Consumers and Their Perception of Private Labels in …gsblibrary.uct.ac.za/researchreports/2010/Karan.pdfWhen a retailer sells a product with a brand ... United States and more than

Copyright UCT

Consumers and Their Perception of Private Labels in Turkey:

A study across categories

A Research Report

The Graduate School of Business

University of Cape Town

In partial fulfillment of the requirements for the

Masters of Business Administration Degree

by

Selen Karan

10 December 2010

Copyright UCT

Copyright UCT

PRIVATE LABELS IN TURKEY

2

Table of Contents

Plagiarism Declaration ......................................................................................................................................... 3

Private Labels in Turkey...................................................................................................................................... 4

Research Area and Problem ............................................................................................................................ 4

Customer Perception of Private Label Brands .............................................................................................. 5

Research Questions and Corresponding Hypotheses .................................................................................... 7

Research Assumptions.................................................................................................................................... 10

Research Ethics ............................................................................................................................................... 11

Literature Review ............................................................................................................................................... 11

Consumer Behavior Across Categories ........................................................................................................ 13

Personal Factors ......................................................................................................................................... 13

Product Factors .......................................................................................................................................... 14

Situational Factors ..................................................................................................................................... 14

Perceived Risk Factors ............................................................................................................................... 14

Conclusion ...................................................................................................................................................... 15

Research Methodology ....................................................................................................................................... 16

Research Approach and Strategy ................................................................................................................. 16

Research Design, Data Collection and Research Instruments ................................................................... 16

Sampling ......................................................................................................................................................... 19

Data Analysis Methods .................................................................................................................................. 20

Research Findings, Analysis, and Discussion ................................................................................................... 21

Research Findings ........................................................................................................................................... 21

Demographics ............................................................................................................................................. 21

Cross-category findings ............................................................................................................................. 23

Research Analysis and Discussion ................................................................................................................ 24

Correlation Analysis .................................................................................................................................. 27

Hypothesis Testing ......................................................................................................................................... 29

Research Limitations ..................................................................................................................................... 32

Research Conclusions ......................................................................................................................................... 33

Manegerial Implications ................................................................................................................................ 35

Future Research Directions ............................................................................................................................... 35

References ........................................................................................................................................................... 37

Appendix: Scale of Research ............................................................................................................................. 41

Copyright UCT


3

Plagiarism Declaration

1. I know that plagiarism is wrong. Plagiarism is to use another’s work and pretend that it is

one’s own.

2. I have used a recognized convention for citation and referencing. Each significant

contribution and quotation from the works of other people has been attributed, cited and

referenced.

3. This report is my own work.

4. I have not allowed, and will not allow anyone to copy my work with the intention of

passing it off as his or her own work.

5. I acknowledge that copying someone else’s assignment or essay, or part of it, is wrong,

and declare that this is my own work.

Signature

Copyright UCT


4

Private Labels in Turkey

Research Area and Problem

When a retailer sells a product with a brand name that the retailer owns, the product is

called a private label product (PL) and the brand is called a private label brand (PLB, also

called “house brands” and “store brands”). American food retailers Kroger and A&P sold the

first PLs a century ago, but worldwide development of PLBs is more recent. In 1976,

Carrefour, the French hypermarket chain, pioneered PLBs in Europe with such success that

PLBs contributed 3.5% of company’s sales turnover by the end of the first year. American

and UK (1978) and German (1978) retailers quickly followed. Rapid worldwide expansion

of PLBs continues today.

PLB consumer packaged goods (CPGs, called fast-moving consumer goods) have been

among the fastest growing product categories for PLBs. CPGs are especially appealing to

retailers because profitability is higher than on national brands, which are promoted by other

retailers and thus subject to price competition. Although penetration varies across countries,

it is now common to see major supermarkets, hypermarkets, drug stores and discounters

offering successful PLBs in many product categories, including fresh, canned, frozen, and

dry foods; snacks, ethnic specialties, pet foods, health and beauty, over-the-counter drugs,

cosmetics, household and laundry products, DIY, lawn and garden, paints, hardware, and

auto aftercare (Herstein & Gamliel, 2006).

According to a recent report from the Private Label Manufacturer’s Association in

2010, “PL gained market share in 14 of the 20 countries tracked by Nielsen and currently

represents at least one of every five products in all but two of the countries” (Private Label

Manufacturers Association, 2010). The report also indicates that PLBs obtained record

market shares across Europe last year. For the first time PLBs account for at least 40% of all

CPGs sold in a total of five countries in Europe. Although the major gains were observed in

the western markets such as France, Germany, Spain, The Netherlands and Belgium, the

growth was strongest in emerging retail markets such as Poland, Hungary, Slovakia and

Turkey. PL share per country can be seen in Figure 1 below. Furthermore, one in every five

items sold in the U.S. supermarkets, mass merchandisers, and drug stores is a private label

product and total PL sales was more than 81billion US Dollars, which accounted for a unit

share of 22.8% (AC Nielsen, 2008). There are around three thousand PLB producers in the

United States and more than half of the branded CPG manufacturers also produce PLBs,

which illustrates that the PLB industry not only offered a market for retailers, but for the

Copyright UCT


5

manufacturers as well.

Figure 1. Private label share of fast-moving consumer good sales in Europe.

Faced with an ever-increasing consumer interest and growth in sales, many retailers

have begun introducing private label as a strategy for creating competitive advantage,

especially in the recent economic downturn (Walsh & Mitchell, 2010). Retailers have

illustrated ambition in expanding their share of PLB sales for several reasons. Among the

most important of these are higher margins, lessened dependency on manufacturers,

improvements in store image and higher consumer store loyalty (Ailawadi & Harlam 2004;

Narasimhan & Wilcox, 1998; Pauwels & Srinivasan, 2004). However, convincing

consumers to shift from their preferred national brand (NB) to private label is still a

challenge. Therefore, it is crucial for retailers, who aim to gain higher customer store loyalty

through PLB products, to have a good understanding of the consumer perception of PLB and

the reasons behind their intention to buy these products (Walsh & Mitchell).

Customer Perception of Private Label Brands

PLBs were originally introduced as a more affordable alternative to national brands. It

is evident that the economic downturns in history, including the recent recession, played

determining roles in changing consumer behavior and causing a major shift in preference

from NBs to PLBs (Corstjens & Lal, 2000; Deloitte Debates, 2010). In Turkey where the

PLBs are on average 30% less expensive compared to NBs, PL sales showed a 50% increase

during the recession in 2001 (“Private Label”, 2008). Although many perceived this change

to be a temporary trend and expected consumers to go back to their usual purchasing habits,

they were proven wrong. It can be seen that PL sales have grown faster than NB sales and

Copyright UCT


6

have gained higher levels of penetration in comparison to figures during recession (Corstjens

& Lal, 2000).

Prior research has found that consumers prefer NBs and are willing to pay more for

them (Walsh & Mitchell, 2010). This has often been the case because PLBs are thought as

having lower quality or social status. However, evidence suggests that this is changing.

According to a recent Deloitte study, more than 90 of 100 people say they have permanently

changed their buying behavior during the recent recession as they found little or no difference

between PLB and NB (Deloitte Debates, 2010). Similarly, AC Nielsen (2005) consumer

survey conducted in 38 European countries supports that many consumers believe PLB to be

a good alternative to NB products and they are willing to spend more on PLB. As the PLB

market matured, more retailers began to offer premium PL products that compete directly

with the NBs in terms of quality and image. The similar packaging, promotion and

increasing number of retailers running TV advertisements make it harder for consumers to

differentiate PLB products from the NB (Putsis & Dhar, 2001).

According to Narasimhan and Wilcox (1998, p. 574) consumers’ intention to switch

from a national brand to a private label is linked to “the risks consumers associate with

making purchases in a given category, as well as measures describing the ability of a retailer

to offer a private label of a comparable quality to the national brands in a given category”. It

has been found that the level of perceived risk, and therefore the attitude toward PLB vary

across product categories (e.g., Batra & Sinha, 2000, Hoch & Banerji, 1993; Narasimhan &

Wilcox, 1998). Aluminum foils, paper products and plastic bags were among the first product

categories in which Turkish major supermarkets and hypermarkets began competing with

PLBs. By 2006, competition was common in paper products, liquid oil, dry pulses, and food

categories, with sales attributable to PLBs growing by 22% on the back of only 4-5% price

inflation (Market Brands Report, 2006, as cited in Private Label Turkey). The turnover

drivers in 2006 were paper products, pulses and beverages. According to Dhar and Hoch

(1997), the main reason for PLB share to vary across retailers is the differences among

product categories.

Previous research on PLB mainly looked at the issue from manufacturer and retailer

perspectives. Researchers have mostly investigated the success of PLB over NB, examining

the promotional activities, financial factors, size and margin variations between categories

and necessary technological investments (Hoch & Banerji, 1993). Belizzi, Hamilton,

Krueckeberg, and Martin (1981) showed that NB buyers, who are more prone to preferring

Copyright UCT


7

name brands, tend to be more influenced by advertising as opposed to PLB buyers. The

authors argue NBs to be more successful in terms of marketing practices, heavy advertising,

and packaging, which in the end determine the perceptual differences. Supporting this

argument, Hoch and Banerji found PLBs to have higher margins in categories where their NB

competitors do not spend much on advertising.

