometer might be improved by
measurements regarding the image of services and customer loyalty. Also, non-public transport users may be
included to find out how the services are perceived. From the results abroad can be concluded that customer
satisfaction is mainly influenced by appreciations of reliability, frequencies, reducing crowding, driving style,
vehicle cleanness and timetable information.
2.52.52.52.5 Supply characteristicsSupply characteristicsSupply characteristicsSupply characteristics
This section describes the influence of transport service types and tendering on customer satisfaction.
2.5.12.5.12.5.12.5.1 Transport service typesTransport service typesTransport service typesTransport service types
Relation between customer satisfaction and transport service typesRelation between customer satisfaction and transport service typesRelation between customer satisfaction and transport service typesRelation between customer satisfaction and transport service types Bovy and Hoogendoorn-Lanser (2005) studied multi modal trip changes and suggest a preference in access
modes. They analysed route choice behaviour of multi-modal train trips. A multi-modal train trip is a journey
whereby train and modality is used. The study focussed on the preference of different modes to reach the train
station. Train travellers and inhabitants of Rotterdam region are interviewed. The study report the following
order of preferred modes to reach the station: walking, after metro, tram, bike, bus and car. Considering only
modes researched in KPVV Klantenbarometer, according to Bovy and Hoogendoorn-Lanser (2005), preferred
mode would be metro, than tram and than bus, in case regional train and ferry are not considered.
Increasing customer satisfaction with public transport 17
Tram bonusTram bonusTram bonusTram bonus Bunschoten (2012) researched differences in preferences between tram and bus services by literature and
stated preference experiment. First, from the literature study is concluded that tram is appreciated above bus
on high aggregation level. The explanations why tram is appreciated above vary by study. Preference of tram
above bus could be based on lower travel time, higher appreciation for new tram vehicles compared to busses
(Axhausen, 2001; in Bunschoten, 2012), network effects (Hüsler, 1996; in Bunschoten, 2012), mode change or
change in traditional supply variables (Kash and Vogts, 2002; in Bunschoten, 2012), location and target group
(Welschen, 2002; in Bunschoten, 2012). Scherer (2011; in Bunschoten, 2012) found no significance preference for
rail above bus on aggregated level; depending on the location preference for trams are found. In addition to
preference, ridership is researched by Welschen (2002); the additional effect of tram on the ridership is
between zero and ten percent on aggregated level in the Netherlands. Second, the stated preference
experiment of Bunschoten (2012) showed the preference for mode depends of the location. In areas were tram
already exist, tram is preferred above bus services. However, in areas where no tram services are offered, bus
is stated as preferred. Furthermore, Bunschoten (2012) reported that a preference for tram can be explained by
the atmosphere in the vehicle, vehicle characteristics and travel information.
Different results KPVV KlantenbarometerDifferent results KPVV KlantenbarometerDifferent results KPVV KlantenbarometerDifferent results KPVV Klantenbarometer
In contrast to the studies above, the KPVV Klantenbarometer (2012) showed a different order of appreciations
by mode. According to the results of KPVV Klantenbarometer, the overall satisfaction with bus services is higher
compared to satisfaction with tram, see Table 2.3. In fact, the bus is more appreciated on the aspects safety,
punctuality and cleanliness compared to the tram. The bus scores only lower on information by delays
compared to the tram. Bus services in urbanized areas (in Dutch: streekdiensten) and in rural areas are
appreciated slightly better than urban bus services. From the table can be concluded that differences exist
between appreciations of transport service types; especially ferries and regional bus services are rated higher.
Table 2.3 Customer satisfaction in The Netherlands by transport service type (source: KPVV, 2012)
2007200720072007 2011201120112011
Urban bus services ± 7.1 7.2
Agglmorational bus services 7.2 7.4
Regional bus services 7.1 7.4
Metro services 6.9 7.1
Tram services 6.8 7.1
Regional train services 6.8 7.2
Ferry services 7.6 8.0
National average 7.0 7.2
Possible explanations different resultsPossible explanations different resultsPossible explanations different resultsPossible explanations different results Bunschoten (2012) give explanations to why KPVV Klantenbarometer may report higher appreciation by bus
compared to tram during 2005-2011. The arguments are based on the methodology of the barometer and
function of the public transport services. First, in the barometer passengers are interviewed within a vehicle.
Therefore, they judge the current trip; they do not state a preference between modes. In addition, it is possible,
tram travellers have higher expectations compared to bus travellers resulting in lower appreciations. Second,
the reported appreciation for bus above tram is explained by function. In the KPVV Klantenbarometer, on the
sub aspect reliability, tram is appreciated higher compared to bus services. The function of urban bus services is
else than tram. At most day parts, urban bus lines have lower ridership compared to tram and serving city
edges. Therefore, higher punctuality could be reached by bus compared to tram services. Therefore, punctuality
should be addressed to location and ridership; however, not to the modality. Third, urban bus lines have stops
where no tram service is offered. As a result, it may be that bus services are suitable for specific journeys and,
therefore, urban bus services may be seen as tailor made. As a result of that, travellers may appreciate bus
higher compared to tram services in the barometer.
Increasing customer satisfaction with public transport 18
2.5.22.5.22.5.22.5.2 NoNoNoNo relationship relationship relationship relationship between customer satisfaction and tendering between customer satisfaction and tendering between customer satisfaction and tendering between customer satisfaction and tendering
Engelsman et al. (2010) compared the KPVV Klantenbarometer results for tendered and non-tendered
concession areas. There is no straightforward relationship between customer satisfaction and the method of
procurement, which could be a public tender or direct award (in Dutch: onderhandse gunning).
2.62.62.62.6 Travel characteristicsTravel characteristicsTravel characteristicsTravel characteristics
This section discusses the contribution of travel characteristics to customer satisfaction.
2.6.12.6.12.6.12.6.1 The nThe nThe nThe number of passengersumber of passengersumber of passengersumber of passengers
Negative influence of repeated crowdNegative influence of repeated crowdNegative influence of repeated crowdNegative influence of repeated crowdinginginging Friman and Gärling (2001) tested a bus service satisfaction model based on the frequency of negative incidents
by structural equation modelling (N = 95). Students of Karlstad University rated situations of four negative
incidents. Negative incidents are loosing a bus connection, early departure, incorrect information and crowding.
Crowding was described as standing in a bus for the whole trip; during the trip, new standing passengers were
entering. The model showed that the more these situations occur, the lower customer satisfaction with the bus
service will be. Crowding showed a correlation of 0.35 with customer satisfaction. In the model, highest
influence on dissatisfaction was found in the next order: incorrect information, crowding, early departure and
losing a bus connection. Therefore, crowding could also influence customer satisfaction with KPVV
Klantenbarometer negatively.
Negative influence ofNegative influence ofNegative influence ofNegative influence of crowding crowding crowding crowding In addition, Beirão and Sarsfield Cabral (2007) conducted depth interviews to investigate dissatisfiers in Porto;
travellers reported, among others, crowding.
2.6.22.6.22.6.22.6.2 Day partDay partDay partDay part
No results are found for the contribution of different day parts to customer satisfaction.
2.6.32.6.32.6.32.6.3 Trip purposeTrip purposeTrip purposeTrip purpose
Needs and expectations are varying by person. This section will discuss links between customer satisfaction and
trip purposes. Must and lust travellers have different needs.
Must travellersMust travellersMust travellersMust travellers Must travellers are passengers who regularly and systematically travel. Trip purposes of must travellers are
mainly commuting and education. Time is important for must travellers. Must travellers are goal oriented. They
expect a fast and reliable service (van Hagen, 2011). Higher expectations lead to lower satisfaction, therefore, it
could be expected that commuters are less satisfied with the aspects in the output factor speed and are more
satisfied with the convenience and comfort aspects than other travellers are.
Lust travellersLust travellersLust travellersLust travellers Lust travellers travel incidental, their trip purpose is mainly social of recreational, e.g., sporting, visiting and
shopping. Where for must travellers, the trip is only meant for transportation, for lust travellers the trip itself
could be seen as a joyful activity. Therefore, time is less important to lust travellers compared to must
travellers. Convenience, comfort and a pleasant experience of the journey are important for lust passengers
(van Hagen, 2011). Therefore, it could be expected that social and recreational travellers have higher
satisfaction with the speed aspects and are less satisfied with the aspects in the output factors of convenience,
comfort and experience. Also interesting to test, whether incidental travellers are more satisfied with speed
aspects, and less satisfied with convenience and comfort.
Increasing customer satisfaction with public transport 19
2.72.72.72.7 Traveller’s characteristicsTraveller’s characteristicsTraveller’s characteristicsTraveller’s characteristics
This section discusses the contribution of travel characteristics to customer satisfaction.
2.7.12.7.12.7.12.7.1 Car availabilityCar availabilityCar availabilityCar availability
Passengers with alternatives, such as car availability, could ‘vote with their feet’. They can move to another
transport service, such as car or bicycle, in case public transport is not serving according their expectations. In
case they left the public transport vehicle, they will not give marks in the KPVV Klantenbarometer. However,
compulsory travellers do. One could argue that travellers without a car are forced to use public transport.
Because of the force, these captives may be appreciating the transport service less compared to voluntary
travellers who made a conscious choice (Savelberg and Harms, 2009). Therefore, car availability may be an
indicator for customer satisfaction.
2.7.22.7.22.7.22.7.2 Travel frequencyTravel frequencyTravel frequencyTravel frequency
Positive contribution to customer satisfaction by seasonal tickets Positive contribution to customer satisfaction by seasonal tickets Positive contribution to customer satisfaction by seasonal tickets Positive contribution to customer satisfaction by seasonal tickets
Woldeamanuel and Cyganski (2011) applied a panel binomial probit model to analyse the relationship between
customer satisfaction with accessibility in PT and socio-economic such as gender, age and car ownership as well
as mode-related. Data is used from the German Mobility Panel 1997-2008 (N over all years > 10,000). The
results suggested a positive relationship between satisfaction and travel frequency; although this relationship
was not significant, a significant relationship is found between satisfaction and ownership of seasonal tickets,
which are assumed to be mainly frequent travellers. Thus, the results of travellers with a seasonal ticket
implied that frequent users may be more satisfied regarding accessibility to public transportation compared to
incidental travellers. However, satisfaction regarding aspects as ease and comfort is not researched.
Different Different Different Different KPVV Klantenbarometer results by travel frequencyKPVV Klantenbarometer results by travel frequencyKPVV Klantenbarometer results by travel frequencyKPVV Klantenbarometer results by travel frequency In contrast, the KPVV Klantenbarometer reports a negative relationship between customer satisfaction and
travel frequency on national level; people who made many PT trips in the last seven days, report lower
customer satisfaction. On concession level, the relationship seems to be valid for the range two to six or more
travels per week (KPVV, 2012). One reason for the negative influence could be that negative aspects such as
delays and crowding have seven times higher influence than positive ones. The change that frequent travellers
experience such a negative aspect is higher compared to incidental travellers. The negative impact is cognitive
linked to traveller’s overall perception of public transport (Ettema et al., 2010).
2.7.32.7.32.7.32.7.3 AgeAgeAgeAge
Hypotheses based on social wellHypotheses based on social wellHypotheses based on social wellHypotheses based on social well----beingbeingbeingbeing Hypotheses regarding age are based on social-well being, because literature of the contribution of age to
customer satisfaction is lacking. Social well-being (SWB) is defined as the degree to which an individual
positively evaluates the overall quality of their live. Jakobsson Bergstad et al. (2009) showed that satisfaction
with life correlate around 0.20 with domain specific well-being in the context of activities and travel by
introducing a satisfaction with travel scale. Customer satisfaction with public transport is also an evaluation of
quality. Therefore, contributors to satisfaction of live may contribute to customer satisfaction with PT. The
relationship between age and live is a U-distribution; SWB is lowest around 40 years. Therefore, a hypothesis is,
customer satisfaction with PT is highest for young and elderly and lowest for travellers around 40 years (Ettema
et al., 2010).
Older travellers are more satisfiedOlder travellers are more satisfiedOlder travellers are more satisfiedOlder travellers are more satisfied Jakobsson Bergstad et al. (2010) researched the contribution to travel satisfaction of travel and traveller’s
characteristics. Travel satisfaction is the degree of positive evaluation of a travel made by a private or public
mode. Therefore, the results might be comparable to customer satisfaction with PT. 1,330 Swedish respondents
filled in a questionnaire by post or on the web. A multiple linear regression analysis was performed to estimate
Increasing customer satisfaction with public transport 20
travel satisfaction by the number of cars in household and car use. The number of cars in household and car use
was corrected for correlations with socio-demographic variables: gender, age classes, cohabiting, children,
education level, income, employment and residential area. Residential areas are divided in urban, semi-rural or
rural area. Car use has a weak significant effect on travel satisfaction; only 2% of the variance is explained
corrected for correlations with the socio-demographic variables. Travel satisfaction with the oldest age group is
significantly higher than in the youngest age group. The oldest age group is set above 55 years; the youngest
people are between 18-35 years. This result is in contrast to the hypothesis based on social well-being. No
significant contribution to travel satisfaction is found for the other social-demographic variables.
2.7.42.7.42.7.42.7.4 GenderGenderGenderGender
WomWomWomWomeeeen feel less safe and appreciate information highern feel less safe and appreciate information highern feel less safe and appreciate information highern feel less safe and appreciate information higher
Yavuz and Welch (2010) researched differences by gender on safety perception and reliability regarding train
services. Data is collected from 1172 train travellers by phone (n = 1172). The interviewees live in the city of
Chicago or its suburbs. Their findings suggest that women appreciate social safety less than men. Women
appreciated frequency and timetable information more than men. However, none of the gender differences is
statistically significant at the 0.05 probability level. Based on linear regression, they concluded, experience with
train services is likely to make passengers feel safer. Experience may eliminate the effects of other variables,
such as being female or elderly.
No clearNo clearNo clearNo clear relationship relationship relationship relationship between age and socialbetween age and socialbetween age and socialbetween age and social----well beingwell beingwell beingwell being The correlation between SWB and gender is unclear. Some studies show that women are more satisfied with
life than men, others show that SWB is independent from gender, and still others show that the relationship
between gender and SWB differ across age classes (Ettema et al., 2010). Thus, the contribution of gender to
customer satisfaction is unclear.
2.82.82.82.8 Trends in customer satisfaction Trends in customer satisfaction Trends in customer satisfaction Trends in customer satisfaction and patronageand patronageand patronageand patronage
The next section will discuss trends in customer satisfaction in The Netherlands. The main message is that
appreciations and the importance of aspects changes through the years.
2.8.12.8.12.8.12.8.1 Changes in Changes in Changes in Changes in satisfaction withsatisfaction withsatisfaction withsatisfaction with aspect in urban and regional public transport in the aspect in urban and regional public transport in the aspect in urban and regional public transport in the aspect in urban and regional public transport in the NetherlandsNetherlandsNetherlandsNetherlands
In the KPVV Klantenbarometer, the nation wide customer satisfaction rate for the overall trip is stable: 7.2 out of
10 in recent years. Customer satisfaction shows deviation by sub aspects through the years. Several sub aspects
are discussed below.
Table 2.4 Three most appreciated aspects in 2011 (Source: KPVV, 2012)
2006200620062006 2011201120112011
1. Seat availability 8.0 8.3
2. Easiness of boarding 8.0 8.3
3. Speed 7.0 7.3
Table 2.5 Three most unappreciated aspects in 2011 (Source: KPVV, 2012)
2006200620062006 2011201120112011
1. Information about delays 4.2 4.8
2. Price 4.9 4.8
3. Noise in the vehicle 5.9 6.2
Increasing customer satisfaction with public transport 21
Table 2.6 Three most called aspects to improve in 2011 (Source: KPVV, 2012)
2006200620062006 2011201120112011
1. Ticketing service (public transport chip card) 4.2 4.8
2. Seat availability 4.9 4.8
3. Cleanness in vehicles 5.9 6.2
Table 2.4 shows the aspects with the highest mark in KVVP Klantenbarometer. The aspects with the lowest
marks are shown in Table 2.5. Seat availability has the highest mark, but it is also marked as a main point for
improvement, as shown in Table 2.6. One reason for this paradox could be in the method: especially travellers
who are sitting in a vehicle are requested to fill in the questionnaire. The other aspects in the top three of
points of improvements are in the first place ticket service and on the third place cleanness of the vehicle.
Information about delays is the most unappreciated aspect; however, it is not in the top of most checked option
for improvements. Easiness of buying a ticket and loading chip card debit has decreased since 2008. This could
be explained by adding the aspect of loading chip card in the question about easiness of ticket buying and start
up problems with the public transport chip card (Konijnendijk & van Beek, 2008; KPVV, 2012). The next
paragraph discusses how the differences may be explained through the years.
2.8.22.8.22.8.22.8.2 Possible explanations of changes in satisfactionPossible explanations of changes in satisfactionPossible explanations of changes in satisfactionPossible explanations of changes in satisfaction
In general, the differences between 2006 and 2011 can be explained by several reasons. First, appreciations
could be increased by improvements of operators. Janse and van Ingen showed that HTM travellers appreciated
the cleanness of the vehicles more after improving the cleaning policy. In fact, the HTM opinion panel slightly
gave higher grades for safety every month. In addition, Brons and Rietveld (2007) analysed the satisfaction and
importance of sub aspects regarding the Dutch railways. They compared the satisfaction and importance in 2001
with 2005. The appreciations regarding punctuality and comfort are increased. The change in satisfaction could
be explained by improvements of these aspects. Second, unsatisfied travellers may be lost; therefore, customer
satisfaction increases on average. The next sub section describes that the relationship between satisfaction and
ridership is not clear.
2.8.32.8.32.8.32.8.3 No cleaNo cleaNo cleaNo clearrrr relationship relationship relationship relationship between customer satisfaction and ridershipbetween customer satisfaction and ridershipbetween customer satisfaction and ridershipbetween customer satisfaction and ridership
Several studies are done regarding the relationship between customer satisfaction and ridership. First,
Savelberg and Harms (2009) compared the changes in the KPVV Klantenbarometer results with the changes in
patronage in several areas. They found no uniform result: they found areas where customer satisfaction
decreased and the ridership increased, where satisfaction increased and the ridership decreased and where
both decreased. Second, a study of Fellesson et al. (2009) showed comparable results. They compared the
relationship between satisfaction and ridership of several services in Oslo, see Figure 2.8. The changes show no
Figure 2.8 Relationship between change in satisfaction and ridership for several services in Oslo (Fellesson et al., 2009)
Increasing customer satisfaction with public transport 22
uniform result. Thus, there is no clear relationship between satisfaction and patronage. The relationship can
vary for several reasons. First, new customers with similar needs compared to the current travellers can be
attracted if customer satisfaction and the offered quality are high. Second, customers with high needs can be
attracted if customer satisfaction is increased. As a result, the average customer satisfaction decreases. Third,
customers can quit the use of PT if the customer satisfaction is low. The travellers that stay are also less
satisfied. Fourth, customers with high needs can quit the use of PT when their satisfaction decreases. As a
result, the average, the average customer satisfaction increases. With situation occur may be dependent on the
feelings spread by word-of-mouth, variation in customer needs and the offered quality.
2.8.42.8.42.8.42.8.4 Customer satisfaction not constant at concession levelCustomer satisfaction not constant at concession levelCustomer satisfaction not constant at concession levelCustomer satisfaction not constant at concession level
For urban and regional public transport, a glance at the individual concession areas shows remarkable
differences through the years. Some concession areas show high increases or decreases in customer
satisfaction thought the years, while others are stable. The concession areas will not be discussed in detail. The
main message from these observations is that customer satisfaction could increase (KPVV, 2012).
2.92.92.92.9 Policy implicationsPolicy implicationsPolicy implicationsPolicy implications
The literature shows interesting policy implications.
2.9.12.9.12.9.12.9.1 Strategy implicationsStrategy implicationsStrategy implicationsStrategy implications
Customer satisfaction can be increased by two strategies: promising less and giving more.
2.9.22.9.22.9.22.9.2 Positive image and marketingPositive image and marketingPositive image and marketingPositive image and marketing
New customers could be gain by pleasant comfort and experience aspects. Marketing about comfort and
pleasant experience is needed to gain new travellers. A positive image by media is important. Because
expectations about comfort and experience differ by traveller group, target groups, e.g. based on travel trip
purpose, are important in marketing (Konijnendijk & van Beek, 2008: van Hagen, 2011). Besides marketing,
customers can be influenced unconsciously. Psychological research shows show that sub aspects such as
colours, scents, sounds, temperature and design are visually very dominant. Customer satisfaction surveys do
not take these pleasant experience sub aspects into account yet (Savelberg & Harms, 2009, van Hagen, 2011).
2.9.32.9.32.9.32.9.3 Most important aspects to improveMost important aspects to improveMost important aspects to improveMost important aspects to improve
The main sub aspects to increase customer satisfaction are of high importance and of low appreciation. From
the matrix of Konijnendijk (2003) can be concluded that Dutch operators and authorities can improve customer
satisfaction by, e.g., improving punctuality, higher frequency, better information in vehicles and at stops and
better information about delays. The result may be outdated. However, the matrix is an useful tool to demine
the main focus points.
2.102.102.102.10 RRRResearch implicationsesearch implicationsesearch implicationsesearch implications
2.10.12.10.12.10.12.10.1 Conclusion literature studyConclusion literature studyConclusion literature studyConclusion literature study and implications for data and implications for data and implications for data and implications for data analysesanalysesanalysesanalyses
Concluded from literature study can be that punctuality and travel time are most important sub aspects in
general. Several studies report also other important contributions. The list with other sub aspects is long, see
Appendix E. This appendix list the sub aspects that influence customer satisfaction, organised by output factor,
including the availability of data. Reason for the long list is that many aspects have a low influence on
customer satisfaction separately. However, all together these aspects explain a lot of customer satisfaction.
Table 2.7 sums up variables about travel characteristics, traveller’s characteristics and external developments
that influence customer satisfaction. Because of the fact that many variables contribute to customer
Increasing customer satisfaction with public transport 23
satisfaction, multivariate data analyses is be needed to correct for correlations with other variables. In addition,
the contributions to the appreciation factors safety, speed, ease and comfort are hardly investigated yet.
Table 2.7 Travel, traveller’s characteristics and external aspects influencing customer satisfaction
Other dimensionsOther dimensionsOther dimensionsOther dimensions AspectsAspectsAspectsAspects Availability in KPVV Availability in KPVV Availability in KPVV Availability in KPVV
Klantenbarometer datasetKlantenbarometer datasetKlantenbarometer datasetKlantenbarometer dataset
Supply characteristics Modality
Concession / research area
Available
Available
Travel characteristics Crowding
Day part
Trip purpose
Indicator available
Available
Available
Traveller’s characteristics Car ownership
Travel frequency
Trip purpose
Age
Expectations
Available
Available
Available
Available
Not (direct) available
2.10.22.10.22.10.22.10.2 Outline data Outline data Outline data Outline data analysesanalysesanalysesanalyses
Data analyses question are formulated to answer the main research question. These questions are based on the
hypotheses and knowledge gaps found in literature study:
■ What appreciation factors and sub aspects contribute the most to customer satisfaction of public transport
in 2011? (chapter 3)
■ How do customer satisfaction and its four appreciation factors differ by the supply, travel and traveller’s
characteristics: transport service type, the number of passenger in the vehicle, travel frequency, trip
purpose, car ownership, age, and gender? (chapter 4)
Increasing customer satisfaction with public transport 24
Figure 3.1 Reading guide chapter 3
3333
Methodology and Methodology and Methodology and Methodology and data preparationdata preparationdata preparationdata preparation empirical customer empirical customer empirical customer empirical customer satisfaction modelsatisfaction modelsatisfaction modelsatisfaction model
Increasing customer satisfaction with public transport 25
To answer the data analyses questions of the previous chapter, empiric research will be done. Therefore, a
model needs to be build, which is called the empirical customer satisfaction model. This chapter is about the
methodology of the model and data preparation.
Model relevanceModel relevanceModel relevanceModel relevance The model is relevant, because it test composition and context effect. Composition effects are influences based
on person characteristics; e.g., differences between customer satisfaction may be explained by differences
within socio-demographics such as age. Context effects are influences based on attributes of a spatial area and
are independent of person characteristics; e.g., customer satisfaction may depend on the transport service type
(Graaf and Kalmijn, 1999). Therefore, the results will show to what extent variance will be explained by supply,
travel and traveller’s characteristics.
Reading guide chapter Reading guide chapter Reading guide chapter Reading guide chapter 3333 Figure 3.1 shows the reading guide. The data of the appreciation factors and input characteristics need to be
prepared. Section 3.1 describes the method, which is linear regression, and the calculation of the four factors.
Moreover, missing values are discussed in section 3.2. Transport service types needs to be classified by spatial
characteristics, as discussed in section 3.3. The next chapter reports the univariate, bivariate and multivariate
results of the model as specified in this chapter.
3.13.13.13.1 Method linear regression and Method linear regression and Method linear regression and Method linear regression and factorfactorfactorfactor preparation preparation preparation preparation
3.1.13.1.13.1.13.1.1 TerminologyTerminologyTerminologyTerminology
Terminology introduced in this chapter is shown in table 3.1.
Table 3.1 Terminology introduced in this chapter
TerminologyTerminologyTerminologyTerminology DescriptionDescriptionDescriptionDescription
Customer satisfaction
(questionnaire)
The answers on the question about overall trip appreciation
Appreciation Appreciation of a factor (calculated average)
Meaningful Significant and relevant result; beta ≥ 0.08
Collinearity Very high correlation between two input variables (correlations above .80)
Multicollinearity Very high correlation between three or more input variables (correlations above .80)
VIF-value Test for multicollinearity: if VIF-value < 10 no problems with (multi)collinearity occur.
Pearson correlation Measure of the linear dependence between two variables. Range: [-1,+1].
Exclude cases listwise (only
used for sensitive analyses)
Exclude any cases with at least one missing value
Exclude cases pairwise (used
in model)
Calculate each correlation as long as values for the two variables are available. Therefore,
the number of respondents used in the analyses is higher than listwise exclusion and vary by
missing values of correlation pairs.
R square R square represents the explained variance. It is a measure for the goodness of fit between
independent and dependent variables.
