Master Thesis Marketing
The Influence of Introducing a Generic Private
Label on Consumer Choice
Author: André Dietz
Student number 1481444
Master Thesis Marketing
The Influence of Introducing a Generic Private
Label on Consumer Choice
July, 2009
Master Thesis
MSc Business Administration
Marketing Management
Faculty of Economics and Business
University of Groningen, the Netherlands
_____________________________________________________________________
Author: André Dietz
Student number: 1481444
Address: Leeuwarderstraat 4, 9718 HX Groningen
Phone number: 0031 6149 77 465
E-mail: andre_dietz@hotmail.com
University supervisors: Dr. J.A. Voerman
Dr. M.A. Tuk
Company Supervisors: Alexander Hendriks
John Dankers
Company: Albert Heijn
Provincialeweg 11 1506 MA Zaandam
Management Summary
The second most important key motive for consumers to decide their supermarket choice, is valuable. The choice criteria for valuable are low prices, attractive promotions and good offer of low priced products (generics). By having a well balanced assortment of low priced products (generics), full service supermarkets can demonstrate that they have a basket that is price competitive against discounters. However, few research exist of the effect of adding a generic PL to the choice set of a supermarket assortment. Therefore, the purposes of this research were aimed at the effect of adding a generic PL to a current two alternatives choice set (manufacturer brand and copycat PL) resulting in a three alternatives choice set (manufacturer brand, copycat PL and generic PL) on consumer choice and on identifying the factors that influence consumer choice.
An extensive literature study has revealed the role generics play in retail assortment and identified the main factors that are expected to influence consumer choice and/or the relation between introducing a generic PL and consumer choice. These factors are: compromise effect, decision uncertainty, product familiarity, quality perceptions, decision involvement, product risks and brand experience.
To analyze the influence of the above factors, data was obtained by means of an online questionnaire. Albert Heijn customers over 18 years old, who were responsible for the groceries within their household and were reachable by mail were targeted. To allow for comparison on quality perceptions of the copycat PL and consumer choice, two groups of respondents were involved in the research. Group A (n=152) were asked questions concerning the three alternatives choice set and the questionnaire of group B (n=156) contained questions concerning a two alternatives choice set.
After analysing the data, the results of this study show that the expected compromise effect was not applicable. When a third alternative is added to the choice set the market share of the compromise option (i.e. copycat PL) did not increase. Although the market share did not increase of the copycat PL, results in this study also show that quality perceptions of the copycat PL do increase when a generic PL is added to the choice set.
involved are less likely to choose the generic PL. For product groups with high financial or social risk, respondents are less likely to buy the generic PL. For product groups with high financial and functional risk, respondents are more likely to buy the manufacturer brand. And when consumers have high manufacturer brand- or generic PL experience they are more likely to choose the manufacturer, respectively the generic PL in the choice set.
Preface
By writing this master thesis during an internship at Albert Heijn I was writing an end to my master Marketing Management and my life as a student.
I enjoyed being a student in the lovely city Groningen. From the very beginning I was satisfied with my study choice. I have experienced and learned about a diverse spectrum of business knowledge fields (Economics, Marketing, Accounting, HRM, Operations and IT). During my bachelor I learned I like to entrepreneur and be creative on a fact-based manor. The Master Marketing, in Groningen, one of the best in the Netherlands, was a good match with my capacities and needs.
Besides studying, my student life meant much more to me. From year one, I have lived in Groningen. In my student life I have met and get to know a lot of interesting people. Not only during my study but also by being a member of the student Society Dizkartes.