Research has been limited on the roles of consumer attitudes and perceptual differences

in determining the success of PLB, especially across product categories. Some studies

examined the role of consumer-level factors in PLB purchasing behavior, but have not

considered cross-category differences (Richardson, Jain & Dick, 1996). Batra and Sinha

(2000), on the other hand, looked at the determinants of the level of perceived risk and how

the perceived risk changes across product categories. They, however, have not looked at the

demographics of PLB consumers.

There is no previous research examined the consumer-level perceptional differences in

PL, across the product categories, in Turkey. In light of the antecedents from existing

research (Batra & Sinha, 2000; Walsh & Mitchell, 2010), this study examined the effect of

four determining factors of perceived risk on consumers’ intention to buy PLBs, across nine

product categories. It is believed that understanding the reasons for causing these cross-

category differences might act as a guide for retailers to increase their PL shares and for NB

manufacturers to develop a more targeted strategy in their fight against the PLB.

Research Questions and Corresponding Hypotheses

Low-income consumers in Turkey commonly prefer discount markets like BIM,

Tansas, SOK, and DiaSA. These markets offer discounts on national brands and also sell

their own private labeled products. However, today PL products are also common in

supermarkets like Migros, Carrefour, and Macro whose target consumer range from middle

to upper income levels. As mentioned earlier, execution of private label brands is no longer a

strategy used to attract and satisfy only low-income consumers. PLB proneness changes

across income levels and product categories. Retailers need to have an insight on the PLB

behaviors of their target consumer in order to succeed.

In general terms, this research tried to understand how middle and upper middle-class

Turkish consumers differ in their approach to PLB. Thus, the main question was if the

selected factors that determined the perceived risks attached to different product categories

explain consumers’ proneness to buy PLBs. Furthermore, the study will also examine to see

if there is any relationship between certain demographic factors, such as income, age and

Copyright UCT


8

gender and one’s propensity to purchase PLB.

Previous research predicted a negative correlation between the perceived risk, and the

purchase of PLB. That is to say, consumers buy PLBs if the perceived risk level is low or in

other words, they will buy NBs if the perceived risk level is high. According to Narasimhan

and Wilcox (1998), consumer’s willingness to choose NB over PLB is linked to the perceived

risk associated with making the purchase in a certain category. Therefore, it is crucial to

understand the factors determining the level of perceived risk in a given category. The four

determinants included in this research were, consumer’s price consciousness in that category,

fear of choosing the wrong brand, the degree of variation in quality across brands, and the

search / experience nature of the product (Taken from Batra & Sinha, 2000).

It is evident that some consumers judge the quality of a product on the basis of price,

rather than its features (Newman & Becknell, 1970). Price conscious consumers are

generally perceived to be low-income people, who are more prone to buying PLBs and more

likely to follow the deals. These consumers do not necessarily associate price with the

quality of the product. Although, the trend is changing and low price does not always equate

to low quality, it is still valid that, the low-income consumers are more price-sensitive, and

therefore, have more positive attitude toward PLBs.

Research Question 1

Does price consciousness affect consumers’ intention to buy PLBs (IBPLB)?

Hypothesis 1

Ho: Price conscious does not determine consumers’ IBPLB.

Ha: Price conscious determines consumers’ IBPLB.

A personal factor, self-perception or in other terms, self-image affects consumer

purchase intention (Blackwell, Miniard, & Engel, 2006). Some customers may choose to

shop at certain stores, as they want to be associated with who they perceive is the store’s

clientele or they may avoid certain brands because of the social status or biases associated

with the buyers of those brands (Blackwell, Miniard, & Engel). Consumers with high degree

of self-perception might be more likely to purchase NB products with high brand reputation

rather than PLB. For example, when the PLB were first introduced in the market, consumers

bought PLB products for their own personal consumption, but avoided being seen buying or

using the product when interacting with others (Sheath & McGoldrick, 1981). Although, this

attitude is seen to have changed, as the quality and brand reputation attached to PLB has

increased over the years, social risk attached to PLB is still an important determining factor

Copyright UCT


9

of the attitude toward the product (Zielke & Dobbelstein, 2007; Walsh & Mitchell, 2010). As

a matter of course, the consumers’ purchasing behavior of PLB shows differences across

categories, as the perceived social risk of PLB changes in between the product categories.

Products like butter, soap or aluminum foil may have low social risks while toilet paper,

coffee and shampoo have higher social risks attached to them.

In addition to social risk factors related to the self-perception, profile risk associated

with a certain PL product category also defines the consumers’ final decision. This perceived

risk is related to consumers’ uncertainty about the product experience (Blackwell, Miniard, &

Engel, 2006). These risks can be physical, performance, or financial. In higher-risk product

categories consumers tend to search for ways to minimize the purchasing risk. This is

because they fear to make a purchasing mistake. Therefore, they may buy the most

commonly preferred brand, the most advertised brand, or the most expensive brand

(Hawkins, Best, & Coney 1986, as cited in Narasimhan & Wilcox, 1998, p. 577). However,

the associated risk changes across product categories. For example, butter might have a lower

risk profile in comparison to baby food.

Research Question 2

Does the level of perceived consequences of making a purchasing mistake affect consumers

IBPLB in a given product category?

Hypothesis 2

Ho: The level of perceived consequences of making a purchasing mistake does not

determine consumers’ IBPLB in a given category.

Ha: Consumers are more prone to buy PLBs in product categories where they believe to

have lower consequences of making a purchasing mistake.

Batra and Sinha (2000) found that the perceived risk was higher when different brands

in the category were seen to vary in quality. When the degree of perceived quality variation

between brands increases in a given category, the fear of purchasing the wrong brand rises as

well. The findings of Hoch and Banerji (1993) support this argument as well. They showed

that PL share is higher in categories where the quality variability was low.

Research Question 3

Does the level of variability in quality affect consumers’ IBPLB in a given product category?

Copyright UCT


10

Hypothesis 3

Ho: The level of perceived variability in quality does not determine consumers’ IBPLB in a

given category.

Ha: Consumers are more prone to buy PLBs in product categories where they perceive lower

variability in quality levels across brands.

The last determinant of the perceived risk used in this research, refers to the search /

experience nature of the product category. Prior research has established that consumers find

NBs to be superior to PLBs in terms of satisfaction, taste, package understandability, purity,

variety, aroma, freshness, and packaging information (Belizzi, Hamilton, Krueckeberg, &

Martin, 1981). It has been argued that the nature of the product features determines

consumer’s PLB proneness. Attributes such as taste, aroma, color, and so forth relate to the

“experience” characteristics of a product, whereas, attributes that can be accessed prior to

purchase, such as ingredients or color fall under the “search” qualities of a product.

Accordingly, it is foreseen that product categories, where the consumer seeks to “experience”

before any purchase, are perceived to have a higher risk. Batra and Sinha (2000) found

consumers to be less likely to purchase PLBs in product categories with many “experience”

attributes.

Research Question 4

Does search versus experience characteristics of a product determine consumers’ IBPLB?

Hypothesis 4

Ho: Informative packaging allowing an easy assessment of product attributes does not

determine consumers’ IBPLB.

Ha: Consumers are more prone to buying PLBs in product categories where they can easily

assess the product attributes or benefits based on the given information on the package alone.

Research Assumptions

There are three main assumptions that have been made to facilitate both collection and

analysis of the data. The first assumption was that the participants of the research would be

diverse enough in their perception of private label brands. That is to say, that the respondents

were assumed to differ in their shopping behaviors. The second assumption was that the

respondents would reflect the general consumer perception of PLBs in Turkey. Lastly, the

assumption was made that the participants would be honest in their answers.

Research Ethics

The research held no tangible risk of harm to its respondents. Yet, the possible

Copyright UCT


11

challenges that could have emerged due to the lack of consent and data protection were

recognized (Bryman & Bell, 2008). Respondents were given the choice to not take part in

the research if they did not feel comfortable enough about either the content of the survey or

their privacy. In order to assure that the respondents were fully informed about the details of

the research and the anonymity of their responses, an introductory notice was given. The

survey was anonymous and no personal information other than age, income and gender was

asked. Hence, identification of any person who participated in the research is unlikely.

Prior to the conducting fieldwork, the Graduate School of Business Research Ethics

Committee approved the research proposal and questionnaire.

Literature Review

There are many research in the literature examined the unexpected success of private

label. Richardson, Jain, and Dick (1996, also cited in Batra & Sinha, 2000, p. 177) developed

a detailed framework that presents three different consumer-level factors, in determining the

intentions for buying PLB. These are (1) demographic factors (income, family size, age, and

education), (2) individual differences (degree of reliance by the consumer on extrinsic cues,

and (3) consumer perceptions of the product category (degree of perceived quality, level of

perceived risk, and perceived value for money) and degree of consumer knowledge about the

category. Their framework, however, lacks information about the fact that some of these

perceptual factors may vary across product categories. In another research, Richardson, Jain,

and Dick (1994) found that consumers perceive PLBs as low quality in comparison to NBs.

They also saw that consumers, whose perception of products were influenced by extrinsic

cues, such as brand, packaging and pricing were less likely to purchase PLBs.

Ailawadi, Neslin, and Gedenk (2001) found the direct effect of demographic variables

on PLB purchasing to be weak. On the other hand demographics were seen to have

significant influence on psychographic characteristics, which determine the shopping

behavior of a consumer and therefore can be useful in segmentation, targeting and marketing.