B Unstandardized regression coefficient that estimates the influence on the mark (1-10),
dependent of variance of the independent variable
beta Standardized regression coefficient, estimates the impact on customer satisfaction,
independent of variance of the independent variable
Standard deviation The standard deviation σ is defined as the square root of the variance σ2, σ = √E[(X-μ) 2],
whereby μ is the sample mean.
t-value t-values above |1.96| are significant at a 95 % confidence interval
Increasing customer satisfaction with public transport 26
3.1.23.1.23.1.23.1.2 Method linear regression analysisMethod linear regression analysisMethod linear regression analysisMethod linear regression analysis
The empirical customer satisfaction model is based on linear regression analyses. Linear regression models the
linear relationship between independent variables by fitting a linear equation to observed data of a dependent
variable. The coefficients of the independent variables explain how these variables are related to the
dependent variable, corrected for correlations with other independent variables in the model.
Independent and dependent variablesIndependent and dependent variablesIndependent and dependent variablesIndependent and dependent variables In the empirical customer satisfaction model, the independent variables to research are the input variables as
described in the research framework: transport service type, region, the number of passengers, day part, trip
purpose, car availability, travel frequency, age and gender. The dependent are customer satisfaction, and the
four appreciation factors.
Interpretation beta coefficients Interpretation beta coefficients Interpretation beta coefficients Interpretation beta coefficients In the empirical customer satisfaction model, all independent variables are entered in the model at once.
Variables that do not explaining the dependent variables are deleted from the analyses. How much a variable
explains, is based on the beta of the variable. The model contains variables of interval level and dummies. As a
result, the input characteristics have different variances. Therefore, meaningful variables are selected on their
standardized regression coefficients. The unstandardized regression coefficients are used to report how much
the meaningful variables contribute to customer satisfaction, because these coefficients are easier to
understand.
Assumptions linear regressionAssumptions linear regressionAssumptions linear regressionAssumptions linear regression The assumptions for linear regression are linearity between the input and dependent variables, independence
of input variables, homoscedasticy and normality. Linearity is checked. Homoscedasticy and normality are no
problems due to many cases because of many interviewees (N=83,513 by pairwise excluded missing values).
Input variables are independent enough, having correlations lower than 0.90. Looking forward, VIFs in the
empirical customer satisfaction model are lower than 10. Therefore, there are no problems with
multicollinearity occur.
Respondents weighted for patronage in all analysesRespondents weighted for patronage in all analysesRespondents weighted for patronage in all analysesRespondents weighted for patronage in all analyses The respondents are weighted for the amount of public transport travellers in the service area in the analyses.
Travellers are weighted as described in Appendix B.3; every respondent is weighted by the response frequency
in a vehicle, offered supply by day part and estimate patronage within the concession area.
3.1.33.1.33.1.33.1.3 Method Method Method Method factor calculationfactor calculationfactor calculationfactor calculation
Customer satisfaction is directly asked in the questionnaire; however, the four factors are not. Therefore, the
factors are calculated. For each of the factors safety, speed, ease and comfort, the appreciations about the sub
aspects within the factor are unweighted averaged. The factor is only computed if no missing values occur for
all variables within the factor. The descriptive statistics of the factors are shown in Table 3.2.
Table 3.2 Descriptive statistics of customer satisfaction and the four factors.
MeanMeanMeanMean Standard Standard Standard Standard
deviationdeviationdeviationdeviation
Customer satisfaction 7.24 1.42
Safety appreciation 7.80 1.39
Speed appreciation 6.22 1.49
Ease appreciation 7.85 1.58
Comfort appreciation 6.81 1.38
Increasing customer satisfaction with public transport 27
Verification: customer satisfaction covered by factorsVerification: customer satisfaction covered by factorsVerification: customer satisfaction covered by factorsVerification: customer satisfaction covered by factors The effect of the simplification is analysed to support the choice of the four factors instead of fifteen sub
aspects. Therefore, Table 3.3 shows the contributions of the factors to customer satisfaction. In addition, Table
3.4 shows the results of a linear regression with as independent variables the fifteen sub aspects. All sub
variables significantly contribute to customer satisfaction on national level. In addition, no collinearity problems
occur in the described regression analyses. The explained variance by the sub aspects is 59%, which is only
slightly more compared to the four factors (57%). As a result, the explained variance by the factors is not much
lower compared to the sub aspects. Therefore, the simplification by factors is allowed.
Table 3.3 Regression for customer satisfaction by the factors (pairwise exclusion, N = 83,513)
Regression
coefficient Beta t
Pearson correlation
with customer satisfaction
Constant of customer satisfaction 1.200* 41.2
Safety 0.086* 0.084 23.6 0.42
Speed 0.403* 0.424 105.8 0.68
Ease 0.124* 0.139 35.7 0.54
Comfort 0.277* 0.270 66.3 0.62
R square 0.57
* significant (P<0.05)
Table 3.4 Pretest regression for customer satisfaction by sub aspects (pairwise exclusion, N = 83,513)
Factor Sub aspect
Regression
coefficient Beta t
Constant 0.876* 33.8
Safety at stop 0.024* 0.028 8.5
Safe
ty
Safety in vehicle 0.050* 0.054 15.6
Travel speed 0.150* 0.190 55.4
Punctuality 0.085* 0.131 39.7
Frequency 0.114* 0.166 53.1
Information at stop 0.036* 0.055 17.1
Delay information 0.032* 0.059 19.3
Speed
Price 0.035* 0.062 20.7
Boarding 0.045* 0.056 17.8
Ticketing facilities 0.028* 0.048 15.8 Ease
Seat availability 0.043* 0.071 23.6
Friendliness personnel 0.079* 0.106 31.3
Driving style 0.079* 0.099 29.5
Noise in vehicle 0.028* 0.038 12.4 Com
fort
Cleanness 0.078* 0.101 31.2
R square 0.59
* significant (P<0.05)
Implication of high explained varianceImplication of high explained varianceImplication of high explained varianceImplication of high explained variance The explained variance of customer satisfaction by the four factors is high. This may be interpreted in two
ways. First, the questions of the questionnaire may cover customer satisfaction well. Second, respondents may
use the ten point scale consistent. This second hypotheses is supported by t-test between customer satisfaction
and sub aspects. The average appreciations of all sub aspects is significantly higher for satisfied travellers
compared to unsatisfied travellers.
Increasing customer satisfaction with public transport 28
3.1.43.1.43.1.43.1.4 Structure customer satisfactionStructure customer satisfactionStructure customer satisfactionStructure customer satisfaction
Considering whether one or more of these factors have a different structure, an additional factor analysis is
performed. The method, results and implications are briefly described. Variables in the factor analyses are the
four factors and customer satisfaction. As a result of the factor analyses, one factor is found. Thus, these five
aspects are within the same factor. Therefore, the structure of the factors and customer satisfaction is highly the
same.
3.23.23.23.2 MissMissMissMissing value analysising value analysising value analysising value analysis
Missing values are investigated to verify the computed correlations. The data file is cleaned by Goudappel prior
to this research. On total, 83,513 travellers filled in all or most of the questions. One average, less than 8% of
Table 3.5 Missing values
Number of valid cases % Missing
Suppl
y ch
arac
-te
rist
ics
Transport service type 83,513 0.0 %
Amount of passengers 83,513 0.0 %
Day part 83,513 0.0 %
Trav
el
char
acte
rit-
ics
Trip purpose 78,578 5.9 %
Car available 77,034 7.8 %
Trip frequency 80,413 3.7 %
Age 79,285 5.1 %
Inpu
t ch
arac
teri
stic
s
Trav
elle
r's
char
acte
rist
ics
Gender 79,472 4.8 %
Customer satisfaction (overall) 80.705 3,4 %
Factor Safety appreciation (calculated) 69,502 16.8 %
Safety at stop 72,930 12.7 %
Safe
ty
Safety in vehicle 73,045 12.5 %
Factor Speed appreciation (calculated) 53,100 36.4 %
Speed 79,550 4.7 %
Punctuality 78,589 5.9 %
Frequency 78,002 6.6 %
Information at stop 77,233 7.5 %
Delay information 68,118 18.4 %
Speed
Price 68,287 18.2 %
Factor Ease appreciation (calculated) 69,492 16.8 %
Boarding 81,330 2.6 %
Ticketing facilities 70,947 15.0 % Ease
Seat availability 82,176 1.6 %
Factor Comfort appreciation (calculated) 71,348 14.6 %
Friendliness personnel 75,617 9.5 %
Driving style 79,359 5.0 %
Noise in vehichle 80,830 3.2 %
Out
put:
sat
isfa
ctio
n a
nd
appr
ecia
tion
s
Com
fort
Cleanness 82,127 1.7 %
Average 77,196 7.6 %
the values are missing pairwise, see Table 3.5. However, 56% of the values are missing listwise. Moreover,
pairwise exclusion is standard in linear regression. Therefore, all analyses are done with 83,513 respondents
from which 8% of the values lack.
Increasing customer satisfaction with public transport 29
3.2.13.2.13.2.13.2.1 Missing values within inMissing values within inMissing values within inMissing values within input variablesput variablesput variablesput variables
The input variables have relatively few missing values. Within the input variables, most missing values are
related to the question whether respondents have an alternative for this trip, from which follows car
availability, see question 17 in Appendix C. Perhaps, it is hard for respondents to evaluate which of the six
boxes regarding alternative travel options should be checked. Perhaps, only asking for car availability is better.
Missing values not at randomMissing values not at randomMissing values not at randomMissing values not at random The missing values are not completely at random according to Little's MCAR test (Chi-Square = 60950.0, df =
53587, sig. = 0.000). The significant result states that the means for customer satisfaction and the four
appreciation factors differ by some missing values. Travellers who left the box for alternative travel options
unfilled, give lower marks compared to travellers who filled in one of these boxes. In addition, travellers who
did not reply their age give lower marks compared to travellers who respond their age. Based on these
differences between present and missing values, concluded can be, missing values are not at random. To solve
this problem, first, the analyses will be done with the available data and, second, the analyses will repeatedly
be done with datasets whereby missing values are imputed. The second step is done for verification of the
robustness of the model. Section 4.4 will report the results of the data imputation. In short, the conclusion is
that the conclusions are robust for different methods of dealing with missing values.
Missing values regarding price appreciationMissing values regarding price appreciationMissing values regarding price appreciationMissing values regarding price appreciation One of the author’s hypotheses was that the trip purpose education shows more missing values for price
appreciation compared to the other trip purposes and input variables. However, more missing values occur
between price and gender and between price and other sub aspects within the factor speed. Therefore, the
assumed hypothesis is not meaningful.
3.2.23.2.23.2.23.2.2 Missing values within output variablesMissing values within output variablesMissing values within output variablesMissing values within output variables
As stated above, the output variables have the most missing values. Nevertheless, missing values within
output variables are less problematic compared to missing values within input variables, because missing
correlations of output variables have less impact on the results. By far, the factor speed has most missing
values (36.4%), compared to the other factors. The factor speed is based on six variables. Within the factor of
speed, appreciation of price has most missings. It has most missing values regarding the appreciations of the
other variables within factor. Several reasons can be given why more missing values occur by the four factors,
especially the factor speed, compared to input variables. First, the factors are composited from several sub
aspects. Second, people could fill in the questionnaire strategically. E.g, price rise may be undesirable by
travellers, therefore travellers may appreciate price insufficient or give no judgement. Thus, missing values
causes no problems, therefore the analyses can continue.
3.33.33.33.3 ClassificationsClassificationsClassificationsClassifications of regions of regions of regions of regions and transport service ty and transport service ty and transport service ty and transport service typespespespes
This section will describe the classification of regions. Due to differences between urbanization degree, a
classification of transport service types is defined in this section.
3.3.13.3.13.3.13.3.1 Classification ofClassification ofClassification ofClassification of regions regions regions regions
The Netherlands can be divided into different areas. The Netherlands can be divided by many categories, such
as, provinces, by Randstad/non-Randstad, by cities and by agglomerations. The choice of classification for this
research about customer satisfaction with public transport is based on the supply of public transport and the
possibilities to make classifications in the dataset.
SSSSpatial classification by city, agglomeration and townspatial classification by city, agglomeration and townspatial classification by city, agglomeration and townspatial classification by city, agglomeration and towns
Supply of public transport differs among cities, agglomerations and towns, e.g. stop distance, frequency.
Therefore, spatial differentiation is preferable between urban, agglomerational and regional services. The
Increasing customer satisfaction with public transport 30
distinction is preferable by line. However, line information is not available in the data file; however, it
distinguishes the service area by research area for bus services. A research area is the same as a concession
area in most cases. Nevertheless, the distinction is made by line in several areas. These lines are combined in a
research area. The distinction is made in the big cities Amsterdam, Rotterdam, The Hague, Utrecht (G4),
Eindhoven, Nijmegen and Groningen. In this report, a city is defined as place with more than 100,000
inhabitants. A spatial distinction of the CPB is used to divide the Netherlands in three zones with different
urbanization degrees: Randstad, non rural areas outside the Randstad and rural areas, see the left part of Figure
3.2. The Randstad is be demarcated between Rotterdam, Utrecht, Almere, Amsterdam and agglomerational
area above Amsterdam to Zaandam and Castricum (Verkade et al., 2006). Rural areas are Friesland, Zeeland and
the Achterhoek compared to the other areas outside the Randstad. As a result, Groningen, Drenthe, Overijssel
and Limburg will be taken into account in the category non rural areas outside the Randstad. Moreover, the big
cities are located in the first two zones. They have a higher activity density compared to the agglomerational
areas around them. The G4 are the biggest four cities in the Netherlands, based on inhabitants. These cities are
denser compared to other cities and areas in the Netherlands. Only transport service types serving mainly the
city, and not much suburbs, are accounted within G4. For that reason, the tram of Utrecht-Nieuwegein, the
Randstadrail The Hague-Zoetermeer, the Randstadrail Rotterdam-The Hague and the erry Amsterdam-Velsen
are excluded from the G4. These services are included in the area Randstad outside G4. The distinction in three
zones and big cities within the zones results in the spatial classification. As a result, it distinguishes five areas in
addition to the existing classification of concession areas in the data file. The value of the addition of these five
areas is researched in a pretest. In addition, due to the fact that other classifications can be proposed, chapter 5
shows that the conclusions are robust for other spatial classifications with the existing data file.
Figure 3.2 left: spatial classification of the Netherlands, input for pretest Right: spatial classification of the Netherlands, results of pretest (original maps adjusted from KPVV, 2012a)
Results Results Results Results spatial spatial spatial spatial pretestpretestpretestpretest
The pretest shows that the G4 has a significant addition in explaining customer satisfaction corrected for
correlations with other input variables, as shown in the right Figure 3.2. The significant addition of the G4 is also
valid for appreciation of safety, ease and comfort, while for appreciation of speed the G4 does not have
significant difference compared to other areas in the Netherlands. The other for areas do not report significant
differences between each other. Thus, customer satisfaction differs spatial between the four big cities and
elsewhere of the Netherlands, while no additional difference in customer satisfaction is found between the
other areas compare to the current distinction between urban, agglomorational and regional bus services at
concession level. Therefore, only the distinction between the G4 and the rest of the Netherlands is valuable in
analyses.
Increasing customer satisfaction with public transport 31
3.3.23.3.23.3.23.3.2 Classification into transport service typesClassification into transport service typesClassification into transport service typesClassification into transport service types
The simplified distinction based on the pretest between regions inside and outside G4 has an impact on the
empirical customer satisfaction model. Availability of transport service types is related to the function and area,
as organised in Figure 3.6. Within the G4, the following transport service types are offered: tram, metro and
urban bus services. Outside G4, the following transport service types are offered: urban bus, agglomerational
bus, regional bus, tram, regional train and ferry. Therefore, only urban bus and tram services are offered in both
the G4 as outside the G4. As a result, eight different dummies for the combination of modalities and regions
will be used. Metro will be used as reference, because metro scores the lowest on customer satisfaction.
Table 3.6 Transport service types divided by region nine
Transport Transport Transport Transport
service service service service
typetypetypetype
RegionRegionRegionRegion
Urban busUrban busUrban busUrban bus AAAAgggggloglogloglo----
mmmmerational buserational buserational buserational bus
(stadsstreek)(stadsstreek)(stadsstreek)(stadsstreek)
Regional Regional Regional Regional
bus bus bus bus
(streek)(streek)(streek)(streek)
TramTramTramTram MetroMetroMetroMetro Regional Regional Regional Regional
traintraintraintrain
FerryFerryFerryFerry
Within the G4 city dummy x x dummy reference x x
Outside the G4 city dummy dummy dummy dummy x dummy dummy
X = not available
3.3.33.3.33.3.33.3.3 No additional contribution by operatorNo additional contribution by operatorNo additional contribution by operatorNo additional contribution by operator
No significant explanations of customer satisfaction are found by the operator, corrected for correlations with
the other independent variables. Therefore, operators are left out of the data analyses.
Increasing customer satisfaction with public transport 32
Figure 4.1 Reading guide chapter 4
4444
Results Results Results Results empirical empirical empirical empirical customer satisfaction customer satisfaction customer satisfaction customer satisfaction modelmodelmodelmodel
Increasing customer satisfaction with public transport 33
The empirical customer satisfaction model estimates the contributions of input variables to customer
satisfaction. This chapter describes the findings. Moreover, the contributions of input variables to the four
factors of safety, speed, ease and comfort is also taken into account. This is described in the methodology
chapter.
Reading guide chapter Reading guide chapter Reading guide chapter Reading guide chapter 4444 Before the results of the empiric customer model can be presented, univariate descriptions (section 4.1) and
bivariate relationships (section 4.2) should be considered, see Figure 4.1. The bivariate descriptions show how
customer satisfaction is related to another variable, without taking into account other variables. The
relationships between input variables are included in the customer satisfaction model. Section 4.3 describes the
results of the empirical customer satisfaction model; the contribution of supply, travel and traveller’s
characteristics to customer satisfaction and the factors are estimated. The results will be shown in several
figures. Section 4.4 describes the robustness of the model to methodical choices. Section 4.5 contains a
summary and conclusion, based on the results described in this chapter. In the next chapter, the results will be
used in the discussion of validity and policy implications.
4.14.14.14.1 Univariate descriptionsUnivariate descriptionsUnivariate descriptionsUnivariate descriptions
The univariate descriptions show the composition of the group of interviewed travellers in the
Klantenbarometer of 2011. The order of the sections 4.1.1-4.1.3 goes by the order of the input characteristics as
shown in the research framework. Section 4.1.4 describes the distributions of given marks for customer
satisfaction and its four factors.
4.1.14.1.14.1.14.1.1 Supply charactSupply charactSupply charactSupply characteristics eristics eristics eristics
This section shows the univariate descriptions for the supply characteristics. Figure 4.2 describes the deviation
of transport service type used by the interviewees. As stated in chapter 4, all results are weighted by, among
others, the expected amount of travellers in a concession area. The bus is the modality that is most used. The
tram is the most used transport service type within the biggest four cities. Regional train and ferry are not used
much compared to the other transport service types.
Distribution of transport service types
City bus w ithin G4
14.6%
City bus outside G4
6.1%
Agglomorational bus
21.5%
Regional bus
14.0%
Tram w ithin G4
24.3%
Tram outside G4
4.0%
Metro
10.3%
Ferry
0.3%Regional train
4.9%
Figure 4.2 Distribution of transport service types by interviewees
4.1.24.1.24.1.24.1.2 Travel characteristicsTravel characteristicsTravel characteristicsTravel characteristics
This section reports the univariate descriptions of the travel characteristics. First, Figure 4.3 shows a histogram
for boarding passengers. In most cases, one service of a line is used by less than 52 travellers. Second, Figure
Increasing customer satisfaction with public transport 34
4.4 shows the distribution of day parts. The most interviewees travel during a workday, especially in the off-
peak or evening peak. Third, the interviewees checked boxes on their form to clarify trip purpose(s). Trip
purposes in the questionnaire are commuting, education, shopping, sporting, visiting and other activities. Most
frequent trip purposes are commuting and education, as shown in Figure 4.4. In addition, many travellers report
a unclear trip purpose as going home or an other activity.
Figure 4.3 Distribution of boarding passengers during one service of a line
Figure 4.4 Distribution by day parts Figure 4.5 Distribution of the trip purposes
4.1.34.1.34.1.34.1.3 Traveller’s characteriTraveller’s characteriTraveller’s characteriTraveller’s characteristicssticssticsstics
This section reports the univariate descriptions of the traveller’s characteristics. Figure 4.6 shows the
distribution of car availability. Most interviewees do not have a car available as an alternative for their trips.
Figure 4.7 describes travel frequencies. Concluded can be, from this figure, most interviewees are regular public
transport users. Only a small amount used public transport not or once in the week before the interview. The
circle diagram in Figure 4.8 shows the distribution of ages. Especially travellers of ages 18 up until 27 years are
well represented. Elderly people are interviewed less compared to other age classes. Figure 4.9 shows that
most interviewees are female. The distribution by gender differs from that of the Dutch population. In the Dutch
population, the amount of woman is only slightly bigger than the amount of men (CBS, 2012).
Figure 4.6 Distribution of gender availability Figure 4.7 Distribution of travel frequency
Increasing customer satisfaction with public transport 35
Figure 4.8 Distribution of age classes Figure 4.9 Distribution of gender
4.1.44.1.44.1.44.1.4 Customer satisfaction and factor appreciationsCustomer satisfaction and factor appreciationsCustomer satisfaction and factor appreciationsCustomer satisfaction and factor appreciations
This section describes the distributions of the marks given. The distributions follow the normal curves in outline.
The marks 7 and 8 were given most to customer satisfaction, see Figure 4.10. Moreover, many tens are given to
all three of the sub aspects of ease, as shown in Figure 4.11. Some suspicious cases are found. These report only
tens or combinations of ones, threes and tens. The factors are correctly calculated.
Figure 4.10 Distribution of marks for customer satisfaction
Figure 4.11 Distribution of marks for the factors safety, speed, ease and comfort
Age class
Gender
Increasing customer satisfaction with public transport 36
4.24.24.24.2 BiBiBiBivariate variate variate variate relationsrelationsrelationsrelationshipshipshipships
This section continues on the univariate descriptions by describing the bivariate relationships between the input
characteristics and customer satisfaction.
4.2.14.2.14.2.14.2.1 BivariateBivariateBivariateBivariate relationships relationships relationships relationships of sof sof sof supply characteristicsupply characteristicsupply characteristicsupply characteristics and customer satisfaction and customer satisfaction and customer satisfaction and customer satisfaction
This section describes the relationship between transport service types and customer satisfaction. Figure 4.12
shows the average marks for customer satisfaction and its four factors for different transport service types. The
colours of the bars correspond with the colours in the research framework. The bar distributions are very similar
across the different transport service types. Therefore, the differences in customer satisfaction between
modalities and regions are related to differences in appreciations of all factors, not by only one factor. Ferry
services score higher on all factors.
Customer satisfaction by transport service type
1
2
3
4
5
6
7
8
9
10
City
bus
with
in G
4 (N
=12,6
11)
City
bus
outs
ide G
4 (N=5,
255)
Agg
lom
erat
iona
l bus
(N=18
,557
)
Regi
onal b
us (N
=12,0
89)
Urb
an tr
am (N
=21,
014)
Agg
lom
orat
iona
l tra
m (N
=3,425
)
Met
ro (N
=8,8
62)
Regi
onal t
rain (N
=4,2
59)
Ferry
(N=
231)
Transport service type
Sati
sfa
cti
on
/ a
pp
recia
tio
n (
mark
)
Customer satisfaction
Safety appreciation
Speed appreciation
Ease appreciation
Comfort appreciation
Figure 4.12 Average marks for customer satisfaction and its four factors by transport service types
4.2.24.2.24.2.24.2.2 BivariateBivariateBivariateBivariate relationships relationships relationships relationships of tof tof tof travel characteriravel characteriravel characteriravel characterissssticsticsticstics and customer satisfaction and customer satisfaction and customer satisfaction and customer satisfaction
This section describes the bivariate relationships between the travel characteristics and customer satisfaction.
Because of the results, classifications for day parts and trip purposes are discussed.
Customer satisfaction by crowdedness indicator
1
2
3
4
5
6
7
8
9
10
100 200 300 400 500
Amount of passengers (passenger)
cu
sto
me
r s
ati
sfa
cti
on
(m
ark
)
Observed Customer satisfaction averages
by the number of passengers from data
Observed Safety appreciation averages by
the number of passengers from data
Observed Speed appreciation averages by
the number of passengers from data
Observed Ease appreciation averages by
the number of passengers from data
Observed Comfort appreciation averages
by the number of passengers from data
Linear estimation Customer satisfaction
(R square = 0.13)
Quadratic estimation Customer
satisfaction (R square = 0.17)
Figure 4.13 Average marks for customer satisfaction and its four factors by the amount of passengers
Increasing customer satisfaction with public transport 37
Bivariate relationship between the number of Bivariate relationship between the number of Bivariate relationship between the number of Bivariate relationship between the number of passengers and satisfactionpassengers and satisfactionpassengers and satisfactionpassengers and satisfaction Figure 4.13 shows the bivariate relationship between the number of passengers and customer satisfaction. The
results shown represent the average values of customer satisfaction. Satisfaction and appreciation decrease as
the amount of passengers increases.
