Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
http://itconfidence2013.wordpress.com
Software Rates vs Price of Function Points
2°International Conference on
IT Data collection, Analysis and Benchmarking
Tokyo (Japan) - October 22, 2014
Rafael de la Fuente
Founder and CEO of LEDAmc
Dácil Castelo
Productivity&Estimations Area
Director
Raúl Fernández
SW Productivity Consultant
Insert here a picture
An Updated Cost Analysis
2 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Software Rates vs Price of Function Points: A cost analysis
Goals of the presentation
Nowadays, in Software Development Contracts the key elements determining the price are: the Rate and the Effort. The possibility to measure the quantity of software produced (the size in Function Points) allows us to assess whether there is a logical connection between the price of the projects and the software actually produced.
The main goal of this presentation is to show what happens with the FP
price in several scenarios:
G1. Implementing estimation models.
G2. Implementing productivity models.
G3. Introducing new competitors in a productivity model.
G4. De-localizing providers.
3 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
• Background
• Objective
• 2014 vs 2012 benchmarking
• Case studies
• Conclusions
Agenda
4 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Background Over the last seven years, LEDAmc has managed the productivity of more than 12,000
development projects of 10 significant clients in Spain (mostly telecommunication and
financial companies). They were mainly small enhancement projects.
The main goal of the measures is to control large contracts of Adaptive Maintenances,
which involves the highest percentage of our clients’ development budget.
In 2012 LEDAmc developed a benchmarking study with some of these projects (we
selected 3,405). The purpose of the study was to analyze the relationship between the
rates of the different software providers and the price of the Function Point that they
were offering their clients. This study was presented:
• In the UKSMA conference in October 2012
• An updated study, in the IT Confidence Conference in Rio, in October 2013.
5 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Background The basic numbers of the 2012 sample were:
• 10 Clients
• 14 Providers
• 3,405 Projects
• 196,356 UFP
• 2,168,192 Man Hours
• 69,926,907 Million Euros
And the main conclusions were:
Pressure to lower rates ends up in lower productivity and higher FP price.
There is not a logical relationship between Rates and FP Price.
The performance of the providers changes dramatically among clients.
Clients with only one development provider have the highest FP Price.
6 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Background
CLIENT-PROVIDER
R
A
T
E
F
P
P
R
I
C
E
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Pressure to lower rates ends up in lower productivity and higher FP price.
7 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Background
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
The same provider for different clients, offers significant differences in FP price.
R
A
T
E
F
P
P
R
I
C
E
CLIENT-PROVIDER
8 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Background
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
F
P
P
R
I
C
E
The price of the FP with only one provider is higher.
CLIENT-PROVIDER
MULTI-PROVIDER CLIENTS
ONE-PROVIDER CLIENTS
9 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Objective LEDAmc has been continuously working for most of those clients, implementing Software
Development Productivity Models over the last two years.
Our productivity benchmarking data base has grown: It has more data for the same
clients and new clients have been included.
The main goal of this presentation is to show what has happened in these two years with
clients who have worked hard to improve their software development productivity.
We will first explain the general evolution of the Function Point Price for those clients,
comparing 2012 and 2014 benchmarks.
Afterwards, we will focus on three case studies that are especially significant.
Finally, we will draw conclusions on the economics of software management, based on
our results.
10 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
2014 vs 2012 benchmarking The 2014 sample were:
• Clients that were in the 2012 study and have implemented some estimation or
productivity control
• 4 Clients
• 12 Providers
• 883 Projects
• 44,897 UFP
• 415,768 Man Hours
• 15,428,772 Million Euros
What has happened with the FP price?
11 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
2014 vs 2012 benchmarking
R
A
T
E
CLIENT-PROVIDER
The rates have remained almost the same.
12 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
2014 vs 2012 benchmarking
The FP price is more homogeneous in 2014 than 2012.
CLIENT-PROVIDER
F
P
P
R
I
C
E
13 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
2014 vs 2012 benchmarking
R
A
T
E
Apparently there is no relation between FP price and rate…
F
P
P
R
I
C
E
CLIENT-PROVIDER
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Rates vs FP Price in 2014
14 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
2014 vs 2012 benchmarking
.. but if we group the data by client, the results look different.
CLIENT A CLIENT B CLIENT C CLIENT D
Rates vs FP Price in 2014
R
A
T
E
F
P
P
R
I
C
E
15 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
2014 vs 2012 benchmarking
There are big differences in the FP price for the same provider with different clients.
CLIENT A CLIENT B CLIENT C CLIENT D
R
A
T
E
F
P
P
R
I
C
E
Rates vs FP Price in 2014
P1 P1 P2 P2 P3 P3 P3
16 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
2014 vs 2012 benchmarking
Conclusions
The results are homogeneous on a client basis because they have implemented
estimation or productivity control models that allow them control the FP price.
Because there is no transparency in software development and the public FP price does
not exist in the market, the FP price stills remains very different among the clients.
The natural differences of the systems architectures and processes should not imply
such FP price differences.
… now we will present some case studies highlighting the change that has been
produced for these clients who have introduced control systems.
17 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Case studies
High FP price
Only one provider
No estimation control
Case Study 1 Competition among providers + Estimation Model
INITIAL SITUATION OF THE CLIENT (2012)
18 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Case studies
During 2013, the implementation of a centralized mechanism of estimation control has
allowed to objectively contrast the costs offered by providers with corporate cost
strategies.