Their research showed that the psychographic drivers of private label purchasing and national

brand purchasing differ substantially. For example, consumers with higher income levels are

less price conscious, older consumers are more likely to be shopping mavens, and women are

more likely to be innovative, impulsive, and planners than men. Furthermore, higher

education was related to higher financial status and higher levels of quality consciousness,

both of which were negatively correlated with intentions to purchase PLB. According to

their research PLB buyers were not quality conscious but price conscious shoppers. Belizzi

Copyright UCT


12

et al also showed that private label shoppers value price, taste and value for money for their

choice of brands, while national brand shoppers base their choice mainly on quality,

packaging, experience and value for money. However, findings of Sethuraman (1992) and

Hoch and Banerji (1993) proved the opposite. Sethuraman empirically demonstrated a

negative correlation between the price difference between PLB and NB and the category

share of PL. Hoch & Banerji on the other hand found perceived quality to be more important

than the price in determining the PL category share.

According to the results gathered from a more recent research conducted by Walsh and

Mitchell (2010), there is an increase in quality perceptions and a more positive attitude

toward PLB. Researchers explained that due to the increased reputation and quality of PLB

products consumers no longer perceive PLB as a separate category of product. Furthermore,

they proved a positive relationship between quality, price and emotional-related value and

consumers’ intentions to buy PLB.

In the name of understanding the success factors for the high market share of PLB

industry, many researchers examined the quality, price and customer related influencers.

Many agreed that high-quality PLBs make store differentiation easier to achieve and lower

NB price premium (Richardson, Jain, & Dick, 1996). On the other hand, higher quality

serves as a justification for higher prices as these products are perceived as premium and seen

equal to NBs (Mills, 1995). De Wulf et al. (2005) showed that as the NBs increase in price

consumers either prefer PLB or do not buy at all. On the contrary, according to Amrouche et

al. (2007), in case of price increase NB consumers are more likely to switch among NB

products rather than to PLB. As for customer related factors, it is agreed that the lower-

income consumers prefer PLB, as they are more price sensitive, especially when the NB-PLB

price gap is high (Putsis & Dhar, 2001; Sethuraman & Cole, 1999). According to Ailawadi et

al. (1999) due to the low-quality perception associated with PLB, belief in the positive

relationship between price and quality discouraged the purchase of lower-priced PLB

products.

In their research Ailawadi et al. (2008) built a simultaneous model looking at the

relationship between a household’s PL share at a chain and its loyalty to that chain, which is

measured as the household’s total spending in the chain (SOW: share of wallet). Their

research proved a two-sided direct correlation between PL share and SOW. Results indicated

a positive relationship between the store loyalty and PL share as SOW initially increased

noticeably along with the PL share until the PL share reached approximately 40% and after

Copyright UCT


13

that SOW actually started to decrease. Although PLB increases store loyalty, retailers should

note that dominant presence of PLB seems to make even the high-level PL buyers feel

constrained in their choices and therefore causes a decline in their overall loyalty to the store.

As noted by the researchers as well, the research fails to analyze the relationship between

SOW and PL across categories. From the previous findings we know that the purchasing

intention for PLB varies across product categories (Sethuraman, 1992). Therefore, it would

be valuable both for the researchers and the retailers to know which categories of products

are more sensitive to changes in the PL share and which are less affected by the PL

dominance in the store.

Consumer Behavior Across Categories

As noted earlier, consumers show changing purchasing behaviors for different product

categories. Previous researches have shown that the level of perceived risk in the category is

a determining factor in PLB purchases (Richardson, Jain, and Dick (1996). Narasimhan and

Wilcox (1998) linked the willingness of consumers to switch from a NB to a PLB to the risk

consumers associate with making purchases in a given category as well as measures

describing the ability of retailer to offer a PL of comparable quality to the NB in the given

category.

Consumers’ decision-making process is a form of problem solving. There are three

distinctive factors identified that the extent of the problem-solving process depends on and

they are (1) degree of involvement, (2) degree of differentiation between alternatives, and (3)

amount of time for deliberation (Blackwell, Miniard, & Engel, 2005). The degree of

involvement is determined by the given importance to the product by the consumers. As the

authors stated within the consumer decision process the involvement happens when intrinsic

personal characteristics like needs, values or self-concept are confronted with a certain

marketing stimuli in a given situation. Blackwell et al. summarized the factors affecting the

level of personal involvement under three categories such as personal, product and

situational. These factors shed some light on the associated risk factors to the decision

making process of consumers for PLB.

Personal Factors

When the outcome of a certain decision affects the person directly in terms of self-

image, health, beauty or physical condition the person tends to feel the need to be involved in

the decision making process on a higher level. Blackwell, Miniard, and Engel (2005)

explained that it is the activation of need and drive that motivates one to get involved

Copyright UCT


14

whereas in the absence of both the person will be less likely to get involved. Buying

cosmetics or medicine are examples of decision processes where there is likely to be high

personal involvement as the products affect the person’s health or self-image directly.

Product Factors

The perceived risk in purchasing or using a product or a brand determines the degree of

involvement. Consumers’ purchase intentions are highly influenced by the risks associated

with the purchasing behavior (Bettman, 1974). Risk may come in many forms such as

physical (risk of physical harm), psychological (social disapproval), performance (fear that

the product will not possess as desired), and financial (fear that the purchasing of the product

will cause loss of earnings).

Situational Factors

Lastly situational factors are mostly related to the perceived social risks. Within this

category the behavior is very much based on the fact whether the product will be consumed

alone or with others. Livesey & Lennon (1978) reported in their research that English

consumers prefer to serve national branded tea to their guests while drinking PL branded tea

when outside of a social context. The more people are concerned about the approval of the

society the more careful and selective they will be in their choice of brands.

Perceived Risk Factors

As mentioned earlier, it is evidently seen that there is a significant difference between

NB proneness and PLM proneness based on the perceived risk associated with the given

product category (Batra & Sinha, 2000; Bettmann, 1974; Richardson, Jain, & Dick, 1996).

Here, the risk is determined based on the degree of fear of making the wrong decision, which

Blackwell, Miniard and Engel (2005) talked about under product factors. Consumer

measures the consequences of making a mistake while deciding between products and

brands. These consequences and therefore the attached risks can be physical, financial, or

psychological risks. For example, in terms of the risk factors involved, purchasing toiletries

is less sensitive than buying baby food as it is less risky to buy beverages as opposed to meat.

There is also a social aspect attached to the level of comfort in making a mistake. As

mentioned under situational factors within a social environment the risk associated with the

purchasing of certain brands or products increases.

As suggested by Batra and Sinha (2000), the PLB proneness increases in product

categories where consumers perceive lower variability in quality levels across brands. In

other words, if there is not much difference in terms of quality between the products within a

Copyright UCT


15

category, consumers are more likely to prefer PLBs. However, on the other hand as the

variety in quality increases consumers are less likely to buy PLBs when they can afford

higher quality products and experience higher satisfaction. Furthermore, the ambiguity of

quality causes the purchasing decision to be based on other things like branding. That is to

say it is likely that a consumer will pick the cheapest national brand rather than purchasing a

private label brand.

Another determinant factor of perceived risk raised by some researchers is the search

versus experience nature (Erdem & Swait, 1998; Walsh & Mitchell, 2000). Sethuraman and

Cole (1999) argued that NBs are more likely to set a premium price if they were higher on

hedonic aspects as opposed to functional (reviewed by Batra & Sinha, 2000). Some

researchers have suspected the relationship between PLB proneness and the nature of product

features. Batra & Sinha suggest that the hedonic attributes like taste, aroma, texture, and

color are characteristics that influence the perception through experience rather than search.

Search on the other hand is defined by pre-purchase attributes that can be verified through

direct observation or product features listed on the back of the package, or any other

accessible sources. Products with high experience attributes require some previous

knowledge. Therefore, the absence of any past experience or sufficient information increases

uncertainty and the level of perceived risk. In such cases, consumers tend to be more prone

to preferring NBs in the name of reducing the risk. Richardson, Jain, and Dick (1996)

advised that consumers might be more prone to prefer PLBs for products with search

attributes rather than experience benefits.

Lastly, the level of price consciousness determines the level of perceived financial risk

attached to a product category. Previous research has proved that a consumer’s level of price

consciousness rises with lower incomes. On the other hand consumers with higher income

are less price conscious. However, it should be noted that low price and PLB are associated

with low quality. Therefore, a quality conscious low-income consumer may still prefer a

relatively cheap national to a private label brand in certain product categories.

Conclusion

As PLBs in the consumer packaged goods industry have been enjoying a worldwide

growth for many years, the subject has become an area of interest for many researchers.

Several studies have examined the socioeconomic and demographic characteristics of PL

buyers, mainly supporting the direct link between PLB’s lower price ranges and consumers’

intention to buy PLB. However, most of the reviews studied either the market structure or

Copyright UCT


16

the PL market from NB manufacturers’ and retailers’ perspectives.

Consumers’ purchasing decisions are highly determined by the perceived risks

associated with buying a product. Following the Batra and Sinha (2000) research as an

example, this paper focused on four consumer related factors driving the perceived risk and

they are fear of making a purchasing mistake (involving self image and social approval),

degree of quality variation in a given category, search / experience nature of the product

category and lastly price consciousness. Previous research (Ailawadi, Pauwels, &

Steenkamp, 2008) also suggested PLB behavior to be investigated across product categories.