Curve estimations forCurve estimations forCurve estimations forCurve estimations for relationship relationship relationship relationship between between between between the number of passengers the number of passengers the number of passengers the number of passengers and satisfactionand satisfactionand satisfactionand satisfaction Customer satisfaction is estimated by, among others, a linear and quadratic relationship in SPSS, without
considering the other variables in the research framework. The lines are fitted to separate cases, not directly to
the shown averages in Figure 4.13. The estimate curves are shown with the blue lines in Figure 4.13. One of the
author’s hypotheses was that satisfaction and ease appreciation decrease strongly when the vehicle is so
crowded that no seats are available anymore and very little standing space is left. This could be shown with an
s-curve. Therefore, the figure should show a strong decrease of customer satisfaction when the vehicle
becomes crowded, for example, a value between 20 and 100. However, Figure 4.13 shows a strong decrease
starting at 1-10 passengers, whereby the vehicle could not be crowded. In addition, the quadratic estimation
(light blue line) is not in line with the hypothesis. Therefore, the hypothesis does not hold. This analysis is done
for every transport service type separately; the variation in capacity and the line length still exist, however,
there is less variation. The results for separated transport service modes are in line with the above. The last
way to eliminate spread in the capacity and line length with the available data is to repeat the analyses for
every research area separately. A research area is a concession area specified by modality. The data file exists
of 87 research areas. Therefore, modalities are researched separately. However, the results are again in line
with the results above; customer satisfaction and the factor appreciations decrease strongly when the amount
of passengers is between 1-20. Seats are still available in those cases. Therefore, the nonlinear curves between
the number of passengers and customer satisfaction do not show a strong decrease when all seats are used
and the vehicles become crowded. Therefore, the inflection point of the s-curve is not at the expected amount
of passengers. Therefore, it is not appreciated to use a complex curve in stead of a simple line. Concluded from
this analyses can be that a linear estimation between the number of passengers and customer satisfaction and
its four factors is the most robust estimation.
BivariateBivariateBivariateBivariate relationship relationship relationship relationship between day part and satisfactionbetween day part and satisfactionbetween day part and satisfactionbetween day part and satisfaction Figure 4.14 shows the bivariate relationship between day parts and customer satisfaction. The patterns of the
bars are similar for every day part. The differences between day parts are very small.
Customer satisfaction by day part
1
2
3
4
5
6
7
8
9
10
Working
day before
10h
(N=16,722)
Workingday
10 - 16h
(N=27,616)
Working
day 16 -
19h
(N=17,437)
Working
day after
19h
(N=8,756)
Saturday
before 19h
(N=6,128)
Sunday
before 19h
(N=1,407)
Sunday
before 19h
(N=4,165)
Sunday
after 19h
(N=1,281)
Day part
Sati
sfa
cti
on
/ a
pp
recia
tio
n (
mark
)
Customer satisfaction
Safety appreciation
Speed appreciation
Ease appreciation
Comfort appreciation
Figure 4.14 Average marks for customer satisfaction and its four factors by day part
Increasing customer satisfaction with public transport 38
Classification of day partsClassification of day partsClassification of day partsClassification of day parts The empirical customer satisfaction model needs a classification of day parts and trip purposes. In this
paragraph, the classification of day parts is discussed. Below the classification by trip purposes is discussed. The
results in Figure 4.14 show that customer satisfaction is influenced by morning peak, evening peak and off-
peak. However, an ANOVA pretest has shown that no off-peak periods significantly differ from each other on
explaining customer satisfaction. In addition, customer satisfaction differs only slightly from day part to day
part. For that reasons, the off-peak periods will be grouped in one category. Therefore, the day parts morning
peak, evening peak and off-peak will be considered in the empirical customer satisfaction model.
BivariateBivariateBivariateBivariate relationship relationship relationship relationship between trip purpose and satisfactionbetween trip purpose and satisfactionbetween trip purpose and satisfactionbetween trip purpose and satisfaction Figure 4.15 shows the relationship between trip purposes and satisfaction. The satisfaction differs slightly by trip
purpose. Travellers with trip purposes of commuting and education report lower customer satisfaction and all
four of the appreciation factors compared to the other trip purposes.
Customer satisfaction by trip purpose
1
2
3
4
5
6
7
8
9
10
Commuting
(N=17,474)
Education
(N=15,420)
Shopping
(N=5,926)
Sporting
(N=1,893)
Visiting
(N=8,231)
To home or
other trip
purposes
(N=34,569)
Trip purpose
Sati
sfa
cti
on
/ a
pp
recia
tio
n (
mark
)
Customer satisfaction
Safety appreciation
Speed appreciation
Ease appreciation
Comfort appreciation
Figure 4.15 Average marks for customer satisfaction and its four factors by trip purpose
Classification of trip purposesClassification of trip purposesClassification of trip purposesClassification of trip purposes
An ANOVA pretest showed commuting, education and visiting differ from the other activities, while no
difference was found for shopping and going sporting. As a result, four types of trip purposes can be
distinguished: commuting, education, visiting and other trip purposes including, among others, sporting and
visiting.
4.2.34.2.34.2.34.2.3 BivariateBivariateBivariateBivariate relationships relationships relationships relationships of tof tof tof traveller’s characteristicsraveller’s characteristicsraveller’s characteristicsraveller’s characteristics an an an and customer satisfactiond customer satisfactiond customer satisfactiond customer satisfaction
This section describes the bivarate relationships between the traveller’s characteristics and customer
satisfaction.
BivariateBivariateBivariateBivariate relationship relationship relationship relationship between car availability and satisfactionbetween car availability and satisfactionbetween car availability and satisfactionbetween car availability and satisfaction Figure 4.16 compares travellers with and without car availability. Travellers with a car at their disposal are
slightly more satisfied compared to travellers with no car at their disposal. Concluded from this can be that
travellers with an alternative consciously choose for public transport, while travellers that have no car available
to them are captive.
Increasing customer satisfaction with public transport 39
Customer satisfaction by car availability
1
2
3
4
5
6
7
8
9
10
Car available (N=18,455) No car available (N=58,578)
Car availability
Sati
sfa
cti
on
/ a
pp
rec
iati
on
(m
ark
)
Customer satisfaction
Safety appreciation
Speed appreciation
Ease appreciation
Comfort appreciation
Figure 4.16 Average marks for customer satisfaction and its four factors by car availability
BivariateBivariateBivariateBivariate relationship relationship relationship relationship between travel frequency and satisfactionbetween travel frequency and satisfactionbetween travel frequency and satisfactionbetween travel frequency and satisfaction Figure 4.17 shows customer satisfaction by travel frequency. The average marks for customer satisfaction,
speed, ease and comfort appreciation decrease when trips are more frequent. However, safety appreciation
shows a weaker correlation with travel frequency. Safety appreciation is constant until five trips a week, on
average. Safety appreciation slightly decreases by a travel frequency of six times a week or more.
Customer satisfaction by travel frequency
1
2
3
4
5
6
7
8
9
10
0 or 1 2 3 4 5 6 or more
Travel frequency in the last week (trips on the line of interviewing)
Sati
sfa
cti
on
/ a
pp
rec
iati
on
(m
ark
)
Customer satisfaction
Safety appreciation
Speed appreciation
Ease appreciation
Comfort appreciation
Figure 4.17 Average marks for customer satisfaction and its four factors by travel frequency
BivariateBivariateBivariateBivariate relationship relationship relationship relationship between agebetween agebetween agebetween age and satisfaction and satisfaction and satisfaction and satisfaction
Customer satisfaction is higher for older travellers, see Figure 4.18. This statement is valid for travellers above
the age of 20. The data suggest that customer satisfaction decreases slightly from the ages of 12 up until 20.
The same bivariate relationships account for age and safety, speed, ease and comfort appreciation. Thus,
travellers above the age of 65 are more satisfied than younger travellers. For this older group, the linear trend
could account. However, a bigger increase than a linear one could account as well. The means of satisfaction for
travellers of 65 years and above show more variation compared means of younger travellers.
Increasing customer satisfaction with public transport 40
Customer satisfaction by age
5
6
7
8
9
10
10 20 30 40 50 60 70 80
age (year)
cu
sto
me
r sa
tis
fac
tio
n (
ma
rk)
Linear estimation Customer satisfaction
(R square = 0.13)
Quadratic estimation Customersatisfaction (R square = 0.17)
Observed Customer satisfaction averages
by age from data
Observed Safety appreciation averages byage from data
Observed Speed appreciation averages by
age from data
Observed Ease appreciation averages by
age from data
Observed Comfort appreciation averages
by age from data
Figure 4.18 Average marks for customer satisfaction and its four factors by age
Curve estimationCurve estimationCurve estimationCurve estimations fors fors fors for relationship relationship relationship relationship between the between the between the between the ageageageage and satisfaction and satisfaction and satisfaction and satisfaction For travellers of 65 and above, a linear trend could account. However, a bigger increase than a linear could
account as well. Means for travellers of 65 and above show more variation compared to the means of younger
travellers. Different curves between age and customer satisfaction are estimated considering all cases. The two
most promising curves are shown with the blue lines in Figure 4.18. The quadratic curve explains more variance
compared to the linear curve. However, the quadratic curve is not in line with the theory in the literature study.
Ettema (2009) expected lowest customer satisfaction around the age of 40, based on satisfaction in live. The
data show that this hypothesis does not hold. No literature is found for the quadratic curve. Too much fitting of
data needs to be prevented in linear regression models. Therefore, the linear relationship between age and
customer satisfaction and its four factors will be used in the empirical customer satisfaction model.
BivariateBivariateBivariateBivariate relationship relationship relationship relationship between gender and satisfactionbetween gender and satisfactionbetween gender and satisfactionbetween gender and satisfaction Gender contributes less to customer satisfaction, see Figure 4.19. Women on average are slightly more satisfied.
Men appreciate safety more than women do.
Customer satisfaction by gender
1
2
3
4
5
6
7
8
9
10
Man (N=32,865) Woman (N=46,606)
Gender
Sati
sfa
cti
on
/ a
pp
recia
tio
n (
mark
)
Customer satisfaction
Safety appreciation
Speed appreciation
Ease appreciation
Comfort appreciation
Figure 4.19 Average marks for customer satisfaction and its four factors by gender
4.34.34.34.3 Results Results Results Results empirical customer satisfaction modelempirical customer satisfaction modelempirical customer satisfaction modelempirical customer satisfaction model on customer satisfaction and on customer satisfaction and on customer satisfaction and on customer satisfaction and factorfactorfactorfactorssss
The previous section described bivariate relationships. In this section, these relationships will be controlled for
correlations with the input variables. The results of the empirical customer satisfaction model are discussed in
this section.
Increasing customer satisfaction with public transport 41
4.3.14.3.14.3.14.3.1 Low eLow eLow eLow explained variancexplained variancexplained variancexplained variance
Table 4.1 Results empirical customer satisfaction model based on input factors (2011)
Sign
ific
ance
P<0
.05
Ref
eren
ces:
met
ro, m
ornin
g pe
ak, o
ther
s tr
ip p
urpo
ses
Mea
nin
gfu
l tra
vel an
d tr
avel
ler’
s ch
arac
teri
stic
s ar
e m
arke
d b
old
Increasing customer satisfaction with public transport 42
The results of the empirical customer satisfaction model are shown in Table 4.1. The explained variance of the
customer satisfaction model is around 5%, which is low. Therefore, the input variables only slightly contribute
to customer satisfaction. The contributions of the input variables to customer satisfaction and the factors are
discussed below.
4.3.24.3.24.3.24.3.2 Results customer satisfactionResults customer satisfactionResults customer satisfactionResults customer satisfaction
The results of Table 4.1 are visualised in five graphs: one by customer satisfaction and four by its factors. Figure
4.20 shows the contribution of input characteristics to customer satisfaction. The bleu line is the average of
customer satisfaction. The figure includes bars for the amount of passengers, trip frequency and age classes.
These variables are analysed linear; however, the analyses are repeated with dummy variables for the classes
resulting in the visualisation. The heights of the bars are calculated based on the regression coefficients and
average for customer satisfaction. The population of a group, e.g., commuters, is considered as weighted factor
for the heights of the bars.
Results for customer satisfaction visualised for the input characteristics
6
7
8
City b
us w
ithin
G4
City b
us o
uts
ide G
4
Agglo
mora
tional bus
Regio
nal bus
Tra
m w
ithin
G4
Tra
m o
uts
ide G
4
Metr
o
Regio
nal tr
ain
Ferr
y
0-2
5 p
assengers
26-5
0 p
assengers
50-1
00
passengers
100-5
05
passengers
Morn
ing p
eak
Evenin
g p
eak
Off
-peak
Com
muting
Education
Vis
itin
g
Oth
er
trip
purp
ose
Car
available
No c
ar
available
0-1
trips p
er
week
2-4
trips
a w
eek
5 o
r m
ore
trip
s a
week
12-1
8 y
ear
18-2
7 y
ear
28-4
0 y
ear
41-6
4 y
ear
65-8
9 y
ear
Man
Wom
an
Input characteristics
Cu
sto
mer
sati
sfa
cti
on
(m
ark
1-1
0)
Figure 4.20 Contribution of input characteristics to customer satisfaction
Significant and relevant variables will be called meaningful variables. The standardized regression coefficient of
meaningful variables is 0.08 or higher. This border is set, because these values show notable higher
contribution to customer satisfaction compared to the others. The meaningful variables on customer satisfaction
are printed bold in Table 4.1: the number of passengers in the vehicle, trip frequency and age. These variables
show much variation in heights of the bars. First, the more passengers in the vehicle, the lower customer
satisfaction. Second, the more one travels, the less one is satisfied. Third, elderly people are more satisfied on
average. Furthermore, positive contributions to customer satisfaction are agglomerational bus, agglomerational
tram services, urban bus services outside the G4, car availability and the gender woman. However, not all
significant variables are relevant such as several day part and trip purpose.
Increasing customer satisfaction with public transport 43
4.3.34.3.34.3.34.3.3 Results Results Results Results safetysafetysafetysafety appreciation appreciation appreciation appreciation
Results for safety appreciation visualised for the input characteristics
7
8
9
City b
us w
ithin
G4
City b
us o
uts
ide G
4
Agglo
mora
tional bus
Regio
nal bus
Tra
m w
ithin
G4
Tra
m o
uts
ide G
4
Metr
o
Regio
nal tr
ain
Ferr
y
0-2
5 p
assengers
26-5
0 p
assengers
50-1
00 p
assengers
100-5
05 p
assengers
Morn
ing p
eak
Evenin
g p
eak
Off
-peak
Com
muting
Education
Vis
itin
g
Oth
er
trip
purp
ose
Car
availa
ble
No c
ar
availa
ble
0-1
trips p
er
week
2-4
trips a
week
5 o
r m
ore
trips a
week
12-1
8 y
ear
18-2
7 y
ear
28-4
0 y
ear
41-6
4 y
ear
65-8
9 y
ear
Man
Wom
an
Input characteristics
Safe
ty a
pp
recia
tio
n (
mark
1-1
0)
Figure 4.21 Contribution of input characteristics to the factor safety
The results by factor will be described, starting in this section. Figure 4.21 shows the contribution of input
characteristics to safety appreciation. The bleu line is the average mark for the factor. The background colour
corresponds to the colour of the factor in the research framework. Gender mainly contributes to safety. Women
feel less safe compared to men (-0.34 points on average). Agglomerational and regional bus service feel safest
compared to other modes. The number of passengers, trip frequency and age have the same influence on
safety appreciation and customer satisfaction. However, the impact of these variables is less for safety
appreciation. Thus, the elderly feel safer than young people do.
4.3.44.3.44.3.44.3.4 Results speed appreciationResults speed appreciationResults speed appreciationResults speed appreciation
Results for speed appreciation visualised for the input characteristics
5
6
7
City b
us w
ithin
G4
City b
us o
uts
ide G
4
Agglo
mora
tional bus
Regio
nal bus
Tra
m w
ithin
G4
Tra
m o
uts
ide G
4
Metr
o
Regio
nal tr
ain
Ferr
y
0-2
5 p
assengers
26-5
0 p
assengers
50-1
00 p
assengers
100-5
05 p
assengers
Morn
ing p
eak
Evenin
g p
eak
Off
-peak
Com
muting
Education
Vis
itin
g
Oth
er
trip
purp
ose
Car
available
No c
ar
available
0-1
trips p
er
week
2-4
trips a
week
5 o
r m
ore
trips a
week
12-1
8 y
ear
18-2
7 y
ear
28-4
0 y
ear
41-6
4 y
ear
65-8
9 y
ear
Man
Wom
an
Input characteristics
Sp
eed
ap
pre
cia
tio
n (
mark
1-1
0)
Figure 4.22 Contribution of input characteristics to the factor speed
Figure 4.22 shows the contribution of input characteristics to speed appreciation. The constant of speed is
remarkably lower than constants of the factors safety and ease. The three meaningful influences are the same
for speed appreciation and customer satisfaction. The transport service types have relatively high influence
compared to the other factors. Tram and urban bus service within the G4 are worst appreciated on speed. Thus,
the construct of speed appreciation is slightly in line with customer satisfaction.
Increasing customer satisfaction with public transport 44
4.3.54.3.54.3.54.3.5 Results ease appreciationResults ease appreciationResults ease appreciationResults ease appreciation
Results for ease appreciation visualised for the input characteristics
7
8
9C
ity b
us w
ithin
G4
City b
us o
uts
ide G
4
Agglo
mora
tional bus
Regio
nal bus
Tra
m w
ithin
G4
Tra
m o
uts
ide G
4
Metr
o
Regio
nal tr
ain
Ferr
y
0-2
5 p
assengers
26-5
0 p
assengers
50-1
00 p
assengers
100-5
05 p
assengers
Morn
ing p
eak
Evenin
g p
eak
Off
-peak
Com
muting
Education
Vis
itin
g
Oth
er
trip
purp
ose
Car
availa
ble
No c
ar
availa
ble
0-1
trips p
er
week
2-4
trips a
week
5 o
r m
ore
trips a
week
12-1
8 y
ear
18-2
7 y
ear
28-4
0 y
ear
41-6
4 y
ear
65-8
9 y
ear
Man
Wom
an
Input characteristics
Ease a
pp
recia
tio
n (
mark
1-1
0)
Figure 4.23 Contribution of input characteristics to the factor ease
Figure 4.23 shows the contribution of input characteristics to ease appreciation. The results are comparable with
the results regarding customer satisfaction. Ease appreciation is mainly negatively influenced by the number of
passengers. Tram and urban bus services within the G4 are highly negative appreciated on ease. Little
difference is found for the other transport service types.
4.3.64.3.64.3.64.3.6 Results comfort appreciationResults comfort appreciationResults comfort appreciationResults comfort appreciation
Results for comfort appreciation visualised for the input characteristics
6
7
8
City b
us w
ithin
G4
City b
us o
uts
ide G
4
Agglo
mora
tional bus
Regio
nal bus
Tra
m w
ithin
G4
Tra
m o
uts
ide G
4
Metr
o
Regio
nal tr
ain
Ferr
y
0-2
5 p
assengers
26-5
0 p
assengers
50-1
00 p
assengers
100-5
05 p
assengers
Morn
ing p
eak
Evenin
g p
eak
Off
-peak
Com
muting
Education
Vis
itin
g
Oth
er
trip
purp
ose
Car
available
No c
ar
available
0-1
trips p
er
week
2-4
trips a
week
5 o
r m
ore
trips a
week
12-1
8 y
ear
18-2
7 y
ear
28-4
0 y
ear
41-6
4 y
ear
65-8
9 y
ear
Man
Wom
an
Input characteristics
Co
mfo
rt a
pp
recia
tio
n (
mark
1-1
0)
Figure 4.24 Contribution of input characteristics to the factor comfort
Figure 4.24 shows the contribution of input characteristics to comfort appreciation. The results are comparable
with the results regarding customer satisfaction. The number of passengers has slightly more impact on
comfort appreciation compared to customer satisfaction. Regional train and ferry services are positively
evaluated on comfort compared to the other modes; however, the urban services in the G4 are appreciated
less. Elderly appreciate comfort much higher compared to the other age groups.
Increasing customer satisfaction with public transport 45
4.44.44.44.4 Robustness of resultsRobustness of resultsRobustness of resultsRobustness of results
This section describes the robustness of the results based on several aspects. First, robustness is verified by
different missing value imputation techniques. Second, robustness is verified by predicted values that stay
within the range. Third, the model is cross validated by another year of interviews.
4.4.14.4.14.4.14.4.1 Results robustResults robustResults robustResults robust for for for for several several several several missing value techniquesmissing value techniquesmissing value techniquesmissing value techniques
The robustness for missing values is checked. Besides pair wise deletion, missing values could also be
estimated and imputed in the data file. Thus, the estimated values could be used in the model. The results may
differ when missing values are imputed. Therefore, the analyses are done again with data files that include
estimations of missing values. Missing values are imputed by, first, linear regression estimations using the
other variables in the file and, second, maximum likelihood estimation by using the expectation–maximization
algorithm (IBM, 2008). The exact results only slightly differ by data imputation method. Moreover, the
conclusions of the models are the same for all imputation methods. Therefore, the conclusions are robust for
different methods of dealing with missing values.
4.4.24.4.24.4.24.4.2 Robust rRobust rRobust rRobust range of predictive valuesange of predictive valuesange of predictive valuesange of predictive values
This section will describe that the range of the predictive values do not provide problems by the linear
regression model. The linear regression model estimates the impact of different variables. The regression
coefficients are used to predict the marks on customer satisfaction and its four factors for every single
respondent. The predicted marks are within the range of 1 to 10 for all single respondents. This is the same
range in which respondents answer.
4.4.34.4.34.4.34.4.3 Results robust Results robust Results robust Results robust over timeover timeover timeover time
Table 4.4 Cross validation results of contributions to customer satisfaction 2009-2011
Significance P<0.05,
References: metro, morning peak, others trip purposes
Meaningful travel and traveller’s characteristics of 2011 are marked bold
Increasing customer satisfaction with public transport 46
To test the robustness of the results, data of several years of fieldwork are analysed in a cross validation. The
empirical customer satisfaction model with respect to 2011 is build for 2010 and 2009 too. Table 4.4 shows
contributions to customer satisfaction for the three years. In all years, the number of passengers, trip frequency
and age are meaningful input characteristics. Therefore, the results are robust for different years of interviews.
Moreover, Appendix F shows contributions to the four factors in 2009 and 2010. Also, these results are mainly
line with the results of 2011.
4.54.54.54.5 Conclusion results Conclusion results Conclusion results Conclusion results empirical customer satisfaction modelempirical customer satisfaction modelempirical customer satisfaction modelempirical customer satisfaction model
The travel and traveller’s characteristics explain more variance than transport service types. The results will be
highlighted in this conclusion section.
4.5.14.5.14.5.14.5.1 Main contribution by travel anMain contribution by travel anMain contribution by travel anMain contribution by travel and traveller’s characteristicsd traveller’s characteristicsd traveller’s characteristicsd traveller’s characteristics
The number of passengers, trip frequency and age mainly contribute to customer satisfaction and the
appreciations of safety, speed, ease and comfort. Customer satisfaction is higher when the number of
passengers is less, the trip frequency is less, the traveller is older and a woman. However, women appreciate
safety less compared to men.
4.5.24.5.24.5.24.5.2 LLLLessessessess appreciation in the four big cities appreciation in the four big cities appreciation in the four big cities appreciation in the four big cities
Spatial differences are found between customer satisfaction for the services within the four big cities
Amsterdam, Rotterdam, The Hague and Utrecht compared to the other services. Tram and urban bus services
within the G4 are worst appreciated in all four factors and on customer satisfaction, compared to other public
transport service types. The differences could not be fully explained by the travel and traveller’s characteristics.
The urban tram is evaluated worse than urban bus service on safety and speed within these cities: however,
tram services score slightly higher on customer satisfaction, ease and comfort compared to urban bus services.
The implications fort the tram bonus are discussed in section 5.3.1.
4.5.34.5.34.5.34.5.3 Appreciations ofAppreciations ofAppreciations ofAppreciations of transport service types transport service types transport service types transport service types are averaged out are averaged out are averaged out are averaged out
The appreciations of transport service type vary by factor. Their impact is highest on the speed appreciation.
Transport service types contribute more to the factors than to customer satisfaction. Therefore, the influences of
transport service types are mainly averaged out on customer satisfaction.
4.5.44.5.44.5.44.5.4 RRRRobust resultsobust resultsobust resultsobust results
The results are robust for different techniques of dealing with missing values and for different years. Also, the
predictive values do not exceed the borders of the answer possibilities. Therefore, the results of the empirical
customer satisfaction model are robust.
Increasing customer satisfaction with public transport 47
Figure 5.1 Reading guide chapter 5
5555
DiscussionDiscussionDiscussionDiscussion empirical empirical empirical empirical customer satisfaction customer satisfaction customer satisfaction customer satisfaction modelmodelmodelmodel
Increasing customer satisfaction with public transport 48
The results of the empirical customer satisfaction model are described in the previous chapter. This chapter
discusses the methodology and results.
Reading guide chapter Reading guide chapter Reading guide chapter Reading guide chapter 5555 The model verification will be discussed to assess the value of the results in section 5.1. Section 5.2 briefly
discusses possible improvements for the model; the most improvements show no addition value. As a result,
the results are stable. Therefore, section 5.3 discusses the results regarding several transport service types. In
fact, the tram bonus is discussed. Research and policy implications are based on the meaningful contributions of
the number of passengers, trip frequency and age in section 5.4. Finally, section 5.5 describes the conclusions
based on this chapter.
5.15.15.15.1 Model verificationModel verificationModel verificationModel verification
This section describes the verification of the model. First, it discusses the implications of the highly explained
variance by the factors. Second, questions that may explain part of the unexplained variation are introduced.
Third, the explained variance of the empirical customer satisfaction model is discussed.
5.1.15.1.15.1.15.1.1 Biases in measurements in verification modelBiases in measurements in verification modelBiases in measurements in verification modelBiases in measurements in verification model
Unexplained variance of customer satisfaction with the verification model (43%) could be based on, among
others, biases of measurement and missing aspects in the questionnaire. First, bias of measurements of
appreciation and customer satisfaction could be based on different expectations between travellers.
Appreciation can differ on expectations. Expectations can be influenced by experience of a trip in the past.
Moreover, expectations could be influenced by a public transport service of a previous concession. E.g., the
frequency drops from eight to four services per hour for a new concession in region A, while the frequency
stays four for a new concession in region B. Travellers in region A could compare the different concessions and
be less satisfied with the new one. In addition, travellers in region A could become even less satisfied
compared to the travellers in region B, because of the drop down in supply. Expectations and objective
measurements are lacking in the barometer. Therefore, supply and performance could be the same while a bias
in appreciation and given mark could still occur. In fact, different travellers can have the same opinion about the
service while they give different marks. E.g., the meaning of the mark seven could differ by traveller. Moreover,
biases in measurements may be caused by different needs, which vary by person and trip purpose (van Hagen,
2011).