It includes functional sizing, as well as other sizing methods (for parametrization,
technical changes, test support,..)
Benchmarking studies helped establish a cost strategy and a target price for FP .
Introducing a new development provider fosters competitiveness among providers and
helps establish a FP price within market values.
Case Study 1 Competition among providers + Estimation Model INPUTS (2013)
19 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0
10
20
30
40
50
60
70
80
90
100
TOTAL 1 2PROV 1 PROV 2
INPUTS (2013)
Case studies
1. The rate was only
reduced by 5%
2. New development
provider worked with
a similar rate
R
A
T
E
2 1
Case Study 1 Competition among providers + Estimation Model
20 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0
10
20
30
40
50
60
70
80
90
100
TOTAL 1 2PROV 1 PROV 2
INPUTS (2013)
Case studies
R
A
T
E
0
10
20
30
40
50
60
70
80
90
100
TOTAL 1 2
F
P
P
R
I
C
E
The FP price was
reduced by 44 %
Similar FP Price
among providers
1. The rate was only
reduced by 5%
2. New development
provider worked with
a similar rate
Case Study 1 Competition among providers + Estimation Model 2
1
PROV 1 PROV 2
21 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Case studies
A good estimating process was used for all projects
44% reduction in FP price
More than 10 Million euros saved
FP Price within market values
Two competing providers in 2013 – Four in 2014
FINAL SITUATION (2014)
Case Study 1 Competition among providers + Estimation Model
22 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Case studies
Specialized providers
On-site development
High rates and very low FP price
Estimation and productivity control implemented
F
P
P
R
I
C
E
INITIAL SITUATION OF THE CLIENT (2012)
Case Study 2 De-localization of providers
23 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Case studies
The client strategy was to proceed to the de-localization of the providers to a
typical outsourced model introducing new development providers with lower rates.
The client also decided to create an in-house development team in order to
preserve and extend the know-how.
Case Study 2 De-localization of providers INPUTS (2013)
24 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Case studies
0,00
10,00
20,00
30,00
40,00
50,00
60,00
70,00
80,00
90,00
100,00
TOTAL 1 2 3 4
R
A
T
E
1. New provider with
lower rate
2. In-house development
rate
1 2
INPUTS (2013)
Case Study 2 De-localization of providers
PROV 1 PROV 2 PROV 3 PROV 4
25 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0,00
10,00
20,00
30,00
40,00
50,00
60,00
70,00
80,00
90,00
100,00
TOTAL 1 2 3 4
Case studies
0,00
10,00
20,00
30,00
40,00
50,00
60,00
70,00
80,00
90,00
100,00
TOTAL 1 2 3 4
R
A
T
E
1. New provider with
lower rate
2. In-house development
rate
F
P
P
R
I
C
E
1 2
Logical behavior of FP
price
In-house developments
at lower FP cost than
in outsourced scenario
INPUTS (2013)
Case Study 2 De-localization of providers
PROV 1 PROV 2 PROV 3 PROV 4
PROV 1 PROV 2 PROV 3 PROV 4
26 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
Case studies Case Study 2 De-localization of providers
FP price controlled
Outsourcing process controlled
2014 situation
F
P
P
R
I
C
E
FINAL SITUATION (2014)
27 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Case studies
Ratio of the rates: 1.1
Ratio of the FP Price: 2.3 F
P
P
R
I
C
E
INITIAL SITUATION OF THE CLIENT (2012)
Case Study 3 Controlling large maintenance contracts
28 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Case studies
During 2012 the client implemented a productivity control for maintenance
development.
An annual productivity SLA for all the providers was established.
The client also implemented a productivity and quality providers ranking.
Every two months the client presented to the providers their position in the ranking.
Case Study 3 Controlling large maintenance contracts INPUTS (2013)
29 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0,00
10,00
20,00
30,00
40,00
50,00
60,00
70,00
80,00
90,00
100,00
TOTAL 1 2 3 4 5 6
Case studies
R
A
T
E
Renegotiation of rates
Ratio 2012: 1.1
Ratio 2014: 1.0
INPUTS (2013)
Case Study 3 Controlling large maintenance contracts
30 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0,00
10,00
20,00
30,00
40,00
50,00
60,00
70,00
80,00
90,00
100,00
TOTAL 1 2 3 4 5 6
0,00
10,00
20,00
30,00
40,00
50,00
60,00
70,00
80,00
90,00
100,00
TOTAL 1 2 3 4 5 6
Case studies
R
A
T
E
Renegotiation of rates
Ratio 2012: 1.1
Ratio 2014: 1.0
F
P
P
R
I
C
E
FP cost control
Ratio 2012: 2.3
Ratio 2014: 1.7
INPUTS (2013)
Case Study 3 Controlling large maintenance contracts
31 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
Case studies Case Study 3 Controlling large maintenance contracts
Uniformity of FP price
9.5 million euros saved
2014 situation
F
P
P
R
I
C
E
FINAL SITUATION (2014)
32 IT Confidence 2014 – October 22, 2014 http://itconfidence2014.wordpress.com
Conclusions
1. It is relatively easy to influence FP price. The main levers are:
>>> A centralized estimation control
>>> A development productivity control
>>> Several providers in competition
2. Working on those key elements the results are much better than working only
on the rates:
>>> The behavior of the unit cost of production (FP price) is more homogeneous.
>>> The cost of the development process can be controlled.
3. Diversification of providers with joint management results in a better uniformity
of the FP price.