Additionally, many researchers (Kumar & Steenkamp, 2007; Narasimhan & Wilcox, 1998;

Richardson, Jain, & Dick, 1996) point out the differing successes of PLBs across categories.

This research, therefore, tried to understand if and how the four determining factors of

perceived risk influence consumers’ changing PLB purchasing behavior across categories.

Research Methodology

Research Approach and Strategy

A quantitative deductive approach was adopted in this research. In several papers, such

as Narasimhan and Wilcox (1998), perceived level of risk was seen to influence PLB

purchasing behavior. The research began with the question whether this observation also

holds in Turkey. The question was modified with the replacement of the independent

variable with four proxies. The theoretical backbone that allowed this modification was

borrowed from Batra and Sinha (2000). Then five hypotheses were postulated to check how

the four proxy variables influenced PLB purchasing behavior. In order to test these

hypotheses across a variety of product categories, a survey containing multiple Likert items

on five point scales was designed and numerical data for the proxy variables were collected

(Bryman & Bell, 2007). The data were later scrutinized using correlation (Leedy & Ormrod,

2010) and regression analyses (Lind, Marchal, & Wathen, 2010). The hypotheses were

confirmed or rejected based on these statistical analyses. The quantitative approach provided

the precision required for this in-depth analysis (Bryman & Bell, 2007).

Research Design, Data Collection and Research Instruments

A social survey research design was used to collect and analyze the data. The main aim

of the research was to see how the risk middle and upper-middle class Turkish consumers

associate with PLBs differs across product categories, and how this perceived risk defines the

end purchasing behavior. Secondly, the research was designed to control if demographics

such as age, gender and income have any effect on one’s level of intention to purchase

Copyright UCT


17

private label brands.

Thirteen questions were administered using an online survey. The survey was directly

e-mailed to a total of 50 people who the researcher is directly or indirectly acquainted with,

and it was advertised on Facebook and LinkedIn pages of different profile groups.

Furthermore, the respondents were asked to share the survey with their family and friends.

Hence, more than half of the total respondents are not acquaintances of the researcher. It is

evident that web-based surveys have increasingly become the choice of many market

research studies. Compared to traditional methods, online surveys have many advantages

such as high quality, speed, and cost-effectiveness (Aaker, Kumar & Day, 2004). Having no

interviewers involved in the process eliminates the possible interviewer error and bias.

Furthermore, an on-line environment assures voluntary participation and therefore

encourages honest and spontaneous answers (Aaker, Kumar & Day). It also allows the

respondent to take the survey in his own time free of any disturbances from his shopping time

and of any pressure from the interviewer. Lastly, and most importantly, increased

availability of Internet and on-line population makes it possible for researchers to reach a

broader range of demographics (Aaker, Kumar & Day). In this particular study, it was

important to attract respondents from different demographic backgrounds and to attain a

representative distribution across the gender, age and income categories indicated in the

survey. According to the latest Internet Usage Stats and Market Report (2010) launched by

ITU (International Telecommunications Union), Internet penetration in Turkey is 45%.

Given the time constraint, the positive attributes of a web-based market research and the cost

involved with the alternative methodologies, an online survey was judged to be the most

appropriate design for the purposes of this research.

The main concern of this research was to identify how the consumer perception of PLB

varies across different product categories along with the level of perceived risk. Therefore,

the selection of product categories was a crucial step at the start of the survey design. Nine

categories selected for this study were as follows: toilet paper, oil, detergent, dry pulses,

yoghurt, tea, cheese, soap and shampoo. They were chosen according to the Retailing

Institute’s private label brands report (2006) that detailed the most popular PLB product

categories, and to the retailer trends data that was gathered by AC Nielsen (2006) from the

six biggest supermarkets and hypermarkets in Turkey. Food, personal care products and

cleaning products were indeed found to be the top three product categories in terms of growth

in 2006. Within these categories, dry pulses, paper products, milk, yoghurt and cheese were

Copyright UCT


18

the top five items with the highest turnover in 2006. Furthermore, these products are

believed to be representative in the sense that the expected risks that consumers assign to

these items occupy a large spectrum. Products like shampoo, toilet paper, cheese, and

yoghurt are expected to have higher perceived risks (both social and physical risks). On the

other hand dry pulses, oil, soap, detergent and tea are predicted to be products with lower

risks. Unlike in previous research, respondents were asked to answer each question for each

of the nine product categories at once. This made it possible to circumvent any annoyance

and carelessness that could arise when the respondent is asked the same question nine times

repeatedly.

Four category-based determinants of the level of perceived risk assigned to PLBs were

included in the survey as independent variables: (1) price-sensitivity, (2) fear of making a

purchase mistake, (3) degree of quality variation, and (3) “search” versus “experience” nature

of product category. Tabachnick and Fidell (as cited in Pallant, 2005, p. 142) suggest a

formula for calculating the samle size requirements and gives information on the number of

independent variables one should use. The formula is: N > 50 + 8m (where m = number of

independent variables). Furthermore, they explain that for stepwise regression the ratio of 40

cases for every independent variable should be expected. Therefore, in the current reserach

with 108 samples four independent variables were judged to be enough. All scales were

measured on a 5-point Likert scale (1: strongly disagree and 5: strongly agree) (Batra &

Sinha, 2000; Shannon & Mandhachitara, 2005), and were later used to explain the variation

of consumer behaviors across categories.

In line with the previous research (Batra & Sinha, 2000; Walsh & Mitchell, 2010;

Zielke & Dobbelstein, 2007), consumption habits were measured with three items on a five-

point scale. Consumers were asked to rate their purchasing of (1) known brands that are sold

in almost every store, (2) private label brands sold at their preferred retailer only, and (3) no

brand in particular. This measurement was done in order to examine the expected

relationship between one’s likeliness to purchase PLBs and the level of perceived risk.

Lastly, gender, age and monthly income were used in the survey as control variables

because in previous studies they have been associated with PLB purchasing behavior (Dar &

Hoch, 1997; Richardson, Jain, & Dick, 1996). On the other hand Walsh and Mitchell (2010)

did not find age and gender to add to the explanation of consumer’s PLB purchasing

behavior. A dichotomous variable was used to measure gender – female were coded one (1)

while males were coded two (2). Age and income were measured using six brackets.

Copyright UCT


19

Sampling

The research population is comprised of mostly middle and upper-middle class Turkish

consumers with Internet access, living mainly in Istanbul. Respondents were consumers with

a disposable income, either from their own personal earnings or from their parents’ income.

Any grocery shopper living in Turkey, regardless of his purchasing behavior, was welcomed

to take part in the study. This helped to have a wide range of respondents from different

demographic background and with different perspectives on PLBs.

The study used two samples. The first 80% (86 people) of the total sample (108

people) was used to test the proposed model and the remaining 20% (22) was used to validate

the findings. Using a second independent sample to test and support the conceptual model

was consistent with the previous research (e.g., Walsh & Mitchell, 2010). The demographic

distributions in the two partitions were in line with that of the whole data set. This ensured

that the model was not invalidated for foreseeable biases and that the test was fair. Table 1

below provides a description of the sample characteristics.

Table 1

Sample Characteristics

Socio-demographic characteristic Sample 1 Sample 2

Age

Gender

Monthly income

Total

<25

25-30

31.35

36-40

41-45

>45

Male

Female

<1,000 TL

1,000 TL– 3,000 TL

3,000 TL – 5,000 TL

5,000 TL – 7,000 TL

7,000 TL – 9,000 TL

>9,000 TL

8 (9%)

39 (45%)

19 (22%)

10 (27%)

2 (2%)

8 (9%)

54 (63%)

32 (37%)

13 (15%)

43 (50%)

15 (17%)

6 (7%)

0

9 (11%)

86 (80%)

1 (4%)

8 (36%)

5 (23%)

5 (23%)

1 (4%)

2 (9%)

15 (68%)

7 (32%)

3 (14%)

12 (55%)

5 (23%)

1 (4%)

0

1 (4%)

22 (20%)

Note. Reported are nominal and proportional results for each subsample.

Copyright UCT


20

Data Analysis Methods

The data for this research was collected from middle and upper middle-class Turkish

grocerry shoppers currently living in Turkey. Turkey was chosen firstly because it is where

the PL market is growing the fastest (AC Nielsen, 2005) and secondly to help address the

lack of insight on the subject in Turkey. Furthermore, there is also a lack of non-US studies

in the field (Walsh & Mitchell, 2005). It is therefore important to test if the already proven

results are applicable across cultures and in different economies.

In order ensure high level of accuracy, a Turkish version of the questionnaire was

prepared. The terminology used in the survey was adapted for Turkish grocery shoppers to

prevent any kind of misunderstanding and to achieve a uniformity of meaning (Nasif, Al-

Daeaj, Ebrahimi, & Thibodeaux, 1991). No back translation testing was performed on the

survey, that would have helped to identify the possible errors, which may have occured

during the translation process. However, since the author, who prepared the research survey,

is the native speaker of the target language, level of error is assumed to be at a minimum, if at

all.

Microsoft Excel was used to construct a table summarizing the average responses in

each product category, and pie diagrams illustrated the demographic features of the total

sample. This allowed a concise way of providing a feel for the whole data set.

SPSS was employed to carry out several correlation analyses and regressions involving

the explanatory variables. The resulting correlation and regression coefficients were

delivered with their p-values indicating their respective statistical significances. The

statistical analyses and inferences were based on the assumption that the involved regression

errors were normally distributed.