5.1.25.1.25.1.25.1.2 Lacking aspects in questionnaireLacking aspects in questionnaireLacking aspects in questionnaireLacking aspects in questionnaire
This paragraph focuses on the questionnaire to verify whether the right topics are addressed. Therefore, lacking
aspects in the questionnaire will be discussed. Lack of questions is found the fifth need experience. Based on
Table E.1 in the appendix, questions about experience could be about, e.g., facilities at stops, image, colours,
scents, sounds, aesthetics of the stop and vehicle, spending time usefully, distraction and privacy. The size of
KPVV Klantenbarometer questionnaire should be restricted, so only a few questions could be added or adjusted.
Which questions could be added can be researched with a test questionnaire. Nevertheless, without the results
of a test questionnaire, some advices could be given. First, spending time usefully is a suitable indicator of a
pleasant experienced, according to van Hagen (2011). Second, image of the service is a convenient indicator of
experience. This question is also used in BEST (Fellesson, 2009). In addition, it also indicates expectations of the
traveller. Furthermore, another lacking aspect is customer loyalty, see the argumentation in the first chapter.
The new questions are reported below.
Questions to add regarding traveller’s experience are:
■ ‘What is your opinion regarding the image of this service?’ Axes: very unpleasant – very pleasant (BEST,
2012).
■ ‘I spend my time usefully during the travel?’ Axes: fully untrue – fully true.
Increasing customer satisfaction with public transport 49
Customer loyalty could be measured by the loyalty question of BEST;
■ ‘I gladly recommend travelling with PT to others.’ Axes: fully untrue – fully true (BEST, 2012).
5.1.35.1.35.1.35.1.3 Interpersonal varianceInterpersonal varianceInterpersonal varianceInterpersonal variance
The explained variance of the model based on input characteristics is low. Therefore, the empirical customer
satisfaction model contains much bias. The bias could be explained by two reasons. First, the empirical customer
satisfaction model may not contain the right variables. The right variables could be appreciations on sub aspects
of customer needs. These appreciations are no input characteristics in the empiric customer satisfaction model.
Other lacking variables could relate to variations between services. The journey satisfaction model (section 2.2)
described that customer satisfaction is based on, among others, the supply of the current journey and traveller’s
expectations. Variables that describe the supply of the current journey and variations in services are lacking in
the survey. The survey considers only variations by the number of passengers (indicator for crowding) regarding
the supply. Lacking variables that could infirm this problem could be, e.g., the objective travel time, frequency,
measured punctuality, seat occupation and noise measurements. In addition, cleanness, driving style, personnel
friendliness could be scored by the interviewers. Second, the bias may caused by interpersonal variance.
Interpersonal variance is the bias of judgements of different travellers, which are offered exactly the same
supply at the same time. Interpersonal variance could be based on, e.g., experience during previous trips, image
because of media attention, traveller’s mood, impact of weather on mood and other external influences (Koots
et al, 2011; Zeithaml et al., 1993). The expert vision of Dick Ettema − traveller’s satisfaction researcher at
University of Utrecht − is that these aspects could explain slightly more variance. However, still many unknown
variables cannot be captured in the model regarding travel and traveller’s characteristics. Thus, the model could
be improved to only a certain extent if more trip specific aspects are researched.
5.25.25.25.2 Discussion of classifications and Discussion of classifications and Discussion of classifications and Discussion of classifications and possible improvementspossible improvementspossible improvementspossible improvements
This section checks the results on robustness for methodological choices before discussing policy implications
with respect to the number of passengers, trip frequency and age. Table 5.1 reports the proposed
improvements. They are tested if possible. The effects on the explained variance are described. The last column
report whether the proposal was useful. Almost no proposed improvements were useful. No improvement
changed the main conclusions with respect to the national level. The spatial classification and trip purposes are
discussed in more detail below.
Table 5.1 Possible improvements
Proposed improProposed improProposed improProposed improvementvementvementvement EffectEffectEffectEffect R square R square R square R square
changechangechangechange
Useful Useful Useful Useful
proposalproposalproposalproposal
Improvement of spatial classification based
on population density
Small effect, results are robust for different classifications < 1.5 % No
Different classifications of trip purposes Results are robust for different classifications No
Different classifications of day parts No effect, results are robust for different classification No
S or other curve between crowding and
customer satisfaction
No logic results, see section 4.2.2 <0.5 % No
Different curves for age No other curve is evident better than a linear relation.
Other curves are fitting on data, theory is lacking. See
section 4.2.3
<0.5% No
Interaction transport service type x trip
frequency to customer satisfaction
Not significant, see Appendix H No
Interaction transport service type x age to
customer satisfaction
Not significant, see Appendix H No
Increasing customer satisfaction with public transport 50
Interaction transport service type x amount
of passengers to customer satisfaction
Crowding needs to measured better before interaction has
a meaning, see section 4.2.1
< 0.4 % No
Improved measurement of crowding (with
chip card data)
Not tested, see discussed in section 5.4.1 Probably
Include non public transport users in the
survey
Not tested, promising method based on BEST (Fellesson,
2009)
Probably
Considering more supply characteristics
that can explain and control for variation in
performed quality.
Not tested Possibly
Correcting for social well-being Not tested on barometer. Ettema et al. (2011) explained
slightly more variance of car traveller’s satisfaction (2%
above 20-30%).
Slightly
Analyses of contribution of sub aspects to
customer satisfaction by individual
concession
Results for regions can differ from nation width analyses.
Therefore, individual analyses are convenient
improvements, see Appendix H
Yes
Analyses of contribution of supply, travel
and traveller’s characteristics to customer
satisfaction by individual concession
Results for HTM did not differ from nation width analyses.
However, an analysis about a specific concession is seen
as a trustful supporting of marketing plans and policies.
Yes
5.2.15.2.15.2.15.2.1 No effects by different spatial classificationsNo effects by different spatial classificationsNo effects by different spatial classificationsNo effects by different spatial classifications
One could discusses the spatial structure in the pre analyses in chapter 3. E.g., the agglomeration around
Amsterdam has another urbanisation degree, offered services and travellers demands compared to towns
around the Green hart (in Dutch: het Groene Hart). Therefore, it might be useful to split the Randstad in more
detail. The robustness of spatial classification of chapter 3 is tested as follows. The data consist of 87 research
areas. A research area is not larger than a concession area. A research area is more specific, because the
modality is defined too. For 86 areas a dummy is made. These dummies replaced the eight dummies of
transport service types in the empiric customer satisfaction model. The additional explained variance is no more
than 1.5% on customer satisfaction and the four appreciation factors. This result may overestimated by change
capitulation due to the many dummies. Moreover, the additional explained variance is low. The dummies that
can be interpreted as constant for the research areas. A different classification would be based on a different
grouping of the 87 research areas. Therefore, a different classification of the research areas could only explain
less than 1.5 % additional variance. Therefore, the classification of concession areas is of minor importance in
explaining contribution to customer satisfaction.
5.2.25.2.25.2.25.2.2 Focus on improvements for students, commuters and Focus on improvements for students, commuters and Focus on improvements for students, commuters and Focus on improvements for students, commuters and scholarsscholarsscholarsscholars
The policy of an operator could be improving customer satisfaction of traveller groups that give low marks. Trip
purpose is a relatively easy aspect of categorizing travellers by sight. The trip purpose of a traveller can be
estimated by visual aspects such as age, luggage, day part and group size. This policy has several advantages.
First, the most unsatisfied group could become satisfied, spread better word-of-mouth publicity and may
become a loyal customer. Second, the average satisfaction mark could increase which could result in earning a
bonus. Appendix G shows that operators should focus on students, commuters and scholars to improve
customer satisfaction and sub aspects that can be influenced by operators such as personnel friendliness,
cleanness and driving style.
5.35.35.35.3 Discussion of Discussion of Discussion of Discussion of tram bonustram bonustram bonustram bonus
This section discusses the differences in the appreciations of tram and bus services, which results in a discussion
regarding the tram bonus.
Increasing customer satisfaction with public transport 51
5.3.15.3.15.3.15.3.1 Tram is higher appreciated after correction compared to busTram is higher appreciated after correction compared to busTram is higher appreciated after correction compared to busTram is higher appreciated after correction compared to bus
The univariate results showed higher averaged marks for urban bus services compared to tram services (2012).
However, the results after correction for the travel and traveller’s characteristics show a different order in
appreciation. The urban scores higher customer satisfaction, appreciation of ease and comfort compared to the
urban bus. These results are in line with many other studies, as described in Bunschoten (2012). Therefore,
possible sources with respect to the tram bonus are described in more detail below.
5.3.25.3.25.3.25.3.2 Higher tram appreciation based on seat availabilityHigher tram appreciation based on seat availabilityHigher tram appreciation based on seat availabilityHigher tram appreciation based on seat availability
The analyses for HTM showed that the urban tram is appreciated above the urban bus on seat availability,
driving style, frequency and punctuality, see Appendix I. Seat availability explains the most of the higher
appreciation of the tram. In fact, one could argue that tram services are appreciated more due to the seats
capacity of the bus. Another explanation could be that tram services are used for shorter trips compared by bus.
Therefore, tram travellers may expect less seat availability. The trip length of travellers is not interviewed;
therefore, this explanation could not be checked in this research. The results of this analysis are compared to
the conclusions of Bunschoten’s tram bonus study (2011). He concluded that the tram bonus is based on
atmosphere in the vehicle, characteristics of the vehicle and travel information. The similarity between the
results of the HTM analyses and Bunschoten is that the tram bonus is partly dependent on some vehicle
characteristics.
5.45.45.45.4 Policy implications regarding thePolicy implications regarding thePolicy implications regarding thePolicy implications regarding the numbernumbernumbernumber of passenge of passenge of passenge of passengers, travel frequency andrs, travel frequency andrs, travel frequency andrs, travel frequency and ageageageage
The main results show that the number of passengers and travel frequency negatively contribute and ages
positive contributes to customer satisfaction and its four factors. Policy implications following from the results
are discussed in this section. The goals are to increase the number of travellers and satisfied travellers.
5.4.15.4.15.4.15.4.1 The nThe nThe nThe numberumberumberumber of passengers of passengers of passengers of passengers
This sub section discusses the bias in the measurement of crowding, how this can be solved and policy
implications following form the number of passengers.
Bias in measurements of Bias in measurements of Bias in measurements of Bias in measurements of crowdingcrowdingcrowdingcrowding The number of passengers is reported as the number of passenger entering a vehicle between first and last
stop of a served line. Therefore, the number of travellers is constant for everyone in the same vehicle during
the operation of one line from arrival to departure. It is an indicator for crowding in the vehicle during the
interview. However, the passengers entering in total measures crowding in a vehicle during the moment of
interviewing not exact, because location of boarding and deboarding is not considered. As a result, spread of
passengers is not considered. The number of passengers could be interpreted as crowding with bias for the
exact number of passengers in the vehicle. Also, line lengths vary. A long line could have more boarding
passengers compared to a short line, while the occupancy rate could be the same for both lines. In addition, the
maximal capacity is not reported either. As a result, the number of passengers could not be divided by the
maximum capacity. One could argue that maximum capacity by vehicle could be estimated knowing the
transport service type. As a matter of fact, the implicit assumption is made that the same transport service
types has the same capacity, which is only partly true.
Measuring dynamic traveller’s occupancyMeasuring dynamic traveller’s occupancyMeasuring dynamic traveller’s occupancyMeasuring dynamic traveller’s occupancy Crowding is an important variable and is measured with much bias. Therefore, the number of travellers is
important during the trip of an interviewee, instead of the total number of passengers of a line. A method to
solve this problem will be described. The number of boarding and deboarding passengers is known, based on
data of public transport chip card. Crowding between stops can be computed if the following is known: the
Increasing customer satisfaction with public transport 52
number of boarding and deboarding passengers by stops and the capacity of the vehicle. Now the location of an
interviewee needs to be considered. One option could be using a tablet pc instead of paper to fill in the
questionnaire. The location and time can be recorded using GPS coordinates when the questionnaire is being
filled in. Based on the GPS coordinates, the location between stops can be specified. The location between stops
can be linked to the number of passengers in that vehicle between that stops. Several alternatives can be
discussed from this, which is too much detail for this report.
Policy implications Policy implications Policy implications Policy implications regarding the regarding the regarding the regarding the numbernumbernumbernumber of passengers of passengers of passengers of passengers The number of passengers contributes meaningfully to customer satisfaction. Therefore, it is important to
prevent crowding. It can be prevented by spreading travellers. Travellers can mainly be spread by increasing the
service reliability (van Oort, 2011). First, plan the service such that the headways between the vehicles are
equal. Second, prevent early departures. Third, real time traffic management: keep the headways between
vehicles the same, even if a small delay occurs regarding high frequent services (Wardman & Whelan, 2010).
Fourth, ensure realistic time tables especially regarding low frequent services. Therefore, measure the travel
times of a previous comparable period and use the 35 % percentile to adjust the time table information. Fifth,
provide multiple lines on shared track if a part of a service is much used. Sixth, hold vehicles at waiting stops if
travel patronages are splitted at that stop. Seventh, split lines if travel patronages are splitted at a stop. Two
improvements to spread travellers without improving service reliability are as follows. First, attract peak
travellers to the off-peak by discounts and marketing. Experiments have shown peak travellers may become
off-peak travellers by price incentives (Spitsmijden, 2009). A peak ticket may be five times expensive as a off-
peak ticket. Second, the frequency on the busiest routes may be increased: use more vehicles on the busiest
routes and less on routes with less demand.
5.4.25.4.25.4.25.4.2 Travel frequencyTravel frequencyTravel frequencyTravel frequency
The result chapter showed, the more one travels, the less one is satisfied. Possible reasons for this negative
influence could be, first, having no alternatives and, second, the impact of negative situations in the past. Both
reasons are argued in chapter 2.
Policy implications Policy implications Policy implications Policy implications regarding regarding regarding regarding travel frequencytravel frequencytravel frequencytravel frequency: more attention for frequent travellers: more attention for frequent travellers: more attention for frequent travellers: more attention for frequent travellers Policy implications will be given based on the results about travel frequency and the theoretical foundation.
Improvements should be focussed on frequent travellers, because they show the lowest satisfaction. How this
can be done is discussed among others with HTM, see Appendix I. First, give more attention (in Ducth:
knuffelaandacht) to frequent travellers, e.g., month/year card holders and other frequent travels could get a
small HTM present for their birthday. In addition, the Dutch Railways letter regarding extinction of the
month/year card could be more polite and on the right moment not being in the summer holiday. Second, the
HTM analyses also showed that frequent travellers could become more satisfied when the vehicles are less
crowded and when delay information is improved, e.g., by displays at the stops and a smart phone app. When
delays occur and the vehicle is crowded, the situation should be managed well. Make sure that travellers get
more and better information from the operator instead of social media. Convenient situational management
should prevent that travellers quit using the service after a negative experience. Furthermore, situational
management may be valuable for the customer satisfaction of all travellers.
Policy implications travel frPolicy implications travel frPolicy implications travel frPolicy implications travel frequency: prevent negative experiences equency: prevent negative experiences equency: prevent negative experiences equency: prevent negative experiences The results can also bee seen from a different perspective. The average mark of customer satisfaction can be
increased by attract incidental passengers. Incidental travellers give higher grades on customer satisfaction,
which could result in a bonus for the operator. Incidental travellers could be attracted by marketing and
communication focussed on comfort and experience. After this, another challenge occurs: try to keep incidental
travellers to increase the revenue. Therefore, marketing is needed too. Comfort and experience aspects need to
be exploited (van Beek et al., 2008; van Hagen, 2011). One might argue that customer satisfaction decrease
when incidental travellers become frequent travellers. In the author’s opinion, this is not the case, as long as
the new frequent travellers do not have negative experience. Therefore, the situational management needs to
Increasing customer satisfaction with public transport 53
function well in advance. Thus, to keep incidental and frequent travellers, negative experiences should be
prevented.
5.4.35.4.35.4.35.4.3 AgeAgeAgeAge
Policy implicationsPolicy implicationsPolicy implicationsPolicy implications regarding regarding regarding regarding ageageageage The results in chapter 4 show that the travellers younger than 40 are less satisfied compared to the older
travellers. Therefore, improvements should be focussed on travellers until the age of 40 years. E.g.,
improvements for this group of travellers could be actual travel information with an app, beach express and
reduced tickets for young families.
Market segmentationMarket segmentationMarket segmentationMarket segmentation The contribution of person characteristics such as age and gender argues for market segmentation in urban,
agglomerational and regional public transport. Market segmentation is valuable for marketing. Because travel
and traveller’s characteristics contribute more to customer satisfaction than spatial characteristics, marketing
should aim at the classification of travellers instead of regions. The text below discusses market segmentations
that are known. Market segmentation is made for national train service by NS and NeedScope (van Hagen,
2011). Sigger (2011) applied the NeedScope methodology on urban and regional public transport. The
segmentation is based on two axes regarding traveller’s characteristics: introversion−extraversion and reaching
a goal−necessary evil. Six segments of travellers are based on the axes: the explorer, individualist, functional
planner, certainty seeker, convenience seeker and socializer. The sizes of the market segments are calculated
by comparing the distribution of trip purposes in train and urban and regional public transport. The six segments
for urban and regional public transport and the sizes of the segments are not yet established by other research.
The results of chapter 4 show that it can be researched whether characteristics such as crowding, travel
frequency and age need to be used as axes for market segmentation.
5.55.55.55.5 Conclusion discussion Conclusion discussion Conclusion discussion Conclusion discussion empirical customer satisfaction modelempirical customer satisfaction modelempirical customer satisfaction modelempirical customer satisfaction model
5.5.15.5.15.5.15.5.1 Probably much interpersonal varianceProbably much interpersonal varianceProbably much interpersonal varianceProbably much interpersonal variance
Many improvements are tried; however, the contribution of transport service types, travel and traveller’s
characteristics stays weak. In contrast, the explained variance of customer satisfaction is much by the sub
aspects appreciations. Therefore, travellers mainly judge by their perceived quality of the aspects Therefore,
travellers judge mainly by their perceived quality of the sub aspects. Interpersonal variance probably explains
the deviation in marks by the same offered supply.
5.5.25.5.25.5.25.5.2 Research about differences within travellResearch about differences within travellResearch about differences within travellResearch about differences within travellers ers ers ers
The differences within travellers are unknown. As a result, a research implication is to investigate the variance
of customer satisfaction within frequent travellers that are mainly using one line. In that case, the research will
show that the variance is high within travellers; therefore, policy implication would be ensuring stability in
supply. Therefore, standardisation of supply by high punctuality (van Oort, 2011) and highly comparable
behaviour of personnel would be advised. Moreover, standardisation is possible by the same courses for every
driver, and the same courses for stuarts. In case, the research shows that the variance is low within travellers,
than can be concluded that the judgements are stable. In that case, the judgements are independent of small
variations in supply. In that case, small differences in supply may not affect customer satisfaction. That may
mean that customer satisfaction could be influenced by marketing.
5.5.35.5.35.5.35.5.3 Questions about experience and loyaltyQuestions about experience and loyaltyQuestions about experience and loyaltyQuestions about experience and loyalty
Several other research implications are found. Lacking questions in the barometer questionnaire are about the
top factor of the pyramid of customer needs defined as experience. Two recommended questions to cover this
factor are about the image of the service and usefulness of the spend time. In addition, a question regarding
customer loyalty is proposed.
Increasing customer satisfaction with public transport 54
5.5.45.5.45.5.45.5.4 Dynamic occupancy measuremDynamic occupancy measuremDynamic occupancy measuremDynamic occupancy measurementsentsentsents
Another research implication is regarding dynamic occupancy measurements. A linear contribution of the
number of passengers of a line is found. Even with a lot of bias, crowding contributes to customer satisfaction. It
is worthily to investigate the contribution in more detail. Therefore, the occupancy during the journey of an
interviewed traveller is important instead of the total number of passengers boarding of a line. Dynamic
occupancy rates can be collected form PT chip card data.
5.5.55.5.55.5.55.5.5 Improvements for the less satisfied travellersImprovements for the less satisfied travellersImprovements for the less satisfied travellersImprovements for the less satisfied travellers
Form the policy implications can be concluded that the improvements should focus on preventing crowding,
frequent travellers and travellers younger than 40 years. First, crowding may be prevented by spreading
passengers. A policy implication is to improve service reliability to spread the travellers over vehicles as
described. Moreover, peak travellers may be attracted to the off-peak by discounts and marketing. In addition,
the frequency may be increased on the busiest routes at the expenses of routes with less demand. Second, the
satisfaction of frequent travellers may be improved by better actual travel information at displays at stops and
on a smart phone app. Negative experiences need to be management well if they could not be prevented.
Third, Frequent travellers could be satisfied more by small attentions. Fourth, marketing should be focused on
travellers younger than 40 years.
Increasing customer satisfaction with public transport 55
Figure 6.1 Reading guide chapter Figure 6.1 Reading guide chapter Figure 6.1 Reading guide chapter Figure 6.1 Reading guide chapter 6666
6666
Conclusions and Conclusions and Conclusions and Conclusions and recommendations to recommendations to recommendations to recommendations to increase customer increase customer increase customer increase customer satisfactionsatisfactionsatisfactionsatisfaction
Increasing customer satisfaction with public transport 56
Reading guide conclusions and recommendationsReading guide conclusions and recommendationsReading guide conclusions and recommendationsReading guide conclusions and recommendations This chapter provides answers to the sub questions (section 6.1) that together answer the main research
question (section 6.2). The main research question has been formulated as: what variables contribute to
customer satisfaction of public transport services in the Netherlands and to what extent? Research and policy
implications are followed by the conclusions and discussions. Research implications will be described in section
6.3. Finally, in section 6.4, policy implications that contribute to the main research goal will be described.
6.16.16.16.1 Answers sub questionsAnswers sub questionsAnswers sub questionsAnswers sub questions
6.1.16.1.16.1.16.1.1 What is customer What is customer What is customer What is customer satisfaction withsatisfaction withsatisfaction withsatisfaction with publ publ publ public tranic tranic tranic transsssport?port?port?port?
Customer satisfaction with public transport is defined as the degree to which an individual positively evaluates
the overall quality of a public transport service. Customer satisfaction of a public transport journey is the result
of an evaluation of offered supply by a public transport operator and authority compared to traveller’s
expectations. Traveller’s expectations are influenced by the promised supply. In addition, customer’s
satisfaction with a journey is positively influenced by the satisfaction of previous journeys. Furthermore,
customer satisfaction could be influenced both positively and negatively by media attention. Travellers have
five main needs: safety, speed, ease, comfort and experience. Comfort and experience are satisfiers. When
satisfiers are evaluated sufficiently, new travellers can be attracted. Travellers consider the five described needs
in the evaluating satisfaction.
6.1.26.1.26.1.26.1.2 WWWWhat hat hat hat variablevariablevariablevariables cos cos cos connnntribute to customer satisfaction of public transport services according tribute to customer satisfaction of public transport services according tribute to customer satisfaction of public transport services according tribute to customer satisfaction of public transport services according to the lito the lito the lito the litttteeeerature?rature?rature?rature?
Concluded from many customer satisfaction researches can be that the most important supply appreciation sub
aspects are travel time and punctuality. Many other sub aspects have low contributions to customer satisfaction
individually; however, together, many appreciation sub aspects properly explain customer satisfaction. Supply
characteristics are offered transport service types (combination of region and modality). According to the
literature, metro is appreciated more than tram; tram more than bus. Travel characteristics are the amount of
passengers, day part and trip purpose. According to the literature, crowding contributes negatively to comfort
and customer satisfaction; commuters and students have lower customer satisfaction compared to social-
recreational travellers. Traveller’s characteristics are car availability, travel frequency, age and gender.
According to the literature, car availability may contribute positively to customer satisfaction; travel frequency
may contribute negatively or positively to customer satisfaction; age may contribute positively; women feel
less save compared to men.
Less is known about the contribution of travel and traveller’s characteristics to customer satisfaction and to
appreciation of factors in the pyramid for urban, agglomerational and regional public transport services in the
Netherlands. Therefore, this is researched with the empirical customer satisfaction model.
6.1.36.1.36.1.36.1.3 How should How should How should How should the the the the variables be variables be variables be variables be clusteredclusteredclusteredclustered in a model in a model in a model in a model????
Supply characteristics are spatial areas, transport services and operators. Because of the relationship between
spatial area and transport service, both are combined to transport service type. A spatial distinction of customer
satisfaction is found between the four biggest cities and other areas of the Netherlands. In addition, some
social recreational trip purposes could be clustered. Off-peak day parts could be clustered too. Sub aspects are
clustered in to four appreciation factors: safety, speed, ease and comfort. Travel speed, frequency, personnel
friendliness, driving style and cleanness contribute most to customer satisfaction.
Increasing customer satisfaction with public transport 57
6.1.46.1.46.1.46.1.4 What What What What dodododo contribution of supply, travel and traveller’s characteri contribution of supply, travel and traveller’s characteri contribution of supply, travel and traveller’s characteri contribution of supply, travel and traveller’s characterisssstics tics tics tics contribute contribute contribute contribute to to to to customer satisfaction and to the four appreciation customer satisfaction and to the four appreciation customer satisfaction and to the four appreciation customer satisfaction and to the four appreciation factorfactorfactorfactors?s?s?s?
The model based on supply, travel and traveller’s characteristics explains variances in customer satisfaction and
its four factors safety, speed, ease and comfort weakly. However, the three variables that meaningfully
contribute to customer satisfaction are: the number of passengers, trip frequency and age.
Contribution of supply characteristicsContribution of supply characteristicsContribution of supply characteristicsContribution of supply characteristics The contribution of supply characteristic to customer satisfaction is weak after controlling for travel and
traveller’s characteristics. On a national level, transport services are appreciated in the following order:
agglomerational bus, agglomerational tram, regional bus, ferry, city bus outside G4, regional train, metro, tram
urban within, city bus within G4. Therefore, transport services are appreciated less in the four big cities.
Transport service type can influence 0.8 points of customer satisfaction at most. Transport service types
contribute higher to the four appreciation factors compared to customer satisfaction. Therefore, most
(dis)advantages of transport services types are averaged out in customer satisfaction.