A general-to-specific selection criterion was used to come up with a linear model for

predicting PLB proneness. At each step, the explanatory variable with the highest p-value

was dropped and a new regression was conducted with the remaining variables. The

reduction process continued until no regression coefficient had a p-value greater than .05.

Meanwhile changes in R2 were observed to take into consideration the trade-off between

accuracy and complexity of the candidate model.

The final model was validated on the part of the sample that was left out of the

calibration process. This was done by employing the conventional method of comparing the

root mean squared error of validation to the standard error of the estimate.

Copyright UCT


21

Research Findings, Analysis and Discussion

Research Findings

Demographics

The survey obtained 108 responses in total and displayed rich demographics (gender,

age and income). 66% of the people participated in the study were female and 34% were

males (see Figure 2). The age profile reflected in Figure 3 below shows that 44% of the

respondents falls in the 25-30 age bracket, 22% in the 31-35 age bracket, 14% in the 36-40

bracket, 9% in the above 45, 8% in the below 25 and 3% fall within the 41-45 age bracket.

Therefore, the results mainly represent the purchasing behavior of consumers between the

ages of 25 and 35.

Figure 2. Gender distribution of participants.

Fifty-one percent of the respondents earn between 1,000 and 3,000 TL1 (Turkish Lira)

a month. Therefore, it can be assumed that the survey results reflect majorly the PLB

perceptions of middle and upper middle-class people. It should be noted that no significant

correlation seen to exist between age and income level. Moreover, one’s income level was

not found to be in correlation with either his price consciousness or private label purchasing.

These findings are possibly culture specific, and they hint at the importance of non-economic

(e.g. psychological) factors underlying the studied consumption behavior. As opposed to

some previous research results (e.g., Kilian, Walsh, & Buxel, 2008), males were observed to

be slightly more prone to buy PLBs than females. (More specifically, the average level of

intention to buy PLBs was 2.12 for males and 1.9 for females).

1 Equivalent of €500 and €1,500

64%

36%

Female

Male

Copyright UCT


22

Figure 3. Age distribution of participants.

Figure 4. Income (monthly) distribution of participants.

IBPLB was regressed onto the three demographics to test if gender, age and / or income

had determining roles in consumers’ propensity to purchase PLBs. Table 2 below shows the

results of the regression analysis. Non of the demographic have a statistically significant

relationship with IBPLB. That means age, gender and income do not necessarily determine

one’s intention to buy private label brands. In terms of gender, male consumers (μ = 2.12)

were seen to be slightly more likely to purchase PLBs compared to females (μ = 1,9).

8%

44%

22%

14%

3% 9%

<25

25-30

31-35

36-40

41-45

>45

15%

51%

19%

6%

0%9%

<1,000 TL

1,000 - 3,000 TL

3,000 - 5,000 TL

5,000 - 7,000 TL

7,000 - 9,000 TL

>9,000 TL

Copyright UCT


23

Table 2

Demographics

Model Unstandardized Coefficients Standardized

Coefficients

t

Significance B Standard Error Beta

(Constant) Gender Age Income

1.820 -.006 .054 .050

.417

.186

.115

.077

-.004 .052 .072

4.370 -.033 .467 .641

.000

.974

.642

.524

Cross-category findings

The results obtained from the survey showed that the four variables (determining the

level of perceived risk) differ across the product categories. According to the descriptive

statistics shown in Table 3, it is possible to separate the product categories into two groups as

per their significantly different risk profiles. The underlined product categories (dry pulses,

toilet paper, detergent and soap) in Table 3, fall in the “low-risk profile” group whereas the

other five product categories (cheese, oil, tea, yoghurt and shampoo) belong to the “high-risk

profile” group. Low-risk product categories vary less in terms of quality and have high

search attributes. They also have lower consequences for a purchasing mistake, and therefore

allow the consumer to be more price-conscious. On the other hand, high-risk product

categories possess higher quality variation among different brands, and therefore have higher

consequences for a purchasing mistake. They also have high experience attributes and can

accommodate more price consciousness. Hence, it was found that the respondents were more

likely to buy PLBs in categories where there is low-risk involved.

Yoghurt and cheese were perceived to have the highest purchasing risk whereas dry

pulses were perceived to have the lowest. Cheese was judged to have the highest quality

variation across brands whereas dry pulses were again judged to have the lowest. Similarly,

respondents were prone to choose the brand with the lowest price when buying dry pulses,

and were less likely to exhibit the same behavior while buying cheese. Lastly, search /

experience attributes of a product was seen to be almost equally important for each of the

product categories. Detergent and soap were rated highest on search qualities while yoghurt

and cheese were rated lowest. (In other words, yoghurt and cheese were judged to highest

experience qualities).

Overall, PLB purchasing behavior was observed to display a relation with each of the

Copyright UCT


24

four determinants of the perceived risk factors, across categories. It increases with high price

consciousness and higher search attributes, and decreases with high variation in quality and

higher consequences of purchasing mistake. The mean results were found to be significantly

high, in all categories, for quality variation and for consequences of making a purchasing

mistake.

Table 3

Descriptive Statistics

Note. Reported mean values (standard deviations) are calculated from the total sample of 108

participants, for each product category. Higher value in the Consequences of Purchasing

Mistake column means higher consequences; higher value in Quality Variation means higher

variability and higher value in Search / Experience attributes means higher search

characteristics.

Research Analysis and Discussion

Findings were analyzed, firstly to see how the four selected variables determine one’s

intention to buy private label brands (IBPLB), and secondly to identify the correlations

between each of the dependent and independent variables.

Since all four variables were predicted to have a determining effect on consumers’ PLB

purchasing behavior, a multiple regression analysis was run on SPSS and IBPLB was

regressed onto the four independent variables. (One item was used for measuring price

consciousness, three items for the effects of search versus experience characteristics of the

product category, two items for the affects of quality variation within the product category,

Product Category PC QV CPM SE IBPLB

Dry Pulses

Toilet Paper

Cheese

Detergent

Oil

Tea

Yoghurt

Soap

Shampoo

2.27 (0.99)

2.24 (1.01)

1.73 (0.79)

2.22 (1.04)

1.98 (0.92)

1.94 (0.85)

1.75 (0.71)

2.20 (1.02)

1.76 (0.75)

3.28 (1.12)

3.88 (0.93)

4.32 (0.82)

3.58 (1.02)

4.03 (0.94)

4.01 (0.97)

4.12 (0.91)

3.55 (1.11)

4.07 (0.97)

3.26 (1.18)

3.43 (1.20)

4.02 (0.98)

3.41 (1.10)

3.81 (1.01)

3.84 (1.05)

4.02 (1.03)

3.41 (1.18)

3.87 (1.14)

2.90 (1.12)

2.86 (1.15)

2.84 (1.07)

2.99 (1.13)

2.93 (1.11)

2.86 (1.09)

2.82 (1.07)

2.97 (1.10)

2.88 (1.07)

2.58 (1.15)

2.29 (1.12)

1.63 (0.98)

2.25 (1.06)

1.73 (0.93)

1.78 (0.95)

1.61 (0.92)

2.38 (1.04)

1.73 (1.00)

Copyright UCT


25

and three items for measuring the affects of making a purchasing mistake.) The analysis was

conducted on sample 1 (first 80% of the total data). The results showed a coefficient of

determination (R²) of .259 and beta values between -.406 and .108 (see Table 4).

Table 4

Multiple Regression Analysis 1


Coefficients

T


(Constant) PC SE CPM QV

3.978 .092 .154 -.537 -.154

.758

.116

.139

.146

.137

.082 .108 -.406 -.124

5.248 .791 .106

-3.687 -1.128

.000

.431

.272

.000

.263

Then the unstandardized coefficients and p-values of each coefficient were examined.

As evident in the above table, CPM (consumer purchasing mistake) is the only explanatory

variable that has statistically significant relations with IBPLB. The effects of PC (price

consciousness) are weakest (β = .08, t (degrees of freedom) = .791, p >.431). The other two

variables (QV and SE) also have weak significance due to their high p-values, which were

respectively .263 and .272. The results indicate a strong negative relationship between PLB

purchases and CPM, meaning that as CPM increases, PLB purchases decrease substantially (-

.537). The underlying reason for the weak significances of PC (price consciousness), SE

(search / experience attributes) and QV (quality variation) may be due to a hidden co-

linearity. The issue of co-linearity will be observed later in this research report.

According to the tradition of general-to-specific modeling philosophy, PC is the first

variable that needs to be dropped as one proceeds to form a better model with higher

prediction power. After eliminating PC from the analysis, IBPLB was regressed on to the

remaining three variables (CPM, QV and SE). According to the regression results (see Table

5), CPM (p ~ 0) again was seen to be the only statistically significant explanatory variable

with SE having the weakest significance (β = .117, t (degrees of freedom) = 1.214, p >.228).

Thereafter, IBPLB was regressed onto CPM and QV (see Table 6). The results were the same

as in the previous analysis, leaving CPM (p ~ 0) as the only significant variable influencing

the PLB purchasing behavior. Finally, CPM was regressed on to IBPLB (see Table 7) and

CPM was found to have a significant effect on the PLB purchasing behavior with a

coefficient of -.625. Put in qualitative terms, as the consequences of purchasing mistake

Copyright UCT


26

increases, PLB purchasing gets reduced.

Table 5



Coefficients

t


(Constant) QV CPM SE

4.300 -.178 -.559 .167

.638

.133

.143

.138

-.143 -.423 .117

6.740 -1.339 -3.920 1.214

,000 .184 .000 .228

Table 6



Coefficients

t


(Constant) QV CPM

4.709 -.184 -.537

.543

.133

.142

-.147 -.406

8.664 -1.375 -3.783

,000 .173 .000

Table 7



Coefficients

t


(Constant) CPM

4.328 -.625

.470

.127

-.472 9.204 -4.906

,000 .000

This final regression was taken as the conceptual model used for predicting PLB

purchasing behavior in the remaining 20% of the total sample. Although R2 was only .223,

as indicated in Table 8, the model had good predictive power over sample 2. Root mean

squared error of validation (RMSE) turned out to be .71, which was only slightly above the

standard error of the estimate .69. RMSE was expected to be greater than the standard error

of the estimate, since the model was tuned with respect to sample 1. However, the smallness

Copyright UCT


27

of the difference between the two quantities indicates that the model was tuned well and was

validated.

Table 8


Model R R Square Adjusted R Square Standard Error of Estimate

1 .472a

.223 .213 .6909807

Correlation Analysis

Previous research found the four independent variables (price consciousness, quality

variation, consequences of making a purchasing mistake and search / experience attributes of

a product) to have direct or indirect relationships with one’s intention to buy private label

brands (Batra & Sinha, 2000). In their proposed model Batra and Sinha showed that PC and

CPM had direct effect on PLB purchasing while SE and QV had indirect relationship with

PLB. They explained the non-significant influence of CPM in the base model partly by the

high correlation between consequences of making a purchasing mistake and quality variation.

This correlation is understandable especially since quality variation in a given product

category increases the uncertainty, which makes it more likely to commit a purchasing

mistake. In the current research CPM is seen to be the only independent variable to show a

statistically significant relationship with IBPLB.

There are several general ways of detecting a multicollinearity problem and they

include the following: (1) an independent variable known to be important has a

nonsignificant regression coefficient (2) a regression coefficient that should have a positive

sign ends up having a negative sign, or vice versa, (3) when an independent variable is added

or removed, there is a drastic change in the values of the remaining coefficients (Lind,

Marchal, & Wathen, 2010). Additionally variance inflation factor (VIF) and tolerance values

can be examined. Table highest VIF is .968 and the lowest tolerance is 1.033, and they do

not indicate multicollinearity to be an issue (see Table 9).

Copyright UCT


28

Table 9

Testing for multicollinearity (VIF and tolerance values)

Independent Variables VIF Tolerance

PC

QV

SE

CPM

.855

.757

.968

.754

1.170

1.320

1.033

1.325

However, it is advised that these values should only be taken as a warning sign, and

correlation between independent variables should be checked (Pallant, 2009). In order to

understand the potential co-varying of the four variables among themselves, correlation

analyses were conducted (see Table 10). The findings show no significant relationship

between the search / experience attributes of a product category and any of the other

variables. Correlations between PC and the other two variables (CPM and QV) are

significant, yet weak (-.295 and -.324). On the other hand CPM and QV tend to have a

moderate positive correlation that is significant at the 0.01 level (.450).

Table 10

Correlation Analysis

Variables PC CPM QV SE

PC

CPM

QV

SE

1

-.295**

-.324**

.080

-.295**

1

.450**

.130

-.324**

.450**

1

.030

.080

.130

.030

1

Note. **Correlation is significant at the 0.01 level.

Furthermore, the correlation between CPM and QV can also be observed through a new

regression analysis where IBPLB is regressed onto only the three independent variables (QV,

PC and SE). The results show that both the QV coefficient (from -.154 to -.351) and its

significance (from .263 to .011) changed considerably. Therefore, it can be concluded that it

is the inclusion of CPM, which renders QV as an insignificant explanatory factor in the most

general model. Note that the above mentioned correlations were expected since coefficient

changes were observed at each step of the general-to-specific modeling (see Table 11). At

step 4, when we removed QV from the regression, the coefficient of CPM increased

Copyright UCT


29

significantly (from -.537 to -.625).

Table 11

Coefficient variations

Constant PC SE QV CPM

3.978

4.300

4.709

4.328

.092

-

-

-

.154

.167

-

-

-.154

-.178

-.184

-

-.537

-.559

-.537

-.625

Hypothesis Testing

The research questions were answered in light of the results gained from the above

regression analyses. The four questions sought clarification of the expected effect of the four

independent variables on consumers’ intention to buy PLBs. While the regression

coefficients were used to explain the relationships between variables, the p-values were

observed to test their statistical significances.

Hypothesis 1 theorized that price conscious consumers are more prone to buying

PLBs. The p-value was higher than .05, which meant that price consciousness is not a

determining factor for consumers’ IBPLB. Its nonsignificance is understandable since the

research sample consists of middle and upper-middle class people who are les likely to be

price-conscious consumers. Given the demographics and the results of the regression

analyses the sample is rather brand conscious. Brand conscious consumers tend to prefer any

national brand to private label brands. Therefore it is still possible for a price sensitive yet

brand conscious consumer to purchase the national branded product with lowest price rather

than purchasing a private label brand.

Table 12

Multiple Regression Results


Coefficients

t


(Constant) PC

3.978 .092

.758

.116

.082 5.248 .791

.000

.431

Note. IBPLB regressed over PC, CPM, QV, and SE.

Copyright UCT


30

Previous research (Ailawadi et al, 2001; Rothe & Lamont, 1973) found PLB products

to be preferred by price conscious people. However, it should be noted that, as explained

under demographic findings section, there was no correlation found between income level

and either IBPLB or price consciousness. Therefore, unlike the findings of some previous

research (Lumpkin, Hawes & Darden, 1986; Richardson, Jain, & Dick, 1996) demographics

such as income, age and gender were not seen to be in direct correlation with consumers’

IBPLB.

Table 13



Coefficients

t


(Constant) CPM

3.978 -.537

.758

.146

-.406 5.248 -3.687

.000

.000 Note. IBPLB regressed over PC, CPM, QV and SE.

Hypothesis 2 theorized that consumers are more prone to buy PLBs in product

categories where they believe to have lower consequences of making a purchasing mistake.

This hypothesis was tested looking at the coefficient and p-value for CPM, in Table 11. The

null hypothesis was rejected and the alternative hypothesis was accepted since the p-value

was less than .05. There was a negative relationship between the two variables (-.537). This

finding was consistent with the previous research (Batra & Sinha, 2000; Narasimhan &

Wilcox, 1998).

Hypothesis 3 suspected that consumers are more prone to buying PLBs in product

categories where they perceive lower variability in quality levels across brands. The p-

value was higher than .05 and the null hypothesis could not be rejected (see Table 14). Its

nonsignificance was partly caused by moderate positive correlation between consequences of

purchase mistake and quality variation (see Table 8). Additionally, CPM was regressed onto

QV, PC and SE. The results revealed a positive relationship between QV and CPM (β =

.365, t (degrees of freedom) = 3.822, p ~ 0). That is to say the smaller the range of quality

across brands, the more reduced the level of perceived risk of making a purchasing mistake.

Previous research findings (Batra & Sinha, 2000; Hoch & Banerji, 1993; Narasimhan &

Wilcox, 1998) confirm that the degree of perceived risk increases with higher degrees of

Copyright UCT


31

perceived quality variation across brands in a given category. Hoch and Banerji found a

negative relationship between PLB share and quality variability of a certain product category.

Table 14



Coefficients

T


(Constant) QV

3.978 -.154

.758

.137

-.124 5.248 -1.128

.000


Hypothesis 4 theorized that consumers are more prone to buying PLBs in product

categories where they can easily assess the product attributes or benefits based on the given

information on the package alone. This hypothesis was tested examining the relevant

numbers given in Table 15. Since the p-value was found to be higher than .05, the null

hypothesis could not be rejected. The coefficient value suggested a positive relationship

between a product category’s search attributes and consumers’ IBPLB. However, the

relationship was not seen to be statistically significant. Therefore, it was concluded that

search characteristics of a product category, meaning an informative packaging that allows an

easy assessment of the product attributes, does not determine consumers’ IBPLB. This

conclusion, however, should not be interpreted as packaging having no effect on the

consumers’ purchasing behavior. Table 3 reveals the interesting fact that there is almost no

variation among the average values of product categories, meaning that packaging is

considered to be almost equally important for all categories included in this research.

These findings did not support the results of the previous research. Search / experience

attributes of a product category were previously examined by Erdem and Swait (1998) and

Batra and Sinha (2000). There, consumers were found to be more likely to purchase NBs in

product categories with greater experience attributes. On the other hand, in cases where it

was possible to make an adequate brand comparison based on the given information given by

packaging (e.g. color, ingredients, calories), it was easier for consumers to choose PLBs.

Copyright UCT


32

Table 15


Model Un-standardized Coefficients Standardized

Coefficients

T


(Constant) SE

3.978 .154

.758

.139

.108 5.248 .106

.000


Finally, the discrepancy between these findings and the existing literature can be

explained as follows. Packaging attributes may still be used as means for judging the product

at the time of purchasing. The lack of relationship between SE and IBPLB may be due the

possibility that, in situations where the consumer needs to make a comparison between

various brands, the PL product is not even considered as an alternative due to its high

perceived risk profile. Hoch and Ha (1986) showed that consumers are more likely to base

their judgments on brand name when faced with ambiguous attributes. Consequently, it can

be argued that, because PLBs are ambiguous in their nature compared to NBs, even in

product categories with high search characteristics consumers do not necessarily choose

PLBs. In such cases, consumers’ IBPLB is more likely to be determined by the level of

perceived risk factors influenced by CPM and QV. The discrepancy may as well be related to

the level of brand consciousness among Turkish consumers. Therefore, it is suggested that

the brand consciousness is included as a variable in future research.