Contribution of travel characteristicsContribution of travel characteristicsContribution of travel characteristicsContribution of travel characteristics The number of passengers in a vehicle contributes negatively to customer satisfaction and the four appreciation
factors. The influence of the number of passengers on customer satisfaction can be around 1 point out of ten,
on appreciation of ease 2 points and on appreciation of comfort around 1.5 points. In contrast, influence of day
part and trip purposes is low. Customer satisfaction is slightly higher during morning peak compared to other
day parts. Furthermore, customer satisfaction is slightly lower for trip purposes commuting and education
compared to social recreational purposes.
Contribution of traveller’s characteristicsContribution of traveller’s characteristicsContribution of traveller’s characteristicsContribution of traveller’s characteristics Frequent travellers are less satisfied compared to incidental travellers; the difference is up to 0.35 points on
customer satisfaction. Moreover, trip frequency contributes relevantly to the factors speed, ease and comfort. In
addition, age contributes relevantly to customer satisfaction. A higher age is associated with higher customer
satisfaction. Age can count for a difference of around 0.8 points on customer satisfaction. In addition,
contribution of age is relevant for the explanation of the factors speed and comfort. Furthermore, gender
contributes to safety. On average, women appreciate safety 0.3 point lower compared to men. In contrast,
women appreciate customer satisfaction and the factors speed, ease and comfort slightly higher than men do.
Furthermore, car availability shows a small contribution to customer satisfaction and its factors. Therefore,
travellers that make deliberate choices are slightly more satisfied compared to captives.
6.1.56.1.56.1.56.1.5 What implications follow form the results?What implications follow form the results?What implications follow form the results?What implications follow form the results?
The results implicate that customer satisfaction mainly depends on the appreciations of the sub aspects of the
offered supply and, perhaps, on interpersonal variance. The interests of sub aspects can differ by concession
area. Dynamic occupancy measurements are recommended. Improvements need to be done for travellers that
are least satisfied: frequent travellers and young travellers. In addition, crowding needs to be prevented.
6.26.26.26.2 Answer to Answer to Answer to Answer to the main the main the main the main research question: what research question: what research question: what research question: what variablevariablevariablevariables contribute to customer s contribute to customer s contribute to customer s contribute to customer satisfaction satisfaction satisfaction satisfaction withwithwithwith public transport services in the Nethe public transport services in the Nethe public transport services in the Nethe public transport services in the Netherrrrlands and to whlands and to whlands and to whlands and to what at at at extent?extent?extent?extent?
The answer to the main research question is given in this section. Customer satisfaction shows much deviation.
The main conclusion is that much interpersonal bias occurs. Furthermore, the variables that contribute the most
to customer satisfaction with public transport are the appreciations of the customer needs. Appreciations are
influenced by the offered supply, expectations, experiences during previous journeys and media attention. Most
appreciated needs are, in order of importance: speed, comfort, ease and safety. In more detail, the
appreciations of travel speed and frequency contribute most to customer satisfaction, followed by personnel
Increasing customer satisfaction with public transport 58
friendliness, driving style and cleanness. However, results show that the order of these interests can differ by
concession.
Travel and traveller’s characteristics contribute little to customer satisfaction. Customer satisfaction is higher
when the number of passengers is less, the trip frequency is less, the traveller is older and a woman. The
number of passengers can influence customer satisfaction to 1.0 point; age until 0.8 points and trip frequency
until 0.35 points. Transport service types contribute less to the overall costumer satisfaction; however, they
contribute to appreciations of travellers needs. Therefore, most (dis)advantages of transport services types are
averaged into customer satisfaction.
Thus, main contributions to customer satisfaction are interpersonal bias, the offered supply regarding speed and
comfort and the evaluation of the offered supply by traveller’s expectations.
6.36.36.36.3 Research implicationsResearch implicationsResearch implicationsResearch implications
The results are translated into useful recommendations for operators and authorities. Some recommendations
given are directly based on the conclusions (section 6.4). Other conclusions lead to research implications as
described below.
6.3.16.3.16.3.16.3.1 Research interpersonal varianceResearch interpersonal varianceResearch interpersonal varianceResearch interpersonal variance
Little variance of customer satisfaction is explained by supply, travel and traveller’s characteristics. Still a lot of
variation in appreciations occurs within a concession and vehicle. Reason may be much interpersonal variance
or variance in offered supply. Therefore, interpersonal variance may occur. Interpersonal variance is bias of
judgements of different travellers, which are offered exactly the same supply at the same time. Interpersonal
variance could be based on, e.g., experienced by previous trips, image by media attention, traveller’s mood,
impact of weather on mood and other external influences. As a result, research implication is to investigate the
variance of customer satisfaction of frequent travellers using mainly one line. In the case that the research
shows that variance between travellers in the same vehicle is large, it means that a group of travellers observe
small deviations in service. In that case, policy implication would be ensuring stability in supply. Therefore,
standardisation of supply by high punctuality and highly comparable behaviour of personnel would be advised.
Moreover, standardisation is possible by the same courses for every driver, and the same courses for stuarts. In
case the research shows that the variance between travellers is small, than can be concluded that the
judgements are stable. In that case, the judgements are independent of small variations in supply. That mean
that small differences in supply may not affect customer satisfaction.
6.3.26.3.26.3.26.3.2 Dynamic Dynamic Dynamic Dynamic crowdingcrowdingcrowdingcrowding
Crowding is an important variable and is measured with much bias. Therefore, crowding is a topic for further
research. The exact number of travellers during the trip of an interviewee is unknown, however, important.
Research question would be: what is the effect of critical (seat) occupation on customer satisfaction and travel
behaviour? A method to reduce the bias is linking the location of the interviewees, e.g. by tablets including GPS
devices, by the crowding based on chip card data.
6.3.36.3.36.3.36.3.3 Market segmentationMarket segmentationMarket segmentationMarket segmentation
The results regarding travel and traveller’s satisfaction support the need market segmentation in urban,
agglomerational and regional public transport. Market segmentation is, e.g., valuable for marketing. The results
show that characteristics such as crowding, travel frequency and age may be used as axes for market
segmentation.
Increasing customer satisfaction with public transport 59
6.46.46.46.4 Policy implicationsPolicy implicationsPolicy implicationsPolicy implications
This section describes policy implications for operators and authorities to increase customer satisfaction cost
efficiently.
6.4.16.4.16.4.16.4.1 Prevent crowdingPrevent crowdingPrevent crowdingPrevent crowding
Crowding can be prevented by spreading travellers. Travellers can mainly be spread by increasing the service
reliability. First, plan the service such that the headways between the vehicles are equal. Second, prevent early
departures. Third, real time traffic management: keep the headways between vehicles the same, even if a
small delay occurs regarding high frequent services. Fourth, ensure realistic time tables especially regarding low
frequent services. Therefore, measure the travel times of a previous comparable period and use the 35%
percentile to adjust the time table information. Fifth, provide multiple lines on shared track if a part of a service
is much used. Sixth, hold vehicles at waiting stops if travel patronages are splitted at that stop. Seventh, split
lines if travel patronages are splitted at a stop. Two improvements to spread travellers without improving
service reliability are as follows. First, attract peak travellers to the off-peak by discounts and marketing.
Second, the frequency on the busiest routes may be increased at the expenses of routes with less demand.
6.4.26.4.26.4.26.4.2 Earning the bonusEarning the bonusEarning the bonusEarning the bonus
Higher marks on customer satisfaction can be reached by attracting more incidental travellers, woman and
elderly. To prevent problems with social safety, incidentally travelling elderly woman could best be attracted
during the day time outside the peak. However, it is better to improve customer satisfaction for the travellers
who are less satisfied.
6.4.36.4.36.4.36.4.3 IncreaIncreaIncreaIncreasing customer satisfaction of frequent travellerssing customer satisfaction of frequent travellerssing customer satisfaction of frequent travellerssing customer satisfaction of frequent travellers
The focus of improvements needs to be on frequent travellers and young travellers, especially commuters
below forty years, students and scholars. Therefore, marketing should be focused on travellers younger than
forty years. Frequent travellers could be satisfied more by small attentions. Besides preventing crowding, the
satisfaction of frequent travellers may be improved by better actual travel information at displays at stops and
on a smart phone app. Negative situations need to be managed well when they occur: provide the right
information to travellers and keep them up to date through the operator. Therefore, a robust level of supply is
needed: high punctuality and customer-friendly behaviour of staff. Thus, services may be improved, mainly on
travel time, frequency, punctuality, customer friendliness and driving style. The last two improvements can be
cost efficient. Moreover, customer satisfaction can increase by lower traveller’s expectations. Therefore, promise
only expectations that can be delivered.
6.4.46.4.46.4.46.4.4 Policy implications from expert visionPolicy implications from expert visionPolicy implications from expert visionPolicy implications from expert vision
More policy implications are incorporated in the epilogue that contains the expert vision of the author, see
section 7.3 and 7.4. A summary is given. First, be polite to students, despite they travel without paying mostly.
The operator and their personnel should build a committed relationship with the traveller. Second, investigate
the most important traveller’s needs and match the traveller’s needs by qualities of the staff in the planning.
Third, the operator and authority should control the service and customer satisfaction by analyses by line and
concession area. Last but not least, make public transport a positive experience, e.g., by vehicle design and
campaigns.
Increasing customer satisfaction with public transport 60
Figure 7.1 Reading guide chapter 7
Reading guide epilogue: expert vision customer satisfaction with public transportReading guide epilogue: expert vision customer satisfaction with public transportReading guide epilogue: expert vision customer satisfaction with public transportReading guide epilogue: expert vision customer satisfaction with public transport The epilogue discusses the expert vision that is developed by the author during the project by sharing visions
with experts, the literature review and data analyses. This epilogue deals with several interesting aspects, as
shown in Figure 7.1. First, the epilogue describes what customer satisfaction is in the author’s opinion. Second,
a convenient way of measuring customer satisfaction is described. Third, improvements based on
measurements are described. Fourth, policy implications are given as a conclusion of the expert vision.
7777
Epilogue: expert Epilogue: expert Epilogue: expert Epilogue: expert vision customer vision customer vision customer vision customer satisfactionsatisfactionsatisfactionsatisfaction
Increasing customer satisfaction with public transport 61
7.17.17.17.1 What customer satisfaction is in public transport accorWhat customer satisfaction is in public transport accorWhat customer satisfaction is in public transport accorWhat customer satisfaction is in public transport according to the authording to the authording to the authording to the author
Customer satisfaction is a feeling. This feeling is strongly based on the comfort and speed of the journey.
Feelings could be influenced by experiences with previous trips, expectations based on promised supply and
media attention. Experiences with previous journeys can be positive and negative. Negative experiences such
as delay and crowding of previous journeys influence the satisfaction of the current trip. Negative aspects have
seven times more impact on people than positive ones. Therefore, seven positive experiences are needed to
compensate one negative. Therefore, it is worthwhile to prevent small as well as big negative experiences.
In the author’s opinion, no hierarchy exists in customer needs on individual level as suggested in chapter 2.
Dick Ettema, expert on traveller’s satisfaction at Univeriteit Utrecht, as well as Wilco van Dijk, expert on
customer psychology at Univeristeit Leiden, share the same vision. It means that sub aspects can compensate
each other. Therefore, linear regression was the advised method by Dick Ettema and Wilco van Dijk. Matthijs
Dicke-Ogenia, expert on mobility psychology at Goudappel and PhD researcher at Rijksuniversiteit Groiningen,
added that people do not have a problem with giving an average judgement considering 15 sub aspects or
even more. The next sub section describes the expert vision of the author whereby the factors of the pyramid
are changed in blocks next to each other. Also, the section considers the impact on usage.
7.1.17.1.17.1.17.1.1 Customer satisfactCustomer satisfactCustomer satisfactCustomer satisfaction and public transport useion and public transport useion and public transport useion and public transport use
Goal of improving customer satisfaction is the continuity of public transport by sufficient patronage and loyal
customers. Customer satisfaction affects the use in several ways, as can be seen in Figure 7. 2.
Figure 7. 2 Use of public transport depends on customer satisfaction
Accessibility requirementsAccessibility requirementsAccessibility requirementsAccessibility requirements Before customer satisfaction can affect the use, accessibility requirements need to be fulfilled. In the author’s
opinion, accessibility requirements of public transport are accessibility of the departure and arrival location by
stops. Accessibility needs to be ensured by a sufficient network of services. The accessibility of a stop is
dependent on traveller’s motivation and capacity (Poiesz, 1999). Motivation and capacity vary by person.
Therefore, the accessibility distance may vary by person, e.g., more stops are accessible by cycle compared to
walking. In addition, the accessibility distance may differ by time and weather, e.g., cycling through a dark
wood feels not safe and cycling in snow is not always possible. With respect to the capacity, one should have a
loaded chip card or cash to pay the trip.
Switching if accessibility requirements are not metSwitching if accessibility requirements are not metSwitching if accessibility requirements are not metSwitching if accessibility requirements are not met If the accessibility requirements not met, one could switch to an alternative travel option. E.g., requirements for
driving by car are having a driving license, car ownership or rent, insurance and fuel. Cycling is not possible
with a flat tire; walking may be too far. If the requirements for alternative travel options are not met, one could
Increasing customer satisfaction with public transport 62
consider the accessibility requirements of public transport. However, if requirements regarding both public
transport as well as private modes are not met, one will change the planning. A planning could be changes by
adjusting the location of the activity, the activity itself or cancel the activity (Schoemaker et al., 1999).
Switching options that depends on the appreciation of the trip will be discussed below.
Dissatisfiers and captivesDissatisfiers and captivesDissatisfiers and captivesDissatisfiers and captives Dissatisfiers are the aspects regarding safety, speed and ease. These aspects are placed next to each other,
instead of above each other to make clear that they probably do not have hierarchic relationship and therefore
are probably compensatable. If travellers are not satisfied about safety, speed and ease, they search for an
alternative travel option. Therefore, an operator will loose public transport travellers by dissatisfiers. However,
it is hard to gain new travellers by improving safety, speed and ease according to van Hagen (2011). Travellers
are captives if they do not have alternatives, due to accessibility requirements that are not met for alternative
travel options. As a result, captives continue using the service even if they are dissatisfied. This negative cycle
is presented in the red loop. The loop continues unless captives change their activity patron.
Satisfiers and loyal customersSatisfiers and loyal customersSatisfiers and loyal customersSatisfiers and loyal customers Satisfiers are pleasant factors such as comfort and experience. They are coloured green.
When travellers are satisfied about the satisfiers comfort and experience they will probably use the service the
next time. As long as no negative incidents such as crowded and delayed services occur, satisfied travellers are
expected to continue using public transport. When travellers become familiar in a positive way, they are
expected to become loyal customers. Loyal customers share positive images word-of-mouth. Therefore, the
image of services can be improved. New passengers can be attracted by an improved image. In addition, other
improvements regarding satisfiers, such as personnel friendliness, driving style, cleanness (based on this
thesis), working facilities, pleasant waiting, lighting and infotainment (van Hagen, 2011), can attract
passengers; new passengers could be attracted and current users could use the service more frequent. Also,
travellers could be attracted from alternative options to a public transport service. Alternative options can be
driving by car, cycling or walking, using other public transport services or not performing an activity which
result in no transportation. Travellers with alternative options can be attracted to public transport when they
are dissatisfied about alternative’s safety, speed or ease and satisfied about public transport, especially about
the comfort and experience. In fact, car drivers may partly switch to public transport when the parking costs are
high and the time for finding a parking slot is a lot. In the author’s opinion, travellers switch more easily from a
public mode to a private mode than otherwise. The loyal use to private modes is visualised in the small green
loop. The vision that travellers are more loyal to their private mode compared to a public is a result of the focus
on dissatisfiers in public transport. The focus on dissatisfiers is based on saillant negative aspects in word-of-
mouth, habitation of car use (Verplanken, 1998) and probably the independency that car drivers experience.
Besides experience, the structure of costs is a advantage for private transport. Car costs are mainly fixed,
besides fuel; however, public transport costs are mainly variable.
7.27.27.27.2 How customer satisfaction How customer satisfaction How customer satisfaction How customer satisfaction should be measured should be measured should be measured should be measured according to the authoraccording to the authoraccording to the authoraccording to the author
The KPVV Klantenbarometer is a sufficient method for customer satisfaction. A convenient aspect is that the
questionnaire is fulfilled at the same moment that the service is experienced. Other positive aspects are the
high number of lines and respondents. However, some improvements are possible according to the author.
First, the vision of the author regarding convenient customer satisfaction measurement in public transport will
be proposed. Second, some improvements will be discussed regarding the barometer in more detail.
Increasing customer satisfaction with public transport 63
7.2.17.2.17.2.17.2.1 Proposed customer satisfaction measurement methodProposed customer satisfaction measurement methodProposed customer satisfaction measurement methodProposed customer satisfaction measurement method
Table 7. 1 describes the proposed customer satisfaction measurements in the Netherlands. Several research
categories are described: the questionnaire in the vehicle, a new telephonic research, qualitative research and
data analyses.
Table 7. 1 Proposed customer satisfaction measurement
Improvement categoryImprovement categoryImprovement categoryImprovement category Keep/add/changeKeep/add/changeKeep/add/changeKeep/add/change Regarding Regarding Regarding Regarding
quequequequesssstionstionstionstions
Questionnaire in vehicle Keep the interview location: in the vehicle
Questionnaire in vehicle Keep the time period October-December
Questionnaire in vehicle Let the questionnaire fulfil on an tablet PC; register location, line and time
Questionnaire in vehicle Focus on appreciation of lines in stead of current trip as described below All questions:
change ‘trip’ to
‘line’
Questionnaire in vehicle Keep the appreciation questions about the current sub aspects Questions 1-13,
22 and 24
Questionnaire in vehicle Add questions about traveller’s experience as described below
Questionnaire in vehicle Keep the question about the overall customer satisfaction (changed to line) Question 14
Questionnaire in vehicle Add a question about the overall customer satisfaction regarding the operator
Questionnaire in vehicle Add a question regarding customer loyalty as described below
Questionnaire in vehicle Keep the questions about travel and traveller’s characteristics Questions 16-18,
37-38
Questionnaire in vehicle Apply the ‘method of the most important question’ to short distance and
standing travellers as described below
Questionnaire in vehicle Report the seat capacity (once per vehicle by an interviewer)
Questionnaire in vehicle Optional: yearly changing topic in stead of many safety questions as described
below
Telephonic questionaire Add a telephonic research to include non users as described below
Qualitative research (once) Discussion with travellers why they filled in the questionnaire as they did (as
described below)
Data analyses Couple location and crowding by PT chip card data
Data analyses Defining aspects of improvement by a pie chart of checked improvements and
an appreciation-interest diagram
The main improvements will be described below. First, the information about the current trip is hardly to trace
back to a line in the barometer. Measures are applicable to a line, not to a trip that is already performed.
Therefore, the questionnaire needs to focus on the satisfaction with a line or a bundle of comparable lines,
instead of the current trip. Second, the location and time can registered when the questionnaire is fulfilled on
an electronic device such as a tablet pc. This information can be coupled to crowding by chip card data. As a
result, the impact of crowding can be investigated in more detail. In fact, the question could be answered ‘what
occupancy rate influences customer satisfaction negatively’. In addition, the input of data can be checked real
time to control the interviewers. Third, questions concerning experience and loyalty should be added. Fourth,
the method of the most important question needs to be applied to investigate the satisfaction of short distance
and standing passengers. Fifth, a telephonic research among non travellers needs to investigate how non
travellers can be attracted in a concession area. Sixth, a special topic can be investigated yearly as done in
Increasing customer satisfaction with public transport 64
BEST. Seventh, figures can help prioritizing focus points. The third to the seventh improvement are described in
more detail in the sub sections.
7.2.27.2.27.2.27.2.2 Changes regarding the questionsChanges regarding the questionsChanges regarding the questionsChanges regarding the questions
The recommended changes in questions are summed up below. The argumentation can be found in section
5.1.2.
Include questions regarding expeInclude questions regarding expeInclude questions regarding expeInclude questions regarding experiencerienceriencerience Questions to add regarding traveller’s experience are:
■ ‘What is your opinion regarding the image of this service?’ Axes: very unpleasant – very pleasant (BEST,
2012).
■ ‘I spend my time usefully during the travel?’ Axes: fully untrue – fully true.
Include a question regarding loyalty Studies abroad show some potential improvements for the method. Customer loyalty could be measured by:
■ ‘I gladly recommend travelling with PT to others’. Axes: fully untrue – fully true (BEST, 2012).
Include a questionInclude a questionInclude a questionInclude a question regarding customer satisfaction with the operator regarding customer satisfaction with the operator regarding customer satisfaction with the operator regarding customer satisfaction with the operator A question to add regarding satisfaction with the operator is (Fellesson, 2009):
■ ‘What is your appreciation with the operator of this line?’ Axes: very bad – very good.
Other changes in the questionnaireOther changes in the questionnaireOther changes in the questionnaireOther changes in the questionnaire
Many barometer questions regarding safety are in to much detail compared to the other questions. Safety
appreciation is measured well be a combination of safety at the stop and vehicle. Therefore, all other questions
regarding safety may be deleted. In addition, it is recommended to delete the option ‘living’ in question 18,
because it is no clear trip purpose.
7.2.37.2.37.2.37.2.3 Applying ‘the method of the most important question’ Applying ‘the method of the most important question’ Applying ‘the method of the most important question’ Applying ‘the method of the most important question’
The method of the most important question is an adequate method to get one or two responses form
travellers, without the fulfilment of the full questionnaire. This is especially useful regarding travellers who are
standing or making a short trip. The method is apply in living research (in Dutch: woononderzoek) by CBS. The
method works as follows. If a person refuses to fill in the total questionnaire, they will asked to answer the
main question of the research. Thus, if a travel refuses to answer the whole questionnaire, the interviewer will
say ‘that’s okay, can you only tell me if you are satisfied with this line on a scale from 1 to 10? And if you are
satisfied with [a random aspect]?’ The random question makes it possible to add a second aspect that changes
by interviewee. Thus, a more representative group of travellers can be studied by this method.
7.2.47.2.47.2.47.2.4 Include non public transport users in the surveyInclude non public transport users in the surveyInclude non public transport users in the surveyInclude non public transport users in the survey
The opinions of non public transport travellers are missing in the KPVV Klantenbarometer. In certain areas,
customer satisfaction may be high due to the leave of not satisfied travellers (Fellesson, 2009). Moreover, It is
necessary to consider whether non users could be attracted, and if how. Therefore, non public transport users
are important in the measurements too. Table 7. 2 shows the advised measurement methods and amount of
respondents.
Table 7. 2 Advised Characteristics measurement
PT usersPT usersPT usersPT users Non PT usersNon PT usersNon PT usersNon PT users
Method Questionnaires in vehicle
(as current)
Telephonic questionnaires (new)
Amount of respondents by concession by transport service type 800 (instead of 1000) 200 (new)
Increasing customer satisfaction with public transport 65
It is interesting to find out whether and how non users could become users. They cannot be reached in the
vehicle; therefore, they can randomly be called. The first question during the telephonic interview should be
whether the interviewee uses public transport. If not, than the main questions should be regarding:
■ why the interviewee does not use public transport;
■ whether the person thinks he/she might change to public transport someday;
■ what aspect should be changed to attract this interviewee to public transport, if even possible;
■ a mark for the image of public transport and
■ person characteristics such as travel frequency, car availability, age and gender.
7.2.57.2.57.2.57.2.5 Yearly special topicYearly special topicYearly special topicYearly special topic
An advantage of BEST is that they research special topics yearly. This is especially interesting if the results are
expected to be useful and stable over several years. Special topic regarding customer satisfaction in the
Netherlands could be:
■ Qualitative research. Why travellers filled in the questionnaire as they did is interesting according to HTM
and the author. One method to investigate the reasoning behind the fulfilment is a group discussion with
different travellers of the same line in a meeting room. Another method is directly asking the traveller
why he/she filled in as he/she did, what is important for him/her and what they took into account to
answer the questionnaire. This interesting research is already done in Oslo as described in section 2.2.2
Variations in aspects such as expectations, feelings and the weather may be considered in the
appreciations according to Dick Ettema and the author.
■ Main interests of travellers. Explicit questions about the interests of sub aspects are asked in Konijnendijk
(2003) as discussed in section 2.4. The results of this thesis showed that the explicitly reported interests of
ten years ago differ from the estimated interest by linear regression in this thesis.
■ Market changes for food and drinks at regional transfers. This topic is interesting because operators may
benefit from the sales and pleasant spent time. The question is whether it is financial interesting to invest
in shops nearby regional stops.
■ Relation between objective and subjective measurements. The relationship between objective and
subjective measurement is not always aligned, e.g. on reliability in Scandinavian cities (Fellesson, 2009).
Therefore, more research is needed regarding the coherence and incoherence between standardized
measurements and subjective feelings and appreciations.
■ Market segmentation. Dutch Railways used Needscope to define six groups of travellers (van Hagen, 2011).
These travellers have their own wishes and need to be treated differently. Urban and regional PT do not
have clear market segmentation yet.
■ Innovations. Investigating traveller’s attitude, behavioural intention and behaviour regarding innovations in
public transport.
■ Investigation of marketing success.
7.2.67.2.67.2.67.2.6 Defining the aspects for improvements Defining the aspects for improvements Defining the aspects for improvements Defining the aspects for improvements
The most important aspects for improvement can be defined with the help of the Klantenbarometer results.
Especially a graph with the judgements and (estimated by linear regression) interest is very useful according to
the author and HTM. The aspects with the highest interest and less appreciations can be ranked as potential
improvements. In addition, a pie chart can be implemented with the improvements as checked by travellers
(question 15 in appendix B). An example of both figures is given in Appendix K. Also, sentiment regarding the
lines and operators can be measured by analyses of Twitter data (Collins et al., 2011). The vision of a public
transport expert is needed to consider the consistency of the result and other aspects such as costs.
Increasing customer satisfaction with public transport 66
7.37.37.37.3 HowHowHowHow the results should be used to improve customer satisfaction according to the results should be used to improve customer satisfaction according to the results should be used to improve customer satisfaction according to the results should be used to improve customer satisfaction according to the authorthe authorthe authorthe author
Discussions with experts are held to investigate how the results can be used to improve customer satisfaction.