Research Limitations

Sample size could be increased in order to strengthen the conclusions of the statistical

arguments made in the analysis. Greater sample size would have led to a more precise

calibration and a more reliable validation of the model.

Current research analyzed four category-based determinants of level of perceived risk

across nine product categories. The final analysis made it clear that the risks associated with

products differed across categories. However it would be valuable to extend the analysis

further and see how other product categories are effected by risk factors. Additionally,

different independent variables could be chosen to form better model with higher fit.

Moreover, paper questionnaires and face-to-face interviews could be used to reach the

part of the low-income population that does not have Internet access. The survey itself could

Copyright UCT


33

also be improved. For instance, PC and IBLB vectors were constructed using single item

measurements. Two or greater number of items could have been used to increase the

reliability of the data.

Research Conclusions

The effect of the level of perceived risk on consumers’ PLB purchasing behavior has

previously been demonstrated by many researchers (Richardson, Jain & Dick, 1996;

Sethuraman & Cole, 1997, Walsh & Mitchell, 2010). There have been various explanations

as to what determines the level of this risk, and four were selected for examination for the

purposes of this study (PC, CPM, SE, and QV). Moreover, previous research either proved

(Batra & Sinha, 2000; Walsh & Mitchell, 2010) or suspected (Ailawadi, Pauwels &

Steenkamp, 2008) that consumers’ intention to buy PLBs varied across product categories.

This study was aimed at understanding how the four selected variables affected Turkish

consumers’ intentions to buy PLBs across nine product categories.

The results indicated that all four variables together do not act as a determining factor

on consumers’ IBPLB. To be more precise, consumer’s PLB purchasing could not be

modeled using all four variables together in a statistically robust sense. CPM was found to be

the only significantly effective construct (-.537, p = 0). This indicates that consumers’

IBPLB increases as the perceived consequences of making a purchasing mistake in a given

product category diminish. This relationship was expected in the sense that, as hypothesized

in H2, the level of perceived risk associated with PLB in a product category is likely to

increase with greater levels of CPM. Consequences of choosing the wrong brand may be

perceived to be high, (1) if the outcome affects the person directly in terms of self-image,

health, beauty or physical condition (Blackwell, Miniard, & Engel, 2005), (2) if the consumer

is eager for social approval. Kumar and Steenkamp (2007) refer to the consumers in the

latter group as “emotional / social benefit seekers”. Because PLBs are often associated with

low quality, the products that are meant to be consumed in the presence of peers tend to hold

high social risks (Zielke & Dobbelstein, 2007). Mean values revealed CPM to be higher for

shampoo, oil, tea, yoghurt and cheese. These product categories may not necessarily have

any social risk attached to them. However, they were precisely the product categories that

consumers perceived to possess the greatest quality variation among brands.

The other three independent variables (QV, SE and PC) were not found to have any

direct influence on IBPLB. Further analysis indicated the presence of a moderate positive

Copyright UCT


34

relationship between QV and CPM. The CPM effect is suspected to mediate the effects of

QV on IBPLB. Accordingly, CPM increases with higher perceived quality variations among

the brands of a given product category (.450, p = .01). The co-varying relationship was

consistent with the findings of Batra and Sinha (2000). Their results explained the non-

significance of the QV effect by pointing out its high correlation with the consequences of

purchasing mistake.

As opposed to the conclusions of previous research, and unlike it was hypothesized in

H4, an informative packaging on a product’s attributes / benefits does not explain consumers’

IBPLB. However, this lack of relationship should lead one to interpret packaging as an

inefficient or insignificant factor in influencing consumers’ purchasing decisions. The result

only points out a correlation between the two variables. Furthermore, when looked at

separately, no significant relationship was found between SE and the other three variables.

Previous research argues that, when faced with ambiguity, consumers tend to base their

decisions on extrinsic characteristics, such as a brand name (Hoch & Ha, 1986). This

argument is plausible in the sense that ambiguity may be expected to cause an increase in

CPM. A situation, which forces the consumer to refer to the packing information to make a

brand decision, is likely to leave the consumer in ambiguity. Following the logic of Hoch

and Ha, in such a case IBPLB will be reduced and consumers will narrow down their options

to NBs only. Therefore, brand consciousness is suspected to play a covert role in the

ineffectiveness of SE in the prediction of any of the rest of the variables, including IBPLB.

However, this potential role was not investigated in this study.

Previous research revealed conflicting results about the influence of demographics as

explanatory factors of consumers’ intention to purchase private label brands. Some proved

that gender and income played determining role on IBPLB (Richardson, Jain & Dick, 1996;

Sethuraman & Cole, 1997), whereas others found that gender and age did not add to the

explanation of PLB purchasing behavior. The current research controlled to understand the

possible effects of gender, age and / or income but couldn’t find any. Neither of the three

demographics do not have any statistically significant relationship with IBPLB. In terms of

gender, male consumers (μ = 2.12) were seen to be slightly more likely to purchase PLBs

compared to females (μ = 1,9).

Finally, as evident in the mean values obtained from the survey results, the perceived

risk associated with each variable was confirmed to be changing across product categories.

Compared to the ranges in other variables, the variation in search / experience attributes

Copyright UCT


35

construct was quite limited. These differences among the mean values made it possible to

group the products with respect to their risk profiles.

Managerial Implications

The study was conducted to shed some light on private label brands from the

consumers’ perspective. Moreover, it was hoped to generate some valuable insights for

retailers and manufacturers. This seemed to be especially important given the lack of

academic research on PLB in Turkey.

The results showed that perceptions on PLBs differ across product categories and vary

along with the respective levels of perceived risk. Manufacturers and retailers may be able to

use this information in order to devise a strategy for the development and rearrangement of

their product portfolios.

Furthermore, findings reveal customer purchasing mistake and quality variation to be

highly correlated with consumers’ IBPLB. This implies that retailers should seek clear and

effective ways of signaling the high quality of their products and thereby preventing the onset

of potential uncertainties. For instance, they could make investments to increase brand

awareness, and engage in strong marketing activities to enhance brand loyalty.

Future Research Directions

Considering the findings and the limitations of this research, several adaptations could

be recommended to enhance the significance and reliability of future research.

This study examined the level of perceived risk across nine different product categories.

The variation among these categories was proved to be significant. Future studies could

either analyze the relationship among other product categories with similar characteristics, or

extend the current product portfolio.

Four different consumer-level constructs were looked at as the determining factors of

IBPLB. CPM was observed to have a direct relationship with IBPLB. Furthermore, possible

indirect influence of QV through CPM was shown. Future research could look into different

factors behind the mechanisms that underlie risk perception (e.g., brand consciousness) and

find better fitting models.

In addition, it is suggested that the sample size should be increased in order to assure

the greater accuracy of the findings. Furthermore, the survey could also be conducted via

face-to-face interviews, as was commonly done in previous research. This would also allow

one to have greater control over the demographics. Although, this research did not find

income level to be an influencing factor on either one’s price consciousness or his intention

Copyright UCT


36

to buy private label brands, low-income consumers and their PLB purchasing behavior are

worth investigating.

Copyright UCT


37

References

Aaker, D. A., Kumar, V., & George D. S. (2004). Marketing Research. USA: John Wiley &

Sons, Inc.

AC Nielsen, Private Label. (2005). A good alternative to other brands, offering the same

quality and value. ACNielsen Global Consumer Survey, http://sg-

ac.nielsen.com/news/20050822.shtml (Accessed 30 August 2010).

AC Nielsen (2008), Higher unit prices, not volume, behind rapid growth of U.S. private label

sales, http://en-us.nielsen.com/main/news/news_releases/2008(june/nielsen_higher_unit

(Accessed 30 August 2010).

Ailawadi, K. L., & Harlam, B. (2004). An empirical analysis of the determinants of retail

margins: the role of store-brand share. Journal of Marketing, 68, 147-165.

Ailawadi, K. L., Pauwels, K., & Steenkamp J.-B. E. M. (2008). Private label use and store

loyalty. Journal of Marketing, 72, 19-30.

Ailawadi, K. L., Neslin, S., & Gedenk, K. (2001). Pursuing the value conscious consumer:

store brands versus national brand promotions. Journal of Marketing, 65, 71-89.

Amrouche, N., & Zaccour, G. (2007). Shelf-space allocation of national and private brands.

European Journal of Operational Research, 180, 648-663.

Batra, R. & Sinha, I. (2000). Consumer-level factors moderating the success of private label

brands. Journal of Retailing, 76, 175-189.

Belizzi, A. J., Hamilton R. J., Krueckeberg H. F., & Martin, W. S. (1981). Consumer

perceptions of national, private, and generic brands. Journal of Retailing, 57, 56-69.

Bergvist, L., & Rossiter, R. J. (2007). The predictive validity of multiple-item versus single-

item measures of the same constructs. Journal of Marketing Research, 44, 175-184.

Bettman, J. R. (1974). Relationship of information-processing attitude structures to private

brand purchasing behavior. Journal of Applied Psychology, 59, 79-83.

Blackwell, R. D., Miniard, P. W., & Engel, J. F. (2006). Consumer Behavior. Canada:

Thomson South-Western.

http://sg-ac.nielsen.com/news/20050822.shtml

http://sg-ac.nielsen.com/news/20050822.shtml

http://en-us.nielsen.com/main/news/news_releases/2008(june/nielsen_higher_unit

Copyright UCT


38

Boyle, P. J., & Lathrop, S. E. (2009). Are consumers’ perceptions of price-quality

relationships well-calibrates? International Journal of Consumer Studies, 33, 58-63. doi:

10.1111/j.1470-6431.2008.0072.x

Bryman, A., & Bell, E. (2007). Business Research Methods (2nd ed.). Oxford: Oxford

University Press.

Burton, S., Lichtenstein, R. D., Netemeyer, G. R., & Garretson, J, A. (1998). A scale for

measuring attitude toward private label products and an examination of its psychological

and behavioral correlates. Journal of the Academy of Marketing Science, 26, 293-306.

Bryman, A. & Bell. E. (2008). Business Research Methods. Oxford: Oxford University Press

Corstjens, M., & Lal, R. (2000). Building store loyalty through store brands. Journal of

Marketing Research, 37, 281-291.

Cotterill, J., Ronals, W., Putsis, W. P., Dhar, R. (2000). Assessing the competitive interaction

between private labels and national brands. Journal of Business, 73, 109-137.

De Wulf, K., Odekerken-Schroder, G., Goedertier, F., & Van Ossel, G. (2005). Consumer

perceptions of store brands versus national brands. Journal of Consumer Marketing, 22,

223-232.

Deoitte Debates (2010), Brand loyalty and the impact of private label products,

http://www.deloitte.com/us/brandloyalty (Accessed 30 August 2010).

Dhar, S. K., & Hoch, S. J. (1997). Why store brand penetration varies by retailer. Marketing

Science, 9, 208-227.

Private label as a solution against recession. (2008, December 14). Ekonomist, 50th issue.

Erdem, T., & Swait, J. (1998). Brand equity as a signaling phenomenon. Journal of

Consumer Psycohology, 7, 131-157.

Herstein, R., & Gamliel, E. (2006). The role of private branding in improving service quality.

Managing Service Quality, 16, 306-319.

Hoch, S. J. (1996). How should national brands think about private labels? Sloan

Management Review, 37, 89-102.

http://www.deloitte.com/us/brandloyalty

Copyright UCT


39

Hoch, S. J., & Banerji, S. (1993). When do private labels succeed? Sloan Management

Review, 34, 57-67.

Hoch, S. J., & Ha, Y. W. (1986). Consumer learning: advertising and the ambiguity of

product experience. Journal of Consumer Research, 13, 221–233.

Kilian, T., Walsh, G., & Buxel, H. (2008). Measurement of attitude towards private labels: a

replication and extension. European Retail Research, 22, 69-85.

Kumar, N., & Steenkamp, J.-B. E. M. (2007) Private label strategy: how to meet the store

brand challenge. USA: Harvard Business School Press.

Leedy, D. P., & Ormrod, J. E. (2010). Practical Research: Planning and Design. Upper

Saddle River, NJ: Prentice-Hall.

Lind, A. D., Marchal, G. W., & Wathen, A. S. (2010). Statistical Techniques in Business &

Economics. New York: McGraw-Hill.

Livesey, F., & Lennon, F. (1978). Factors affecting consumers’ choice between manufacturer

brands and retailer own brands. European Journal of Marketing, 12, 158-170.

Mills, D. E. (1995). Why retailers sell private labels. Journal of Economic Management

Strategy, 4, 509-528.

Narasimhan, C., & Wilcox, R. T. (1998). Private labels and the channel relationship: a cross-

category analysis. Journal of Business, 71, 573-600.

Nasif, E. G., Al-Daeaj, H., Ebrahimi, B., & Thibodeaux, M. S. (1991). Methodological

problems in cross-cultural research: an updated review. Management International

Review, 31, 79-91.

Newman, D. & Becknell, J. (1970). The Price Quality Relationship as a Tool in Consumer

Research. American Psychological Association, 5, 729–730.

Pallant, J. (2005). SPSS Survival Manal. Glasgow: Bell & Bain Ltd.

Pauwels, K., & Srinivasan, S. (2004). Who benefits from store brand entry? Marketing

Science, 23, 364-390.

Copyright UCT


40

Private Label Manufacturers Association. (2010). Retrieved August 27, 2010, from, http://sg-

acnielsen.com/news/20050822.shtml.

Private Label Turkey. (2007). Retrieved August 30, 2010 from,

http://www.plturkey.org/yazilar.asp?Konu=PrivateLabel

Putsis, W. P. Jr., & Dhar, R. (2001). An empirical analysis of the determinants of category

expenditure. Journal of Business Research, 52, 277-291.

Richardson, P. S., Jain A. K., & Dick, A. (1994). Extrinsic and intrinsic cue effects on

perceptions of store brand quality. Journal of Marketing, 58, 28-36.

Richardson, P. S., Jain A. K., & Dick, A. (1996). Household store brand proneness: a

framework. Journal of Retailing, 72, 159-185.

Rothe, J. T., & Lamont, L. M. (1973). Purchase behavior and brand choice determinants.

Journal of Retailing, 49, 19-33.

Sethuraman, R. (1992). Understanding cross-category differences in private label shares of

grocery products. Marketing Science Institute Working Paper, 92-128.

Sethuraman, R., & Cole, C. (1999). Factors influencing the price premium that consumers

pay for national brands over store brands. Journal of Product and Brand Management, 8,

340-351.

Shannon, R., & Mandhachitara, R. (2005). Private-label grocery shopping attitudes and

bahevior: a cross-cultural study. Brand Management, 12, 361-474.

Sheath, K, J., & McGoldrick, P. J. (1981). Generics: their development in grocery retailing

and the reactions of consumers. Manchester: UMIST.

Walsh, G., & Mitchell, V-W. (2010). Consumers’ intention to buy private label brands

revisited. Journal of General Management, 35, 3-24.

Zielke, S., & Dobbelstein, T. (2007). Customers willingness to purchase new store brands.

Journal of Product and Brand Management, 16, 112-121.

http://sg-acnielsen.com/news/20050822.shtml

http://sg-acnielsen.com/news/20050822.shtml

http://www.plturkey.org/yazilar.asp?Konu=PrivateLabel

Copyright UCT


41

Appendix: Scale of Research

1- Strongly disagree 5-Strongly agree

Q1. Gender

1- Male

2- Female

Q2. Age

1- <25

2- 25-30

3- 31-35

4- 36-40

5- 41-45

6- >45

Q3. Monthly income

1- <1,000 TL

2- 1,000 – 3,000 TL

3- 3,000 – 5,000 TL

4- 5,000 – 7,000 TL

5- 7,000 – 9,000 TL

6- >9,000 TL

Q4. When I choose a brand of this product category, it is not a big deal if I buy a wrong brand

1 2 3 4 5

Dry pulses

Toilet paper

Cheese

Detergent

Oil

Tea

Yoghurt

Soap

Shampoo

Copyright UCT


42

Q5. All brands of this are basically the same in quality

Q6. Brands of this category do not vary a lot in terms of their quality

Q7. When I buy a product of this category, it is ok if the brand is not the best quality

1 2 3 4 5

Dry pulses

Toilet paper

Cheese

Detergent

Oil

Tea

Yoghurt

Soap

Shampoo

1 2 3 4 5

Dry pulses

Toilet paper

Cheese

Detergent

Oil

Tea

Yoghurt

Soap

Shampoo

1 2 3 4 5

Dry pulses

Toilet paper

Cheese

Detergent

Oil

Tea

Yoghurt

Soap

Shampoo

Copyright UCT


43

Q8. For this product category, an informative package with all features helps me choose a

brand

Q9. Information on the package is enough for me when buying:

Q10. I need to have tried and tested the brand of this category in order to buy it

1 2 3 4 5

Dry pulses

Toilet paper

Cheese

Detergent

Oil

Tea

Yoghurt

Soap

Shampoo

1 2 3 4 5

Dry pulses

Toilet paper

Cheese

Detergent

Oil

Tea

Yoghurt

Soap

Shampoo

1 2 3 4 5

Dry pulses

Toilet paper

Cheese

Detergent

Oil

Tea

Yoghurt

Soap

Shampoo

Copyright UCT


44

Q11. When buying a brand of this product category, I look for the recognized brands

Q12. When buying a brand of this product category I look for the cheapest product available

Q13. When I think about my purchases of this product category during the past years, I have

purchased private label brands

1 2 3 4 5

Dry pulses

Toilet paper

Cheese

Detergent

Oil

Tea

Yoghurt

Soap

Shampoo

1 2 3 4 5

Dry pulses

Toilet paper

Cheese

Detergent

Oil

Tea

Yoghurt

Soap

Shampoo

1 2 3 4 5

Dry pulses

Toilet paper

Cheese

Detergent

Oil

Tea

Yoghurt

Soap

Shampoo

Documents

Consumers and Their Perception of Private Labels in …gsblibrary.uct.ac.za/researchreports/2010/Karan.pdfWhen a retailer sells a product with a brand ... United States and more than