Several analyses used in this thesis are also applied to operator HTM and a public transport vision (in Dutch: OV
Visie) of PT authorities. Based on that specific analysis and the discussion with experts, recommendation can be
done regarding the demanded information. The level of analysis follows by the level of demanded information.
7.3.17.3.17.3.17.3.1 Policy implications for HTMPolicy implications for HTMPolicy implications for HTMPolicy implications for HTM
HTM could use the findings to support their marketing plan. Therefore, information that is more detailed is
needed. Therefore, the contribution of supply, travel and traveller’s characteristics is analysed to all 15 sub
aspects. The results and policy implications for HTM can be found in Appendix I. The results regarding the
contribution of the sub aspects to customer satisfaction differ from the nation wide analyses. The result
regarding the supply, travel and traveller’s characteristics are in line with the nation wide results. However, the
detailed information gave more insight. In this expert vision chapter, some conclusions and policy implications
described below.
■ Personnel friendliness contributes highly to customer satisfaction with HTM. Personnel friendliness
contributes highest to customer satisfaction with HTM RandstadRail; the appreciation of it is relatively low
(6.5).
■ Improvements should focus on travellers below the age of forty years, because they are less satisfied
compared to older travellers.
■ Give attention (in Dutch: knuffelaandacht) to frequent passengers. E.g., month/year card holders (in Dutch:
abonnementhouders) and other frequent travels should get a small HTM present for their birthday.
■ Prevent earlier departures. In addition, provide real time travel information about earlier departures for
travellers who planned their trip or signed up by an app. The REISinformatiegroep can build such an app.
■ Prevent crowded and delayed services and improve delay information, especially for frequent travellers.
Focus on personnel friendliness to the customer, also at crowded services.
■ Be polite to students, despite they travel without paying mostly.
■ Urban trams should get the same boarding easiness as RandstadRail vehicles by a flat floor.
■ Provide enough personnel space to travellers.
7.3.27.3.27.3.27.3.2 Use for public transport vision DrechtstUse for public transport vision DrechtstUse for public transport vision DrechtstUse for public transport vision Drechtstedenedenedeneden
As stated above, public transport authorities can use the results in their public transport vision. Public transport
authorities can get descriptive information about the age, trip purpose, trip frequency and car availability of
their travellers. In addition, the KPVV Klantenbarometer data can give insight in patronage during different day
parts and transport service types. Analyses can show the contribution of crowding, trip frequency, age and the
other characteristics to customer satisfaction in the areas of the PT vision. Trend lines regarding customer
satisfaction and appreciation of sub aspects give insight in the development of the perceived quality. The
descriptives, trend lines and results of the empiric customer satisfaction model are shown in Appendix J for the
bus concession Drechtsteden. The outcomes of the trend lines, appreciations, (estimated) interest and stated
main points for improvements by travellers can be used to prioritize quality aspects in a tender. The
contribution of the characteristics give insight in which group of travellers is less satisfied; therefore, the
improvements could focus on these groups. An expert should consider this information and costs too. The focus
points can be included in a bonus/malus system.
7.3.37.3.37.3.37.3.3 Discussion Discussion Discussion Discussion customer satisfaction bonus within concessionscustomer satisfaction bonus within concessionscustomer satisfaction bonus within concessionscustomer satisfaction bonus within concessions
As described in the introduction of this report, operators could achieve bonuses based on customer satisfaction.
Only the single question about customer satisfaction on average is used to determine the bonus. The results
gave arguments whether this single question is a good measure. Arguments in favour of bonuses based on the
single question results are:
■ The appreciation factors are correlated well with customer satisfaction.
Increasing customer satisfaction with public transport 67
■ Customer satisfaction and the four factors have highly comparable structure.
■ One measure is less workload for public transport authorities compared to considering the appreciation
factors.
In the author’s opinion, sub aspects should only temporary be included in the bonus structure as an authority
intends to focus on specific sub aspect.
7.3.47.3.47.3.47.3.4 Supporting operator’s and authority’s policies by barometerSupporting operator’s and authority’s policies by barometerSupporting operator’s and authority’s policies by barometerSupporting operator’s and authority’s policies by barometer
In general, the results show that it is very worthwhile to investigate customer satisfaction on a more detailed
level of comparable services. Analyses on detailed level of sub aspects give relevant information for a specific
concession area. Thus, an advice is to repeat such analyses for other concession areas. This is worthwhile for
both operators and authorities. First, operators can include the results about important sub aspects in the
decision of focus points. They can also include the impact of crowding, travel frequency and age in their policy.
Improvements should focus on the aspects of traveller’s groups that influence customer satisfaction negatively.
Second, public transport authorities can use the information about traveller groups and important aspects in
their public transport vision and tender. The analyses for a specific area are important to support policies. The
descriptives of travel and traveller’s characteristics and their relationships with customer satisfaction are
interesting for policy makers concerning a certain area. The area specific analyses will be trusted more and
therefore have more value compared to general results for the whole Netherlands. Thus, marketing plans,
public transport vision and bonus structure should be supported by area specific analyses.
7.3.57.3.57.3.57.3.5 Applying Dutch Railways strategies to regional public transport Applying Dutch Railways strategies to regional public transport Applying Dutch Railways strategies to regional public transport Applying Dutch Railways strategies to regional public transport
Satisfiers such as comfort and experience are becoming more important in public transport. Among others, HTM
and Dutch railways are trying to improve the personal pleasant experience in public transport. First, HTM runs
the campaign ‘Make my day’. Travellers get surprised by a special moment or gift in the vehicles. The surprise
needs to be an experience for all travellers in the vehicle. E.g., distribute many ice creams to one person, such
that he/she has to divide it among the other travellers. In addition, famous persons could surprise fans in the
vehicles. Social media will spread the event. Second, Dutch Railway has many strategies to improve the
experience, see Table 7. 3. In the author’s opinion, urban and regional public transport could also learn from this
development. The results of this thesis have shown that satisfiers such as personnel friendliness, driving style
and cleanness are important according to the travellers. Urban and regional public transport could apply the
Table 7.3 Promising Dutch Railways strategies for regional public transport
Dutch Railways strategyDutch Railways strategyDutch Railways strategyDutch Railways strategy Change of success for urban and regional PTChange of success for urban and regional PTChange of success for urban and regional PTChange of success for urban and regional PT
Improved vehicle design for introvert and extrovert
persons and group travellers
Promising: silence areas in tram, metros and regional trains
Perhaps promising: long lounge banks in new tram, metros and regional
trains
Increase usefulness of in-vehicle time Difficult; driving style and short travel distance are not in advantage of
working in PT. Silences and call areas might be promising, but need to
be research first by travellers needs.
Compaction of time: avoiding much access and
egress time by attracting activity locations to the
stop (van Hagen, 2011)
Promising
Positive branding: Television commercials about
experience with Nick and Simon
Promising with famous feel good person(s)
Personnel training with focus on customer:
customer friendliness, driving style and cleanness
Very promising: train customer-friendly behaviour, e.g., by mentor
drivers; audit personnel friendliness and other comfort aspects, e.g., by
mystery guest.
Applying market segmentation such as NeedScope
(van Hagen, 2011)
Only promising if different market segmentation approaches show
comparable results.
Increasing customer satisfaction with public transport 68
focus on satisfiers. The right column of the he table shows the potency of the strategies according to the
author.
Market segmentationMarket segmentationMarket segmentationMarket segmentation Market segmentation is described in more detail below. If a robust market segmentation is found for urban and
regional public transport, than different policies are possible. Operators should try to get the same traveller
segmentation in the same vehicle. The driving and social skills and attitudes of the crew can be used to match
drivers and stuarts segmentations. Leisure travellers may wish an extravert driver, while, commuters demand
an introvert driver. Elderly people may prefer a quiet driver that drives more carefully. Probably, all these types
of drivers are working at an operator. Therefore, drivers could be matched to the target group of a service.
Thus, the staff planning should be based on travellers needs as much as possible. The current KPVV
Klantenbarometer data can be used to make a segmentation on trip purpose, day parts, transport service types,
trip frequency, car availability and age. However, no psychographic aspects such as underlying wishes of
travellers are taken into account yet.
7.47.47.47.4 Conclusion expert vision: PConclusion expert vision: PConclusion expert vision: PConclusion expert vision: Policy implications olicy implications olicy implications olicy implications customer satisfaction customer satisfaction customer satisfaction customer satisfaction
The higher goals of customer satisfaction are customer loyalty, profitability and continuity. Therefore, more
travellers and more satisfied travellers are needed. The following policy implications are shown to reach and
increase customer satisfaction with convenient urban, agglomerational and regional PT, based on the expert
vision:
■ Ensure sufficient accessibility by a sufficient network;
■ Ensure that the services are safe and understandable to use;
■ Ensure acceptable travel times;
■ Ensure reliable departure and arrival times (or high frequent services);
■ Prevent crowding. Therefore, increase the service reliability by, e.g., preventing early departures, adjusting
time table information to the current supply, keep headways between high frequent services equal by real
time traffic management, providing multiple lines on shared track if a part of a service is much used and
hold vehicles at waiting stops or split lines if travel patronages are splitted at a stop. Crowding can also be
prevented by attracting peak travellers to the off-peak with price incentives and higher frequencies on the
busiest routes.
■ Promise only expectations that can be delivered;
■ Provide customer-friendly behaviour of staff;
■ Investigate the most important traveller’s needs and match the traveller’s needs by qualities of the staff in
the planning;
■ The operator and their personnel should build a committed relationship with the traveller;
■ The operator and authority should control the service and customer satisfaction by analyses by line and
concession area;
■ Make public transport a positive experience, e.g., by campaigns and vehicle design.
Increasing customer satisfaction with public transport 69
Beek, P. van, van Ingen, A., & Konijnendijk, T. (2008). De reiziger spreekt. Het ov gaat gericht aan de slag met
de mening van de reiziger. Verkeerskunde (4), 32-33.
Beirão, G., & Sarsfield Cabral, J.A. (2007). Understanding attitudes towards public transport and private car: A
qualitative study, Transport Policy, 14(6), 478-489.
BEST (2012). Results of the 2011 survey. Helsinki: Benchmarking in European Service of public Transport.
Brons, M., & Rietveld, P. (2007). Betrouwbaarheid en klanttevredenheid in de OV-keten: een statistische analyse. Amsterdam: Vrije Universiteit.
Bunschoten, T. (2012). To tram or not to tram. Exploring the existence of the tram bonus. Delft: TU Delft.
CBS (2012). StatLine. Statline.cbs.nl.
Collins, G., Hasan, S., & Ukkusuri, S. (2011). A Novel Transit Riders’ Satisfaction Metric: Riders’ Sentiments
Measured from Online Social Media Data. Transportation Research Record. Washington: Transportation
Research Board.
Engelsman, J.C., Groenendijk, J.M., & Timmermans, V. (2010). Efficiencygegevens regionaal stad- en
streekvervoer. Den Haag: Twynstra Gudde.
Ettema, D., Gärling, T., Olsson, L. E., & Friman, M. (2010). Out-of-home activities, daily travel, and subjective
well-being. Transportation Research Part A: Policy and Practice, 44(9), 723–732.
Fellesson, M, Haataja, S, Friman, M, Leite, M.E., Lovén, L, Vihervuori, M & Roer, A (2009). The relationship
between objective and customer satisfaction. Karlstad: Karlstad University.
Friman, M., Edvardsson, B., & Gärling, T. (2001). Frequency of negative critical incidents and satisfaction with
public transport services. I. Journal of Retailing and Consumer Services, 8, 95-104.
LiteratureLiteratureLiteratureLiterature
Increasing customer satisfaction with public transport 70
Gemke, J., & de Bruyn, M. (2010). Competentieontwikkeling in dienst van bedrijfsresultaat en
klanttevredenheid. Hoe een onderzoeksafdeling bij een openbaar vervoerbedrijf proactief inspeelt op
toenemende verzakelijking in de sector. Railforum Colloquium ‘Oog voor de klant.’
Graaf, P.M. de, & Kalmijn, M. (1999). Verschillen in echtscheidingscijfers tussen Nederlandse gemeenten.
Maandstatistiek van de Bevolking, november, 15-24.
Grigoroudis, E., & Siskos, Y. (2004). A survey of customer satisfaction barometers: Some results from the
transportation-communications sector. European Journal of Operational Research, 152(2), 334-353.
Heskett, J.L., Jones, T.O., Loveman, G.W., Sasser, W.E & Schlesinger, L.A. (1994). “Putting the Service-Profit Chain
to Work,” Harvard Business Review, 72, 164-174.
Heskett, J.L., Jones, T.O., Loveman, G.W., Sasser, W.E & Schlesinger, L.A. (2008). “Putting the Service-Profit Chain
to Work,” Harvard Business Review, July-August.
Jakobsson Bergstad, C., Gamble, A., Gärling, T., Hagman, O., Polk, M., Ettema, D., Friman, M., & Olsson, L. E.
(2010). Subjective well-being related to satisfaction with daily travel. Transportation, 38(1), 1-15.
Jakobsson Bergstad, C., Gamble, A., Gärling, T., Hagman, O., Polk, M., Ettema, D., et al (2009). Impacts of routine
out-of-home activities on subjec tive well-being. Göteborg Psychological Reports, 4(39), 1–16.
Janse, J., & van Ingen, A. (2009). Klanten tellen mee. Breda: DTV.
Johansson, J.K. (1979). Advertising and the S-Curve: A New Approach. Journal of Marketing Research, 16(3), 346-
354.
Konijnendijk, T., & van Beek. P. (2008). Klantoordelen opnieuw onderzocht en gerubriceerd. Colloquium Vervoersplanologisch Speurwerk.
Konijnendijk, T. (2003). Klantenbarometer: Niet de exacte temperatuur, wel de ontwikkeling van het openbaar
vervoer. CVOV Bericht. CVOV.
Koots, L., Realo, A., & Allik, J. (2011) The Influence of the Weather on Affective Experience: An Experience
Sampling Study. Journal of individual differences, 32(2), 74-84.
KPVV (2012). OV-klantenbarometer 2011. Onderzoek klanttevredenheid in het regionale openbaar vervoer. Resultaten. Breda: Kennisplatform Verkeer en Vervoer, Goudappel Coffeng B.V. / DTV Consultants B.V.
KPVV (2012a). OV-Concessieposter 2012. Utrecht: KPVV.
Lai, W.-T., & Chen, C.-F. (2011). Behavioral intentions of public transit passengers — The roles of service quality,
perceived value, satisfaction and involvement. Transport Policy, 18(2), 318-325.
Lu, J.W., & Beamish, P.W. (2004). International Diversification and Firm Performance: The S-Curve Hypothesis,
The Academy of Management Journal, 47(4), 598-609.
Maslow, A. H. (1943). A theory of human motivation. Psychological Review. 50(4), 370-396.
Oort, N. van (2011). Service Reliability and Urban Public Transport Design. TRAIL, Delft.
Poiesz, T.B.C (1999). Gedragsmanagement. Wormer: Inmerc.
Increasing customer satisfaction with public transport 71
Savelberg, F., & Harms, L. (2009). Kansen voor klanten, Meerwaarde van klanttevredenheidsonderzoek in het openbaar vervoer. Presented at Railforum colloquium ‘oog voor de reiziger’. Den Haag: Kennisinstituut
voor Mobiliteitsbeleid.
Schoemaker, T., Koolstra, K., & Bovy, P. (1999). Traffic in the 21st century - a scenario analysis for the traffic
market in 2030. In E. t. Weijnen M.P.C., The infrastructure playing field in 2030 (pp. 175-194), Delft: Delft
University Press.
Sigger, B. (2011). If Disney ran regional public transport, Applying the Disney philosophy to the service and organisation of Dutch regional public transport. Delft: TU Delft.
Spitsmijden (2009). Effecten van belonen in Spitsmijden in het OV; hoe verleid je OV-reizigers? Antwerpen: Alke
de Coker.
Stradling, S., Anable, J. & Carreno M., (2007). Performance, importance and user disgruntlement: A six-step
method for measuring satisfaction with travel modes, Transportation Research Part A, 41(1), 98-106.
Verkade, E. Vermeulen, W., & Brouwer, J. (2006). Regionale ontwikkelingen, bevolking en werkgelegenheid.
Den Haag: Centraal Planbureau.
Verplanken, B., Aarts, H., van Knippenberg, A., & Moonen, A. (1998). Habit versus planned behaviour: A field
experiment. British Journal of Social Psychology, 37(1), 111-128.
Wardman, M., & Whelan, G.A. (2011). 20 Years of Rail Crowding Valuation Studies: Evidence and Lessons from
British Experience. Transport Reviews 31(3), 379-398.
Woldeamanuel, M.G., & Cyganski, R. (2011). Factors affecting travellers's satisfaction with accessibility to public
transportation. Proceedings European Transport Conference.
Yavuz, N. and E. W. Welch. (2009). Addressing Fear of Crime in Public Space: Gender Differences in Reaction to
Safety Measures in Train Transit. Urban Studies, 47(12), 2491-2515.
Zeithaml, V.A., Berry, L.L., & Parasuraman, A. The Nature and Determinants of Customer Expectations of Service.
Journal of the Academy of Marketing Science, 2(1), 1-12.
ExpertsExpertsExpertsExperts
■ Dick Ettema (traveller’s satisfaction researcher at Universiteit Utrecht)
■ Gerard van Kesteren (KPVV, Klantenbarometer)
■ Hendrik Bouwknegt (public transport specialist at Goudappel Coffeng)
■ Joep ten Brink (senior productmanager at HTM)
■ Matthijs Dicke-Ogenia (expert on mobility psychology at Goudappel and PhD researcher at Rijksuniversiteit
Groiningen)
■ Mark van Hagen (traveller’s satisfaction specialist at Dutch Railways)
■ Niels van Oort (public transport specialist at Goudappel Coffeng)
■ Ronald Coelman (service manager at HTM)
■ Wilco van Dijk (customer psychology researcher at Universiteit Leiden)
Increasing customer satisfaction with public transport 72
Increasing customer satisfaction with public transport Appendixes A-1
A. TerminologyA. TerminologyA. TerminologyA. Terminology
TermTermTermTerm DescripDescripDescripDescriptiontiontiontion
Agglomerational government In Dutch: stadsregio.
Appreciation Appreciation of a factor (calculated average)
B Unstandardized regression coefficient that estimates the influence on the mark (1-
10), dependent of variance of the independent variable
BEST BEST stands for Benchmarking in European Service of public Transport. It is a
customer satisfaction survey of public transport in some big cities of Europe.
beta Standardized regression coefficient, estimates the impact on customer satisfaction,
independent of variance of the independent variable
Bivariate data analysis Analysing relationships between two variables
Car availability The opportunity to drive a car by yourself. Therefore, a traveller needs a car – not
used by someone else at that moment – and a driver license. The main difference
between car availability and ownership and is that an owned car can be
unavailable to use.
Collinearity Very high correlation between two input variables (correlations above .80)
Concession The unique right to supply public transport in a certain concession area with one or
more predefined modes.
Concession area A geographical area containing one or more concessions. It could be that a
concession area has more concessions, e.g., one for each modality.
Customer satisfaction (definition) Customer satisfaction with public transport is defined as the degree to which an
individual positively evaluates the overall quality of a public transport service.
Customer satisfaction (in questionnaire) The answers on the question about overall trip appreciation
Exclude cases listwise Exclude any cases with at least one missing value
Exclude cases pairwise Calculate each correlation as long as values for the two variables are available.
Therefore, the number of respondents used in the analyses is higher than listwise
exclusion and vary by missing values of correlation pairs.
Factor / appreciation factor A group of public transport travellers needs. Four factors of travellers needs are
distinguished: safety, speed, ease and comfort, based on van Hagen (2011) and van
Beek et al. (2008). Factors are grouped by sub aspects of customer satisfaction
AppendixesAppendixesAppendixesAppendixes
Customer satisfaction analyses for HTM A-2
using a factor analysis. See section 2.3 for more explanation about the distinction
of factors.
Factor appreciation Appreciation of a group of sub aspects.
Importance Importance is defined as ‘how important is it to you that this aspect is delivered
well?’ (Stradling, 2007).
Input variables Indepedent variables: transport service types, travel and traveller’s characteristics.
KPVV Klantenbarometer KPVV Klantenbarometer is national survey about customer satisfaction with public
transport in the Netherlands, also called (customer satisfaction) barometer.
Meaningful Significant and relevant result; beta ≥ 0.08
Multicollinearity Very high correlation between three or more input variables (correlations above
.80)
Multivariate data analysis Analysing relationships between multiple variables controlled for correlations with
other independent variables
Output variables Customer satisfaction and its four factors safety, speed, ease and comfort.
Pearson correlation Measure of the linear dependence between two variables. Range: [-1,+1].
PTA See Public transport authority.
Public transport authority A public transport authority grants, modifies or withdraws concessions for public
transport. A decentralized PTA is a province or agglomerational government. A
decentralized PTA manages concessions about regional train lines, metros, trams,
busses and/or ferries.
R square R square represents the explained variance. It is a measure for the goodness of fit
between independent and dependent variables.
Region A spatial area, e.g., a city or rural area
Research area Defined by data file as concession area x mode, sometimes even more divided by
service area.
Social well-being Social well-being is defined as the degree to which an individual positively
evaluates the overall quality of their live.
Standard deviation The standard deviation σ is defined as the square root of the variance σ2, σ =
√E[(X-μ) 2], whereby μ is the sample mean.
Sub aspect Sub aspects are visualized within the pyramid in Figure 1.2, e.g., customer
friendliness.
Supply characteristics Modality and region
Supply characteristics Transport service type, region and operator.
Transport service type A transport service type can be seen as a modality with a function. Therefore, bus
services are divided into urban bus, urban-regional bus and regional bus. Several
transport service types are distinguished: urban bus, urban-regional bus, regional
bus, tram, metro, regional train and ferry.
Transport service type A transport service type can be seen as a modality with a function. Therefore, bus
services are divided into urban bus, urban-regional bus and regional bus. Several
transport service types will be distinguished: urban bus, urban-regional bus,
regional bus, tram, metro, regional train and ferry.
Travel characteristics The number of passengers, day part and trip purpose
Travel characteristics The number of passengers, day part, trip purpose.
Traveller characteristics Car availability, travel frequency, age and gender.
Traveller’s characteristics Car availability, travel frequency, age and gender
t-value t-values above |1.96| are significant at a 95 % confidence interval
Univariate data analysis Reporting descriptives of variables
Variable General description for an aspect, input characteristic, factor or sub aspect that is
related to customer satisfaction.
VIF-value Test for multicollinearity: if VIF-value < 10 no problems with (multi)collinearity
occur.
Increasing customer satisfaction with public transport Appendixes A-3
B. B. B. B. Klantenbarometer methodKlantenbarometer methodKlantenbarometer methodKlantenbarometer method
The text of this appendix is copied from ‘Rapportage OV-Klantenbarometer 2011’ (KPVV, 2012). The method is
the same during 2004-2011. Before 2004, the Klantenbarometer is executed by surveys outside the vehicle.
B.1B.1B.1B.1 InleidingInleidingInleidingInleiding
Het OV-Klantenbarometeronderzoek 2011 is wederom uitgevoerd door middel van een grootschalig onderzoek
onder openbaar-vervoerreizigers. De reiziger heeft de vragenlijst zelf ingevuld gedurende zijn rit. Het
onderzoek is per onderzoeksgebied uitgevoerd, waarvan er in totaal 87 zijn. Per onderzoeksgebied is de
enquête uitgevoerd op een aselecte steekproef van 100 ritten. Op de apart gemeten lijndiensten, zoals de
regionale treindiensten, Qliner 315, de lijnen Breda / Oosterhout - Utrecht en de sneltram Utrecht – Nieuwegein
/ IJsselstein bestond de steekproef uit 50 ritten, de steekproef bij elk van de drie veerdiensten bedroeg 33
vaarten. Hierbij is rekening gehouden met verschillende soorten openbaar vervoer en de tijdsmomenten
waarop de ritten zijn gereden. De enquête is uitgevoerd in de periode tussen 24 oktober en 10 december 2011.
In alle onderzoeksgebieden hebben in totaal 83.513 reizigers van 7.299 ritten de enquête ingevuld.
B.1.1B.1.1B.1.1B.1.1 SteekproeftrekkingSteekproeftrekkingSteekproeftrekkingSteekproeftrekking
Aselectheid is een voorwaarde voor de mogelijkheid voor het doen van zuivere schattingen over de populatie
en voor het bepalen van de betrouwbaarheid van die schattingen. Voor een uitgebreide verantwoording van de
steekproeftrekking wordt verwezen naar paragraaf 2.
B.1.2B.1.2B.1.2B.1.2 Ophoging en wegingOphoging en wegingOphoging en wegingOphoging en weging
Bij het onderzoek is een deel van het totaal aantal reizigers gedurende de onderzoeksperiode geënquêteerd.
Op basis van de ingevulde enquêtes worden uitspraken gedaan over alle reizigers. Om dit te kunnen doen is
een weging en ophoging van de resultaten noodzakelijk, zowel voor het algemene klantenoordeel als de
klantenoordelen per item. Hiertoe is per geënquêteerde rit het aantal instappers geteld. De methode van
ophoging en weging bestaat uit een aantal stappen die zijn beschreven in paragraaf 3.
B.1.3B.1.3B.1.3B.1.3 Betrouwbaarheid van de uitkomstenBetrouwbaarheid van de uitkomstenBetrouwbaarheid van de uitkomstenBetrouwbaarheid van de uitkomsten
Omdat op basis van de ingevulde enquêtes uitspraken worden gedaan over alle reizigers1, bestaat er spreiding
in de onderzoeksuitkomsten rondom de werkelijke, maar onbekende gemiddelden. Niet iedere reiziger is
immers naar zijn of haar oordeel over het openbaar vervoer gevraagd. Bij de interpretatie dient hierdoor
rekening gehouden te worden met een zekere onbetrouwbaarheid van de onderzoeksuitkomsten. Om deze
reden is een berekening gemaakt van het 95%-betrouwbaarheidsinterval. Dit is een gebruikelijke maat voor de
marge. Voor een nadere uitwerking wordt verwezen naar paragraaf 4.
B.2B.2B.2B.2 Verantwoording steekproeftrekkingVerantwoording steekproeftrekkingVerantwoording steekproeftrekkingVerantwoording steekproeftrekking
Voor de steekproeftrekking van de OV-klantenbarometer is een aantal keuzes gemaakt. Deze zijn hieronder
opgenomen.
B.2.1B.2.1B.2.1B.2.1 KeuzesKeuzesKeuzesKeuzes ■ Jaarlijkse steekproef. Jaarlijks wordt een nieuwe steekproef getrokken uit een steeds opnieuw gemaakt
rittenbestand voor geheel Nederland. Dit rittenbestand wordt opgebouwd uit gegevens die de
vervoerbedrijven verstrekken. Het rittenbestand betreft de ‘normale’ dienstregeling welke tijdens de
uitvoering van het veldwerk geldig is:
1 Strikt genomen zijn het niet alle reizigers maar alle reizen: een reiziger kan meerdere reizen maken en dus meerdere
enquêtes invullen.
Increasing customer satisfaction with public transport Appendixes A-4
■ Steekproef per onderzoeksgebied. Er is een methode gehanteerd die recht doet aan de uitgangspunten
van het bestek en de uitvoering. Daarbij wordt ook zoveel mogelijk rekening gehouden met de benodigde
variatie over dagtypen en tijdsperiodes en er wordt voor gezorgd dat de methodiek herhaalbaar is in
nieuwe metingen waardoor er geen methode effecten optreden. Kern van de methode is dat de trekking
van ritten random plaatsvindt en dat deze in ketens geroosterd worden:
■ Per rit worden alle reizigers ondervraagd. Dit is een belangrijk gegeven voor het OV-
Klantenbarometeronderzoek. Er is niet gekozen voor een aselecte trekking van een steekproef van
reizigers maar voor een volledig enquêtering van alle reizigers per rit. Gemiddeld gaat het om circa 10
geslaagde enquêtes per rit, zodat de gemiddelde steekproefomvang per concessie circa 1.000 geslaagde
enquêtes is:
■ Vervoerbedrijven. Vervoerbedrijven zijn niet op de hoogte van de te onderzoeken ritten om te voorkomen
dat men strategisch zou handelen.
B.2.2B.2.2B.2.2B.2.2 StappenStappenStappenStappen
In onderstaande stappen wordt de steekproeftrekking omschreven en waar nodig met een voorbeeld
verduidelijkt: 1. Door de vervoerders is een rittenbestand aangeleverd met daarin alle ritten (van een bepaalde modaliteit
van begin naar eindpunt) per onderzoeksgebied.
2. Op basis van dit rittenbestand is per onderzoeksgebied bepaald wat de procentuele verdeling is over de
dagsoort (als bijvoorbeeld 10% van de ritten zich in een bepaald dagdeel bevindt wordt ook 10% van de
steekproefritten hieruit getrokken)
Werkdag: spits (7.00 – 10.00, 16.00 – 19.00): dalperiode (10.00-16.00): avonduren (na 19.00):
zaterdag/zondag: voor 19.00 uur: na 19.00 uur. 3. De procentuele verdeling uit stap 2 is uitgangspunt voor de verdeling van de 100, 50 of 33 te enquêteren
ritten (van begin naar eindpunt).
4. Het rittenbestand wordt gesorteerd op grond van de indeling in de dagsoort en binnen de dag (om te
bepalen binnen welk tijdsblok een rit behoort wordt uitgegaan van het tijdstip halverwege de rit). Er
ontstaat nu een excelbestand waarin het aantal rijen per tijdsblok duidelijk is aan te geven. Bijvoorbeeld:
tijdsblok 1 bevat de rijen 1 t/m 900 (weekdagen spits), tijdsblok 2 bevat de rijen 901 t/m 2000
(weekdagen dal), ….., tijdsblok n t/m m (zondag avonduren)
5. In stap 2 is bepaald hoeveel ritten per tijdsblok geënquêteerd dienen te worden. Met behulp van een
excelfunctie zijn “random nummers” gegenereerd. Deze randomnummers corresponderen met ritten in het
rittenbestand. Deze ritten maken deel uit van de steekproef.
Bijvoorbeeld: tijdsblok 1 bevat de rijen 1 t/m 900, er dienen 15 ritten in dit tijdblok geënquêteerd te worden (zoals bepaald in stap 2). In totaal zal er dan 15 keer een randomnummer worden gegenereerd dat ligt tussen 1 en 900. De 15 nummers die worden gegenereerd staan voor 15 rijnummers in het rittenbestand en derhalve voor 15 te enquêteren ritten.
6. Na stap 5 is gecontroleerd of de ritten die deel uitmaken van de steekproef een redelijke afspiegeling zijn
voor alle ritten binnen het onderzoeksgebied. Over het algemeen kan worden aangenomen dat een
aselecte steekproef uit het totale rittenbestand resulteert in een representatieve verdeling van het aantal
te onderzoeken ritten binnen het onderzoeksgebied. Bij het roosteren is dit gecontroleerd. Wanneer bleek
dat het toch voorkwam dat een bepaalde lijn sterk onder- of oververtegenwoordigd was, is voor het totale
onderzoeksgebied een nieuwe aselecte steekproef getrokken.
7. Vervolgens zijn de aselect getrokken ritten zoveel mogelijk geautomatiseerd geroosterd door ketens van
getrokken ritten te maken. I.v.m. de uitvoering van het veldwerk zijn deze ritten eventueel aangevuld met
“tussenritten”. Op deze tussenritten is niet geënquêteerd.
De gehanteerde methode komt tegemoet aan de volgende wensen: ■ Goede vertegenwoordiging van dagtypen en periodes op de dag. De gehanteerde methodiek garandeert
dit geheel. De verdeling naar dagtypen (werkdag, zaterdag en zondag) wordt gegarandeerd doordat in de
Increasing customer satisfaction with public transport Appendixes A-5
tien beginritten de juiste verdeling wordt aangehouden. De verdeling naar periodes volgt vanzelf omdat er
negen vervolgritten daarna worden bevraagd zodat vanzelf een goede verdeling verkregen wordt:
■ Goede vertegenwoordiging van lijnen in een gebied. Dit is van belang omdat lijngebruik vaak
motiefafhankelijk is en daarmee invloed zou kunnen hebben op het eindresultaat. Er wordt op een goede
spreiding van lijnen gelet en er wordt voorkomen dat dezelfde rit opnieuw getrokken wordt:
■ Recht doen aan a-selecte trekking. De vormen van randomiseren komt tegemoet aan dit criterium:
■ Recht doen aan optimale inzetschema’s. Aan deze veldwerkvoorwaarde wordt aan tegemoetgekomen
door vervolgritten optimaal te placeren:
■ Herhaalbaarheid. De methode is perfect herhaalbaar in een gebied. Wellicht zou nog idealer zijn steeds
van volledig dezelfde ritten uit te gaan. Alhoewel hier methodische voordelen aanzitten zijn er ook
nadelen zodat we hier niet voor kiezen:
- Dynamiek in het rittenbestand. Vaak bij een concessiewijziging vindt een overgang plaats van een
nieuwe dienstregeling zodat er nieuwe lijnen kunnen zijn. Ook bij bestaande lijnen kan de lijnvoering en
dienstregeling wijzigen:
- Dynamiek in het reizigersbestand. Het kan voorkomen dat eenzelfde lijn gebruikt wordt door een ander
publiek zodat wellicht wel de lijn hetzelfde blijft maar de gebruiker niet.
B.3B.3B.3B.3 VVVVerantwoording ophoging en wegingerantwoording ophoging en wegingerantwoording ophoging en wegingerantwoording ophoging en weging
Bij de weging/ophoging worden de volgende stappen doorlopen: ■ Van geënquêteerde reizigers per rit naar alle reizigers per rit. De kleinste meeteenheid is de reiziger van
wie een ingevulde enquête is ontvangen. Het oordeel van alle reizigers per rit is verkregen door ophoging
naar het aantal getelde reizigers per rit. Hierbij is verondersteld dat alle geënquêteerde reizigers per rit
beschouwd mogen worden als een enkelvoudige aselecte steekproef uit alle reizigers van deze rit. Anders
gezegd: er is verondersteld dat de non-respons niet selectief is:
■ Van onderzochte ritten per tijdsblok naar alle ritten per tijdsblok. Het oordeel van alle reizigers van alle
ritten per tijdsblok is verkregen door een ophoging naar het totaal aantal ritten per tijdsblok:
■ Van alle ritten per tijdsblok naar alle ritten per onderzoeksgebied. Het oordeel van alle reizigers van alle
tijdsblokken per onderzoeksgebied is verkregen door een weging van het totaal aantal op basis van de
telling geschatte reizigers per tijdsblok. Het geschatte aantal reizigers per tijdsblok is verkregen door de
getelde reizigers in de onderzochte ritten op te hogen naar het totaal aantal ritten in het tijdsblok:
■ Van alle ritten per onderzoeksgebied naar alle ritten in Nederland. Het oordeel van alle reizigers over alle
onderzoeksgebieden in Nederland is verkregen door een weging van het totaal aantal geschatte reizigers
per onderzoeksgebied.
B.4B.4B.4B.4 Verantwoording betrouwbaarheid onderzoeksuitkomstenVerantwoording betrouwbaarheid onderzoeksuitkomstenVerantwoording betrouwbaarheid onderzoeksuitkomstenVerantwoording betrouwbaarheid onderzoeksuitkomsten
Het betrouwbaarheidsinterval geeft de marge weer waarin de werkelijke waarde zich met een bepaalde zekerheid
bevindt. Bij de interpretatie van de uitkomsten van het onderzoek dient rekening te worden gehouden met deze marge:
wanneer een vergelijking wordt gemaakt tussen verschillende gebieden of verschillende jaren dient niet alleen op de
cijfers zelf gelet te worden, maar ook op de betrouwbaarheidsintervallen. Het 95%-betrouwbaarheidsinterval laat zich
het gemakkelijkst uitleggen als zou men "95% zeker" zijn dat het onderzochte populatiegemiddelde zich tussen de
grenzen van het interval bevindt.
B.4.1B.4.1B.4.1B.4.1 Berekeningswijze betrouwbaarheidsintervalBerekeningswijze betrouwbaarheidsintervalBerekeningswijze betrouwbaarheidsintervalBerekeningswijze betrouwbaarheidsinterval
Variantie per tijdsblok Bij deze schatters bestaat de mogelijkheid om de steekproefvarianties en daarmee de
betrouwbaarheidsmarges te bepalen. Uitgegaan wordt van de formule voor de variantie van het
Increasing customer satisfaction with public transport Appendixes A-6
populatietotaal zoals gegeven in het boek van Cochran2. Deze wordt omgewerkt volgens bovenstaande notatie
en voor de schatting van het populatiegemiddelde.
waarbij
Deze formules lijken wat ingewikkeld, maar zijn rechtstreeks toe te passen op de gegevens van de steekproef.
Te zien is dat de formule uit twee termen bestaat, de eerste staat voor de bijdrage aan de variantie door de
spreiding in de gemiddelden over de ritten en de tweede term voor de spreiding van de waarden binnen de
ritten.
Variantie over meerdere tijdsblokken Omdat de primaire eenheden gestratificeerd getrokken zijn dienen de stratumvarianties gewogen opgeteld te
worden, evenals bij de schatters voor de populatiegemiddelden. Hiervoor geldt3:
De betrouwbaarheidsmarge is dan:
Bij een 95%-marge wordt voor z de waarde 1,96 genomen.
De variantie wordt dus bepaald door de variatie in de doelvariabele binnen een rit en de variatie in de
gemiddelden tussen de ritten.
2 Cochran, Sampling Techniques, John Wiley & Sons, New York, 1977, p. 305. Zie ook: J. Muilwijk, T.A.B. Snijders, J.J.A.
Moors, Kanssteekproeven, Stenfert Kroese, Leiden/Antwerpen, 1992, p.129/130
3 Cochran, o.c. p.92
Increasing customer satisfaction with public transport Appendixes A-7
C. C. C. C. Klantenbarometer questionnaireKlantenbarometer questionnaireKlantenbarometer questionnaireKlantenbarometer questionnaire
As example, questionnaire Klantenbarometer 2011 Haaglanden is shown. Questionnaires for other regions and
years are almost the same. One example a small change will be given: for the surveys in trains, ‘bestuurder’ is
replaced by ‘machinist’. In addition, detail questions about safety are asked in the questionnaire, but not part of
KPVV Klantenbarometer.
Increasing customer satisfaction with public transport Appendixes A-8
Increasing customer satisfaction with public transport Appendixes A-9
D. Customer satisfaction measurements and important supply aspects abroadD. Customer satisfaction measurements and important supply aspects abroadD. Customer satisfaction measurements and important supply aspects abroadD. Customer satisfaction measurements and important supply aspects abroad
Benchmarking in European service of public Transport (BEST) measures customer satisfaction and objective
quality in, among others, Stockholm, Helsinki and Oslo. The questionnaires are about appreciation of sub
aspects, overall judgement for the journey or service, travel and traveller’s characteristics. These measurements
are described for two reasons. First, convenient aspects of these measurements will be included in the expert
vision regarding how to measure customer satisfaction in chapter 7. BEST measures the image of services and
customer loyalty. BEST interviews non-public transport users too. In fact, questions of BEST will be used to
improve the KPVV Klantenbarometer. Second, an important aspect of this study regarding references is to sum
up factors what variables influence customer satisfaction most in these cities. Stockholm, Helsinki and Oslo are
chosen as reference studies, because of similarity in method with KPVV Klantenbarometer. In BEST, urban and
regional transport services are included. BEST includes trams, busses, trams, metro and light rail; excluded are
national train services. In Oslo, ferry services are included as well (Fellesson et al., 2009). Therefore, the
comparability between the studies in Stockholm, Helsinki and Oslo and the KPVV Klantenbarometer is high. In
addition, the Dutch train services are compared to that of other European countries at the end of this section.
D.1D.1D.1D.1 Customer satisfaction with StoCustomer satisfaction with StoCustomer satisfaction with StoCustomer satisfaction with Stockholmckholmckholmckholm
D.1.1D.1.1D.1.1D.1.1 Method in StockholmMethod in StockholmMethod in StockholmMethod in Stockholm
In Stockholm, the Perceived Quality survey is conducted each month. The “Perceived Quality” is a
questionnaire, which is filled in by respondents in the vehicle. The response is about 5,000 interviewees
monthly spread across busses, metros, light rail services and express trains. Interviewees rate different aspects
at a seven-point scale. The lowest three points on the scale are translated to dissatisfaction. The box in the
middle is translated to neutral. The last three points on the scale are translated to satisfaction (Fellesson et al.,
2009). Main difference between this method and the KPVV Klantenbarometer is the scale length and the
translation of the scale. The other aspects of the method are highly comparable to KPVV Klantenbarometer.
D.1.2D.1.2D.1.2D.1.2 Important aspects in StockholmImportant aspects in StockholmImportant aspects in StockholmImportant aspects in Stockholm
In Stockholm, the most important sub aspects relating to customer satisfaction are ranked as follows:
punctuality, frequencies and reducing crowding. Customers also demand safe and environmental friendly
vehicle. Furthermore, frontline staff should be visible and friendly. Commercial services at stations are
important for a pleasant experience. Based on customer surveys, the public operator of Stockholm AB SL states
to know that the most significant sub aspect is obtaining new passengers is simple and easily accessible
information about how to travel with the operator (Fellesson et al., 2009).
D.2D.2D.2D.2 Customer satisfaction with HelsinkiCustomer satisfaction with HelsinkiCustomer satisfaction with HelsinkiCustomer satisfaction with Helsinki
D.2.1D.2.1D.2.1D.2.1 Method in HelsinkiMethod in HelsinkiMethod in HelsinkiMethod in Helsinki
In Helsinki, public transport is provided by Helsinki City Transport (HKL), which is owned by the city of Helsinki.
Customer satisfaction is measured via surveys filled in by passengers in busses, trams and metros. Surveys are
done continuously and reported quarterly (Fellesson et al., 2009).
D.2.2D.2.2D.2.2D.2.2 Important aspects in HelsinImportant aspects in HelsinImportant aspects in HelsinImportant aspects in Helsinkikikiki
In Helsinki, the most important sub aspect according to passengers is reliability, followed by route network,
intervals of departure times and travel time. All these sub aspects referred to the second factor of the pyramid,
which was called speed. Reliability has 11% impact on the perceived overall quality. Remarkable, the objective
reliability has been very variable during 2000-2008, were the judgements about reliability stays stable. For
improving customer satisfaction cost-efficient it is important to know that a convenient driving style is more
important than the quality of the vehicles in Helsinki. Other important aspects are stops, stations and terminals,
information and passenger service. Interaction with other passengers has the lowest impact on overall
customer satisfaction with Helsinki (Fellesson et al., 2009).
Increasing customer satisfaction with public transport Appendixes A-
D.3D.3D.3D.3 Customer satisfaction with OsloCustomer satisfaction with OsloCustomer satisfaction with OsloCustomer satisfaction with Oslo
D.3.1D.3.1D.3.1D.3.1 Method in OsloMethod in OsloMethod in OsloMethod in Oslo
In Oslo, public transport is operated by Oslo Sporveier, later turned into the name Ruter. The surveys are explicit
about the present journey only. The respondents are asked to rate their satisfaction with sub aspects on a five
point scale. After the questionnaire, qualitative surveys are done with the main question why did respondents
filled in as they did. The grades 1-2 were chosen when the respondent was dissatisfied with that sub aspect.
The grades 4-5 are used when passengers are satisfied. Depending on the question and respondent, scale 3
was chosen as neutral or positive. The qualitative survey also showed that passengers tend to include previous
travel experiences with the operator Oslo Sporveier, and their opinions about the operator in general. Taking
into account more aspects than the present journey only influences the answers negatively (Fellesson et al.,
2009). Thus, the actual satisfaction for the present journey is higher than responded, because of the bias of
previous travel experience and image of the operator. The same underestimation of customer satisfaction could
account for the KPVV Klantenbarometer, which also emphasize the current trip.
D.3.2D.3.2D.3.2D.3.2 Important aspects in OsloImportant aspects in OsloImportant aspects in OsloImportant aspects in Oslo
In Oslo, most important aspects about the current journey are in order or importance: punctuality, driving style,
vehicle cleanness, vehicle temperature and timetable information at stops. Driving style, cleanness and
temperature could be optimized by the operator. A second questionnaire is done about the satisfaction of public
transport in general in Oslo, not only considering this journey. In the second study, passengers ranked the
following sub aspects in order of importance about their satisfaction with the public transport system in Oslo:
frequency, punctuality and transfer time. Frequency is twice important as a low transfer time. Other important
sub aspects are door-to-door travel time and price. Many other aspects have a low influence on customer
satisfaction when looking separately at that aspect, but all together they are important: comfort in vehicle,
egress distance, seat occupation, direct connections, information, sheltered waiting, and access distance. In
both studies, there is no systematic relationship between impact and mark (Fellesson et al., 2009). Concluded
can be, from this paragraph, the main differences between the studies are as follows. If the questionnaire is
explicitly about the current journey, driving style, cleanness and temperature become important compared to
satisfaction with the public transport system in general. Frequency and transfer time are less important by
evaluating the current trip.
D.4D.4D.4D.4 Conclusions of measurements abroadConclusions of measurements abroadConclusions of measurements abroadConclusions of measurements abroad
From the measurements of BEST can be concluded that the KPVV Klantenbarometer might be improved by
measurements regarding the image of services and customer loyalty and including non-public transport. From
the results of BEST can be concluded that customer satisfaction is mainly influenced by appreciations of
reliability, frequencies, reducing crowding, driving style, vehicle cleanness and timetable information.
Increasing customer satisfaction with public transport Appendixes A-11
E. E. E. E. List of supply aspects influencing customer satisfactionList of supply aspects influencing customer satisfactionList of supply aspects influencing customer satisfactionList of supply aspects influencing customer satisfaction
Table E.1 Aspects influencing customer satisfaction organised by output factor
Output factorOutput factorOutput factorOutput factor AspectsAspectsAspectsAspects Availability in KPVV Klantenbarometer datasetAvailability in KPVV Klantenbarometer datasetAvailability in KPVV Klantenbarometer datasetAvailability in KPVV Klantenbarometer dataset
Safety Social safety in vehicle
Social safety at stop
Lightening at stops
Lightening in vehicle
Available
Available
Not available
Not available
Speed Punctuality
Frequency
Travel speed
Travel time
Door-to-door travel time
Information about delays
Egress time
Transfer time
Access time
Information in vehicle
Information at stop
Tariff
Direct connections
Transfer walking distance
Egress distance
Route network
Access distance
Value for money
Available
Available
Available
Not available
Not available
Available
Not available
Not available
Not available
Not available
Available
Available
Not available
Not available
Not available
Not available
Not available
Not available
Ease Easiness ticket buying / loading chip card
Low number of transfers
Signs to/from stop
Knowing ticket selling points
Seat availability
Easiness of boarding
Available
Not available
Not available
Not available
Available
Available
Comfort
Driving style
Cleanness vehicle
Cleanness station
Noise inside vehicle
Noise outside vehicle
Possibility to eat and drink
Crowding
Friendliness personnel
Friendliness driver
Interaction with other passengers
Vehicle temperature
Parking facilities
Available
Available
Not available
Available
Not available
Not available
Indicator available
Available
Not available
Not available
Not available
Not available
Pleasant experience Commercial services at station
Facilities at stop
Sheltered waiting
Privacy
Image
Colours
Scents
Sounds
Aesthetics
Relaxation
Spending time usefully
Distraction
Not available
Not available
Not available
Not available
Not available
Not available
Not available
Not available
Not available
Not available
Not available
Not available
Increasing customer satisfaction with public transport Appendixes A-12
F. Cross validation results F. Cross validation results F. Cross validation results F. Cross validation results empirical customer satisfaction empirical customer satisfaction empirical customer satisfaction empirical customer satisfaction model 2009 and 2010model 2009 and 2010model 2009 and 2010model 2009 and 2010
Table F.1 Results empirical customer satisfaction model based on input factors (2009)
* Si
gnif
ican
ce P
<0.0
5
Ref
eren
ces:
met
ro, m
ornin
g pe
ak, o
ther
s tr
ip p
urpo
ses
Mea
nin
gfu
l tra
vel an
d tr
avel
ler’
s ch
arac
teri
stic
s ar
e m
arke
d b
old
Increasing customer satisfaction with public transport Appendixes A-13
Table F.2 Results empirical customer satisfaction model based on input factors (2010)
* Si
gnif
ican
ce P
<0.0
5
Ref
eren
ces:
met
ro, m
ornin
g pe
ak, o
ther
s tr
ip p
urpo
ses
Mea
nin
gfu
l tra
vel an
d tr
avel
ler’
s ch
arac
teri
stic
s ar
e m
arke
d b
old
Increasing customer satisfaction with public transport Appendixes A-14
G. Methodology and results interaction testG. Methodology and results interaction testG. Methodology and results interaction testG. Methodology and results interaction test
Interaction terms are analysed to try explaining more variance. An interaction is the influence by a third
variable to the contribution of an independent variable to a dependent variable. No interactions meaningful are
found; as a result, this appendix shortly describes the tested interactions and translates the conclusions to
understandable text.
G.1 G.1 G.1 G.1 Method interaction testsMethod interaction testsMethod interaction testsMethod interaction tests
Models with interaction terms will be compared to the empirical customer satisfaction model. Series interaction
terms will be added to the model if the F-change is significant at a 95% confidence interval and the regression
coefficient meaningful.
G.2 Results interaction analysesG.2 Results interaction analysesG.2 Results interaction analysesG.2 Results interaction analyses
Interaction terms to analyse are based on the meaningful travel and traveller’s characteristics of the empirical
customer satisfaction model: the number of passengers, trip frequency and age, see Table G.1. Interactions are
research on customer satisfaction and the four factors. No meaningful interactions for the groups are found on
high aggregation level. As a result, transport service types, the number of passengers, trip frequency and age
contribute only separately to customer satisfaction and appreciation of safety, speed, ease and comfort. In fact,
the contribution of the number of passengers, trip frequency and age to customer satisfaction is not influenced
by transport service types; their contribution is independent of modality and region. Therefore, the empirical
customer satisfaction model stays as described above.
Table G.1 Analysed interaction, no meaningful interactions found
Group 1: Interactions with Group 1: Interactions with Group 1: Interactions with Group 1: Interactions with the the the the
number of passengersnumber of passengersnumber of passengersnumber of passengers
Group 2: Interactions with trip Group 2: Interactions with trip Group 2: Interactions with trip Group 2: Interactions with trip
frequencyfrequencyfrequencyfrequency
Group 3: Interactions with ageGroup 3: Interactions with ageGroup 3: Interactions with ageGroup 3: Interactions with age
City bus within G4 x Passengers City bus within G4 x Trip frequency City bus within G4 x Age
City bus outside G4 x Passengers City bus outside G4 x Trip
frequency
City bus outside G4 x Age
Agglomerational bus x Passengers Agglomerational bus x Trip
frequency
Agglomerational bus x Age
Regional bus x Passengers Regional bus x Trip frequency Regional bus x Age
Tram within G4 x Passengers Tram within G4 x Trip frequency Tram within G4 x Age
Tram outside G4 x Passengers Tram outside G4 x Trip frequency Tram outside G4 x Age
Regional train x Passengers Regional train x Trip frequency Regional train x Age
Ferry x Passengers Ferry x Trip frequency Ferry x Age
Metro x Passengers (reference) Metro x Trip frequency (reference) Metro x Age (reference)
Increasing customer satisfaction with public transport Appendixes A-15
HHHH. Potential cost efficient . Potential cost efficient . Potential cost efficient . Potential cost efficient focus on lowly focus on lowly focus on lowly focus on lowly satisfied travellerssatisfied travellerssatisfied travellerssatisfied travellers
HHHH.1 Policy opportunities.1 Policy opportunities.1 Policy opportunities.1 Policy opportunities
One policy of an operator could be improving customer satisfaction of traveller groups that give low marks. This
policy has several advantages. First, the most unsatisfied group could become satisfied, spread better word-of-
mouth publicity and may become a loyal customer. Second, the average satisfaction mark could increase which
could result in earning a bonus.
HHHH.2 .2 .2 .2 DDDDefiningefiningefiningefining traveller groups traveller groups traveller groups traveller groups
To perform this policy, focus group with the lowest marks need to be defined. The results above have shown
already that an operator could focus on high frequent travelling man with low age, because they give the
lowest judgments compared tot other travellers. It is important that members of the focus group are
recognisable easily by drivers and stuarts. Personnel can easily distinguish on age and gender, but not on travel
frequency. Therefore and other classification is chosen in this analyse of traveller groups. It is assumed that
personnel can distinguish commuters groups based on trip purposes; students defined as travellers with trip
purpose education and age of 18 years or older, secondary school students defined as travellers with trip
purpose education and age below 18 years, commuters, visitors, sportspersons, travellers going home and
shoppers. Sportspersons could be recognised on their sport clothes and baggage. Shoppers could be recognised
on their shopping bags and department location nearby a shopping area. Travellers going home could be
recognized by the time of the day, mainly evening peak. Travellers going home could have other trip purposes
as well.
HHHH.3 Results for traveller groups.3 Results for traveller groups.3 Results for traveller groups.3 Results for traveller groups
Table H.1 shows the results of the analyses for policies based on focus groups. Variables excluded of the table
are excluded of the analyses. The results are visualised in several figures. To improve the average mark of
customer satisfaction, an operator should focus on improving satisfaction of students and scholars, see Figure
H.1. To improve the appreciation of personnel friendliness drivers and stuarts should be more friendly to
students and scholars, see Figure H.2. In addition, commuters appreciate personnel friendliness low. Personnel
could also friendlier to this traveller group. However, some commuters want to be on there selves (Sigger,
2011). Therefore, personnel should approach commuters carefully. Driving style and cleanness should be
improved one lines where and day part when many scholars travel, see Figure H.3 and H.4. Delay information
should mainly be improved for students, scholars and commuters, as shown in Figure H.5.
Customer satisfaction by traveller groups
6
7
8
Com
mute
rs
Sec
ondary
schola
r stu
dents
(18-)
Stu
de
nts
(18+
)
Sh
oppers
Sport
sp
ers
ons
Vis
itors
Tra
velle
rs g
oin
g h
om
e
Tra
velle
rs w
ith o
ther
trip
pu
rposes
Men
Wo
men
Traveller groups
Cu
sto
mer
sati
sfa
cti
on
(m
ark
1-1
0)
Appreciation of personnel friendliness by traveller groups
6
7
8
Co
mm
ute
rs
Secon
dary
sch
ola
r stu
dents
(18
-
)
Stu
den
ts (
18+
)
Sho
ppers
Sport
sp
ers
ons
Vis
itors
Tra
ve
llers
goin
g h
om
e
Tra
velle
rs w
ith o
the
r tr
ip
purp
ose
s
Me
n
Wo
men
Traveller groups
Ap
pre
cia
tio
n o
f p
ers
on
nel
frie
nd
lin
ess (
mark
1-1
0)
Figure H.1 Contribution of traveller groups to customer satisfaction, other variables excluded (left) Figure H.2 Contribution of traveller groups to personnel friendliness appreciation, other variables excluded (right)
Increasing customer satisfaction with public transport Appendixes A-16
Table H.1 Appreciation of aspects, which can be influenced by operators reported, by focus groups
Sign
ific
ance
P<0
.05
Ref
eren
ces:
oth
ers
trip
pur
pos
e a
nd
wom
an
Mea
nin
gfu
l tra
velle
r gr
oups
are
mar
ked b
old
Increasing customer satisfaction with public transport Appendixes A-17
Appreciation of driving style by traveller groups
6
7
8
Com
mute
rs
Secondary
schola
r stu
dents
(18-)
Stu
dents
(18
+)
Shoppers
Sport
spers
ons
Vis
itors
Tra
velle
rs g
oin
g h
om
e
Tra
velle
rs w
ith o
ther
trip
pu
rposes
Men
Wom
en
Traveller groups
Ap
pre
cia
tio
n o
f d
rivin
g s
tyle
(m
ark
1-1
0)
Appreciation of cleanness in vehicle by traveller groups
6
7
8
Com
mute
rs
Secondary
schola
r stu
dents
(18-)
Stu
dents
(18+
)
Shoppers
Sport
spers
ons
Vis
itors
Tra
velle
rs g
oin
g h
om
e
Tra
velle
rs w
ith o
ther
trip
purp
oses
Men
Wom
en
Traveller groups
Ap
pre
cia
tio
n o
f cle
an
ness i
n v
eh
icle
(m
ark
1-1
0)
Figure H.3 Contribution of traveller groups to driving style appreciation, other variables excluded (left) Figure H.4 Contribution of traveller groups to cleanness appreciation, other variables excluded (right)
Appreciation of delay information by traveller groups
4
5
6
Com
mu
ters
Second
ary
sch
ola
r stu
dents
(18-)
Stu
de
nts
(18
+)
Sho
ppers
Sport
spe
rsons
Vis
itors
Tra
velle
rs g
oin
g h
om
e
Tra
velle
rs w
ith o
ther
trip
purp
oses
Men
Wom
en
Traveller groups
Ap
pre
cia
tio
n o
f d
ela
y i
nfo
rmati
on
(m
ark
1-1
0)
Figure H.5 Contribution of traveller groups to delay information appreciation, other variables excluded
HHHH.4 Focus on improvements for students, commuters and s.4 Focus on improvements for students, commuters and s.4 Focus on improvements for students, commuters and s.4 Focus on improvements for students, commuters and scholarscholarscholarscholars
The results for traveller groups are compared to the results of gender. In general, operators could focus on
improving the satisfaction of men. However, the differences by gender are low compared to traveller group.
Therefore, operators should focus on students, scholars and commuters to improve customer satisfaction and
sub aspects that can be influenced by operators. The amount of travellers within the groups should be taking
into account by prioritizing the focus groups. As a result of that, the priorities should be as follows: students,
commuters and scholars.
Increasing customer satisfaction with public transport Appendixes A-18
IIII. Customer satisfaction analyses specified to HTM. Customer satisfaction analyses specified to HTM. Customer satisfaction analyses specified to HTM. Customer satisfaction analyses specified to HTM
The analyses of the study to customer satisfaction in the Netherlands are applied on the HTM concessions. The
analyses are useful for the marketing plan of HTM. First, this appendix discusses the important sub aspects
contributing to customer satisfaction within the HTM concessions. Second, supply, travel and traveller’s
characteristics are analysed with the empirical customer satisfaction model. Third, the vision of the author
about the usefulness of these analyses is given.
I.I.I.I.1 Important customer satisfaction sub aspects1 Important customer satisfaction sub aspects1 Important customer satisfaction sub aspects1 Important customer satisfaction sub aspects
Table I.1 shows the contribution of fifteen sub aspect to customer satisfaction for the HTM services. The
contributions are presented for HTM urban bus services, HTM urban tram services, HTM RandstadRail and all
HTM travellers together.
Main conclusions regarding important sub aspects are as follows:
■ The conclusions of the analyses for HTM differ slightly for the results found in chapter 4 for the whole
Netherlands.
■ Information at stops and noise in vehicle contribute meaningfully to customer satisfaction for HTM bus
services;
■ Frequency and personnel friendliness contribute meaningfully to customer satisfaction for HTM urban tram
services;
■ Travel speed and personnel friendliness contribute meaningfully sub aspects to customer satisfaction for
HTM RandstadRail tram services;
■ Punctuality is mainly an issue for the urban tram and bus services;
■ Cleanness is important in urban trams and RandstadRail;
■ Driving style is of medium importance for urban tram and bus services;
■ Safety at urban bus stops is important. However, safety in the vehicles shows almost no significant
contribution.
Policy implications regarding important sub aspects are as follows:
■ Personnel friendliness earns the highest focus (see Table I.2). Personnel friendliness could be improved by
friendly contact of the drivers and stuarts to the travellers. Due to the high importance of personnel
friendliness continue awareness is needed. Friendliness to the customer should be an explicit aspect
during drive training. Therefore, the instructors of the driving instructions could focus on personnel
friendliness next to routes and infrastructure.
I.2 Contribution of supplyI.2 Contribution of supplyI.2 Contribution of supplyI.2 Contribution of supply, travel and traveller’s characteristics, travel and traveller’s characteristics, travel and traveller’s characteristics, travel and traveller’s characteristics
The empirical customer satisfaction model estimates the contribution of input characteristics to customer
satisfaction and the four factors safety, speed, ease and comfort. HTM preferred a more detailed view on all
fifteen sub aspects. The idea behind this preference is that the results can be used to improve, e.g., information
about delays specific. Thus, HTM preferred a more detailed level of analyses. Therefore, Table I.3 shows the
contribution of supply, travel and traveller’s characteristics to customer satisfaction and the fifteen sub aspects
for HTM urban bus, urban tram and RandstadRail services. Five sub aspects are highlighted in graphs, see Figure
I.1-I.5. These figures visualise the results as presented in Table I.3. However, the contribution of the amount of
passengers, trip frequency and age are analysed with dummies. Therefore, the contribution of age seems not
to be linear in all cases. More about this discussion could be read in section 4.2.3.
Increasing customer satisfaction with public transport Appendixes A-19
Table I.1 Important sub aspects for HTM services
Table I.2 Advised focus for cost efficient policy directions
Cost efficient policy directions HTM Urban bus HTM Urban tram HTM RandstadRail
Friendliness personnel Medium focus High focus High focus
Driving style Small focus Small focus -
Cleanness - Medium focus Medium focus
Sign
ific
ance
P<0
.05
Mai
n c
ontr
ibut
ions
to c
usto
mer
sat
isfa
ctio
n a
re m
arke
d bol
d
Increasing customer satisfaction with public transport Appendixes A-20
Mean results of these regression analyses are as follows:
■ The conclusions of chapter 4 regarding all concession areas in the Netherlands hold mainly for HTM
services too.
■ The number of passengers contributes negatively to all sub aspects. The amount of passenger contributes
high to punctuality, delay information, personnel friendliness, noise in the vehicles and cleanness.
Therefore, these aspects are more important when the services are crowded.
■ Frequent travellers are less satisfied about every sub aspect compared to incidental travellers. Main
explanation why frequent travellers are less satisfied is their low appreciation of punctuality, price, ticket
facilities, seat availability and all comfort aspects.
■ Elderly travellers are more satisfied about every sub aspect besides boarding compared to young
travellers. Thus, elderly people report higher marks. This is in line with other studies were elderly people
report higher marks compared to young people. Based on the results of chapter 4, improvements should
be focused on people until 40 years.
■ Elderly travellers are least satisfied about boarding.
■ The differences in appreciations of the most sub aspects are related to travel and traveller’s characteristics.
Modalities contribute only highly to delay information, boarding, noise and cleanness.
■ The appreciation of delay is mainly related to the modality. RandstadRail scores the highest appreciation
on this sub aspect. This can be explained by the information panels of RandstadRail. Urban travellers and
younger people appreciate delay information the lowest.
■ The characteristics of a tram contribute highly to the difficulty of boarding compared to RandstadRail. The
differences are more than one point on the boarding appreciation between urban tram and RandstadRail.
■ Noise and cleanness are appreciated less in urban tram services compared to RandstadRail.
■ Price is appreciated less.
■ Customer satisfaction decreases highly when the amount of travellers become around 10 for bus travellers.
A reason might be that travellers like to have their own two or four seats.
Policy implications regarding input characteristics:
■ Give attention (in Dutch: knuffelaandacht) to frequent passengers. E.g., month/year card holders and other
frequent travels could get a small HTM present for their birthday by mail; the letter regarding extinction of
the month/year card could be more polite and on the right moment not being in the summer holiday;
giving away ice creams to month/year card holders on a hot day;
■ Avoid early departures, especially at lines and day parts whereby the frequencies are low. In case early
departures occur, information can be provided by an app. The travel information app can be build by 9292.
■ Delay information need to be clear, especially if travellers have alternative travel options. In addition, only
a few alternatives for the route may be offered to avoid stress.
■ Prevent crowded delayed services, especially for frequent travellers.
■ Improve delay information, especially for crowded services. Therefore, HTM will place panels at al tram
and bus stops.
■ Focus on personnel friendliness to the customer, also at crowded services.
■ Be polite to students, despite they do not have to pay mostly.
■ Urban trams should get the same boarding easiness as RandstadRail vehicles by a flat floor.
■ Price experienced may be improved by special and cheaper tickets such as beach express. The offers
should focus on travellers until 40 years.
■ Provide enough personnel space to travellers.
Increasing customer satisfaction with public transport Appendixes A-21
Table I.3 Results empirical customer satisfaction model based on input characteristics for HTM (2011)
* Si
gnific
ance
P<0
.05
Ref
eren
ces:
HTM
Ran
dsta
drai
l, m
ornin
g pe
ak, o
ther
s tr
ip p
urpo
ses
Mea
nin
gfu
l tra
vel an
d tr
avel
ler’
s ch
arac
teri
stic
s ar
e m
arke
d b
old
Increasing customer satisfaction with public transport Appendixes A-22
Table I.3 Results empirical customer satisfaction model based on input characteristics for HTM (2011) (continued)
Increasing customer satisfaction with public transport Appendixes A-23
Table I.3 Results empirical customer satisfaction model based on input characteristics for HTM (2011) (continued)
Increasing customer satisfaction with public transport Appendixes A-24
Results customer satisfaction
6
7
8
HT
M U
rba
n b
us s
erv
ice
s
HT
M U
rban t
ram
se
rvic
es
HT
M R
andsta
dR
ail
0-2
5 p
asse
ngers
26-5
0 p
asse
ngers
50-1
00 p
asse
ngers
100-5
05 p
asse
ngers
Mo
rnin
g p
eak
Eve
nin
g p
eak
Off
-peak
Com
muti
ng
Edu
catio
n
Vis
itin
g
Oth
er
trip
pu
rpo
se
Ca
r av
aila
ble
No c
ar
av
aila
ble
0-1
trip
s p
er
we
ek
2-4
trips
a w
eek
5 o
r m
ore
trips
a w
eek
12-1
8 y
ea
r
18-2
7 y
ea
r
28-4
0 y
ea
r
41-6
4 y
ea
r
65-8
9 y
ea
r
Man
Wo
man
Input characteristics
Cu
sto
me
r sati
sfa
cti
on
(m
ark
1-1
0)
Figure I.1 Contribution of input characteristics to customer satisfaction (HTM)
Results appreciation of safety in vehicle(HTM)
6
7
8
HT
M U
rba
n b
us s
erv
ice
s
HT
M U
rban t
ram
se
rvic
es
HT
M R
andsta
dR
ail
0-2
5 p
asse
ngers
26-5
0 p
asse
ngers
50-1
00 p
asse
ngers
100-5
05 p
asse
ngers
Mo
rnin
g p
eak
Eve
nin
g p
eak
Off
-peak
Com
muti
ng
Edu
catio
n
Vis
itin
g
Oth
er
trip
pu
rpo
se
Ca
r av
aila
ble
No c
ar
av
aila
ble
0-1
trip
s p
er
we
ek
2-4
trips
a w
eek
5 o
r m
ore
trips
a w
eek
12-1
8 y
ea
r
18-2
7 y
ea
r
28-4
0 y
ea
r
41-6
4 y
ea
r
65-8
9 y
ea
r
Man
Wo
man
Input characteristics
Safe
ty i
n v
eh
icle
ap
pre
cia
tio
n (
ma
rk 1
-10)
Figure I.2 Contribution of input characteristics to safety in vehicles (HTM)
Results appreciation of travel speed (HTM)
6
7
8
HT
M U
rban b
us s
erv
ices
HT
M U
rba
n t
ram
serv
ices
HT
M R
an
dsta
dR
ail
0-2
5 p
assen
gers
26-5
0 p
assen
gers
50-1
00 p
assen
gers
10
0-5
05 p
assen
gers
Morn
ing p
eak
Evenin
g p
eak
Off
-pe
ak
Co
mm
uting
Educ
ation
Vis
itin
g
Oth
er
trip
purp
ose
Car
ava
ilable
No c
ar
ava
ilable
0-1
trips
per
week
2-4
tri
ps a
week
5 o
r m
ore
tri
ps a
week
12-1
8 y
ear
18-2
7 y
ear
28-4
0 y
ear
41-6
4 y
ear
65-8
9 y
ear
Man
Wom
an
Input characteristics
Tra
vel
sp
ee
d a
pp
recia
tio
n (
mark
1-1
0)
Figure I.3 Contribution of input characteristics to travel speed (HTM)
Increasing customer satisfaction with public transport Appendixes A-25
Results appreciation of punctuality (HTM)
6
7
8
HT
M U
rban b
us s
erv
ices
HT
M U
rba
n t
ram
serv
ices
HT
M R
and
sta
dR
ail
0-2
5 p
as
seng
ers
26-5
0 p
as
seng
ers
50-1
00
pas
seng
ers
100
-505
pas
seng
ers
Morn
ing p
eak
Evenin
g p
eak
Off
-pe
ak
Com
mu
ting
Educa
tion
Vis
itin
g
Oth
er
trip
purp
ose
Car
availa
ble
No c
ar
availa
ble
0-1
trips p
er
week
2-4
tri
ps a
week
5 o
r m
ore
tri
ps a
week
12-1
8 y
ear
18-2
7 y
ear
28-4
0 y
ear
41-6
4 y
ear
65-8
9 y
ear
Man
Wom
an
Input characteristics
Pu
nctu
ali
ty a
pp
recia
tio
n (
mark
1-1
0)
Figure I.4 Contribution of input characteristics to punctuality (HTM)
Results appreciation of personnel friendliness
6
7
8
HT
M U
rban b
us s
erv
ice
s
HT
M U
rban
tra
m s
erv
ice
s
HT
M R
ands
tadR
ail
0-2
5 p
ass
engers
26
-50 p
ass
engers
50-1
00 p
ass
engers
100-5
05 p
ass
engers
Morn
ing p
ea
k
Ev
enin
g p
eak
Off
-peak
Com
mutin
g
Ed
ucatio
n
Vis
itin
g
Oth
er
trip
pu
rpos
e
Car
availa
ble
No c
ar
availa
ble
0-1
trips p
er
wee
k
2-4
trips
a w
ee
k
5 o
r m
ore
trips
a w
ee
k
12-1
8 y
ea
r
18-2
7 y
ea
r
28-4
0 y
ea
r
41-6
4 y
ea
r
65-8
9 y
ea
r
Ma
n
Wo
man
Input characteristics
Pers
on
ne
l fr
ien
dli
ness a
pp
recia
tio
n (
mark
1-1
0)
Figure I.5 Contribution of input characteristics to personnel friendliness (HTM)
I.I.I.I.3 3 3 3 ExpertExpertExpertExpert vision about additional findings vision about additional findings vision about additional findings vision about additional findings
The analyses for HTM suggest that frequent travellers are less satisfied because of delays and delay
information. The amount of passenger is negatively related to the appreciation of punctuality and delay
information. In addition, comfort aspects such as personnel friendliness, noise and cleanness are appreciated
worse when the number of passengers is higher.
In general, the results show that it is very worthwhile to investigate customer satisfaction on a more detailed
level of comparable services. Services within a concession are comparable. Therefore, analyses on detailed
level of sub aspects give relevant information for a specific concession area. Thus, it is advised to repeat such
analyses for other concession areas. This is worthwhile for both operators as authorities. First, operators can
include the results about important sub aspects in the decision of focus points. They also can include the impact
of crowding, travel frequency and age in their policy. Second, public transport authorities can use the
information about traveller groups and important aspects in their public transport vision and tender. The
customer satisfaction results can support public transport visions for a specific concession area.
Increasing customer satisfaction with public transport Appendixes A-26
J. J. J. J. Input OV ViInput OV ViInput OV ViInput OV Visie Drechtstedensie Drechtstedensie Drechtstedensie Drechtsteden
Several descriptives and analyses are used as input for the public transport vision of Drechtsteden (DAV). The
descriptives show the customer group compared to the nation wide average in the first section. Therefore,
snapshots from a powerpoint presentation are showed below in Dutch, because the Province of Zuid-Holland
requires a Dutch report. The second section compares customer satisfaction of DAV to other concessions. Trend
lines are also shown for the DAV concession. The third section describes the contribution of the supply, travel
and traveller’s characteristics regarding the DAV concession.
J.1 J.1 J.1 J.1 Beschrijving van de Beschrijving van de Beschrijving van de Beschrijving van de reizigersreizigersreizigersreizigers binnenbinnenbinnenbinnen DAV en DAV en DAV en DAV en de de de de MerwedelingelijnMerwedelingelijnMerwedelingelijnMerwedelingelijn
Motieven: waarom reist een reiziger?Motieven: waarom reist een reiziger?Motieven: waarom reist een reiziger?Motieven: waarom reist een reiziger?
Conclusie: DAV heeft meer scholieren en minder woon-werk reizigers dan het landelijk gemiddelde
Dagdelen: wanneer reist een reizigerDagdelen: wanneer reist een reizigerDagdelen: wanneer reist een reizigerDagdelen: wanneer reist een reiziger
Conclusie: DAV heeft reizigers overdag doordeweeks
AutobeschikbaarheidAutobeschikbaarheidAutobeschikbaarheidAutobeschikbaarheid Geslacht Geslacht Geslacht Geslacht
Conclusie: de DAV-reiziger beschikt vaker over een auto
Conclusie: man-vrouw verdeling van DAV gelijk aan landelijke verdeling
DAV bussen Trein Merwedelingelijn
DAV bussen Trein Merwedelingelijn
DAV bussen Trein Merwedelingelijn DAV bussen Trein Merwedelingelijn
Increasing customer satisfaction with public transport Appendixes A-27
Hoe vaak reist een reiziger per week?Hoe vaak reist een reiziger per week?Hoe vaak reist een reiziger per week?Hoe vaak reist een reiziger per week?
Conclusie: DAV heeft relatief meer frequente en minder incidentele reizigers.
LeeftijdLeeftijdLeeftijdLeeftijd
Conclusie: de DAV-reiziger is gemiddeld jonger dan landelijk
J.2 J.2 J.2 J.2 Klanttevredenheid in vergelijking met andere concessies en jarenKlanttevredenheid in vergelijking met andere concessies en jarenKlanttevredenheid in vergelijking met andere concessies en jarenKlanttevredenheid in vergelijking met andere concessies en jaren
Goede klanttevredenheidGoede klanttevredenheidGoede klanttevredenheidGoede klanttevredenheid
Conclusie: DAV scoort beter dan het landelijk gemiddelde
DAV bussen Trein Merwedelingelijn
DAV bussen Trein Merwedelingelijn
DAV bussen Trein Merwedelingelijn Landelijk gemiddelde (bus)
Landelijk gemiddelde (alle diensten)
Increasing customer satisfaction with public transport Appendixes A-28
TTTTrendlijn totaaloordeelrendlijn totaaloordeelrendlijn totaaloordeelrendlijn totaaloordeel
Klanttevredenheid (totaaloordeel)
6,6
6,8
7
7,2
7,4
7,6
7,8
8
2006 2007 2008 2009 2010 2011
jaar
gem
idd
eld
cij
fer
(1-1
0) DAV
Merwedelingelijn
Landelijk gemiddelde
(alle diensten)
Landelijk gemiddelde
(stads-streekdiensten)
Conclusies: DAV scoort door de jaren heen iets lager dan andere stads-streekdiensten
TTTTrendlijn rendlijn rendlijn rendlijn deeloordelendeeloordelendeeloordelendeeloordelen
]
Conclusies: Schommelende vooruitgang
J.3 J.3 J.3 J.3 Relaties reizigers en klanttevredenheidRelaties reizigers en klanttevredenheidRelaties reizigers en klanttevredenheidRelaties reizigers en klanttevredenheid (DAV en Merwedelingelijn)(DAV en Merwedelingelijn)(DAV en Merwedelingelijn)(DAV en Merwedelingelijn)
De focus kan het beste uitgaan naar reizigers die minder tevreden zijn:
■ Frequente reizigers zijn gemiddeld 0,4 punt ontevredener dan incidentele reizigers!
■ Studenten, scholieren en commuters zijn minder tevreden (-0,1 punt t.o.v. andere motieven)
■ Reizigers tot 40 jaar zijn gemiddeld 0,4 punt ontevredener dan 65+ers
■ Reizigers in de avondspits zijn minder tevreden (-0,1 punt t.o.v. andere momenten)
Bovenstaande focusgroepen zijn gelijk aan de landelijke uitkomsten.
Increasing customer satisfaction with public transport Appendixes A-29
KKKK.... Methods to determine Methods to determine Methods to determine Methods to determine the the the the priority of improvements priority of improvements priority of improvements priority of improvements
This Appendix show different tools to determine what the most important aspect is to improve, see Figure K.1
and K.2. Besides these tools, a public transport expert is needed.
Figure K.1 Result of question 15: ‘what aspect should be improved first?’ (Klantenbarometer data 2010)
Figure K.2 shows an interest x satisfaction graph. The graph can be used to prioritize. The aspects with low
appreciation and high interest are the main aspects to improve.
Figure K.2 Example of ‘Interest-satisfaction graph’ (Klantenbarometer data 2010)
gemak kopen vervoerbewijs;
19,3%
kans op een zitplaats; 9,9%
netheid van het voertuig; 8,5%
geluid in het voertuig; 8,3%
anders; 8,2%stiptheid van deze rit; 6,7%
reissnelheid van deze rit; 6,7%
niets; 6,7%
klantvriendelijkheid personeel;
5,9%
informatie bij vertragingen;
5,9%
rijstijl van de bestuurder; 5,2%
aantal vertrekmogelijkheden;
3,6%
gemak van in- en uitstappen;
2,5%
informatie op de instaphalte;
2,5% Improvements wished by passengers
Increasing customer satisfaction with public transport Appendixes A-30
Vestiging Den Haag
Verheeskade 197
2521 DD Den Haag
T (070) 305 30 53
F (070) 389 66 32
Postbus 16770
2500 BT Den Haag
www.goudappel.nl
goudappel@goudappel.nl
Increasing customer satisfaction with public transport Appendixes A-1
Recommended