An other important part of my life as a student was starting my own company in 2006, called Promo Bears. We offer a creative way of doing marketing; we create brand characters and bring them a life by making custom made mascots and plush toys. Promo Bears offered me the perfect opportunity to put knowledge in practice and to learn facets, which I believe can only be learned in the “real world”. I would like to take this opportunity to thank some persons who supported me during my research. First of all I would like to thank Alexander Hendriks and John Dankers for their support during my internship. I am very grateful for all the knowledge they have shared with me. Next, I would like to thank Petra Janssens for the interesting and funny conversations during lunch breaks in the beautiful city of Zaandam. Furthermore I would like to thank the entire team of Price Management for their support, they were the reason I enjoyed every day of my internship. Finally a special thanks for Joost Poelgeest a fellow intern at Albert Heijn for his help, table soccer qualities and pleasant coffee breaks during my research.
I specially want to thank my supervisor Liane Voerman for here advise and helping me during the various phases of this thesis. With her advice and feedback she made sure I maintained an academic focus during the research.
Index
1. Introduction & Problem statement... 8
1.1 Background problem... 8
1.2 Problem statement and research questions ... 11
1.3 Theoretical and social relevance ... 12
2. Theoretical framework... 13 2.1 Assortment ... 13 2.1.1 Assortment width... 14 2.1.2 Assortment depth... 15 2.2 Brand architecture... 16 2.2.1 Manufacturer brands ... 16 2.2.2 Private labels ... 17
2.3 Factors influencing consumer choice... 20
2.3.1 Compromise and attractiveness effect... 21
2.3.2 Private labels versus manufacturer brands ... 24
3. Research design ... 27
3.1 The Dutch supermarket industry and company... 27
3.2 Method ... 27
3.2.1 Groups of respondents... 27
3.2.2 Product selection... 28
3.2.3 Measurement scales ... 30
3.3 Population and sampling ... 31
3.4 Reliability, validity and representativeness ... 33
3.4.1 Sample characteristics... 33
3.4.2 Response statistics... 34
3.5 Plan of analysis ... 35
4. Analysis and results ... 39
4.1 Compromise effect ... 39
4.1.1 Decision uncertainty ... 40
4.1.2 Product familiarity... 42
4.2 Private labels versus manufacturer brands... 44
4.2.1 Quality perceptions... 44
4.2.2 Decision involvement... 45
4.2.3 Product risks ... 47
4.2.4 Brand experience ... 49
5. Conclusions and recommendations ... 52
Appendix 1 Questionnaire preliminary research ... 63
Appendix 2 outcome preliminary research product groups ... 65
Appendix 3 questionnaire main research ... 66
Appendix 4 invitation mail/message... 76
Appendix 5 sample characteristics... 77
Appendix 6 outcome independent sample T-tests ... 79
1.
Introduction & Problem statement
1.1
Background problem
In the supermarket industry, price is and likely will remain the predominant basis for cross-formula competition (e.g., Garry 1994; Kahn and McAlister 1997; Mathews 1997). Dutch consumers rank price image “valuable” as the second most important key motive for store choice (see table 1) (EFMI and CBL, 2008).
Table 1.1: consumers key motives for supermarket choice (EFMI, CBL, 2008)
Key motive Choice criteria
1. Quality & choice • Good product quality
• Good offer of fresh products • Large assortment
• Customer friendly staff
2. Valuable • Low prices
• Attractive promotions
• Good offer low priced product (generics)
3. Location • Supermarket is near by
4. Efficiency • Long opening hours
• Speed at the cash out • Parking space
5. Store appearance • Supermarket is looking nice
• Shop is looking neatly
6. Extra’s • Good offer extra services
• Child friendliness supermarket
• Enough shops in the surrounding of the supermarket The key motive “valuable” is influenced by three criteria: low prices (manufacturer brands), attractive promotions and good offer of low priced products. Accordingly, establishing a positive store-price image is a common priority among grocery firms (e.g., Cox and Cox 1990; Dickson and Urbany 1994). The rapid growth of discounters, such as Aldi and Lidl, in the Netherlands during the 1990s has put pressure on traditional full service supermarkets and their price images. Dutch consumers bought their fresh fruit and vegetables at full service supermarkets and the remaining part of their groceries at the much cheaper discounters. The leading Dutch supermarket chain Albert Heijn suffered from an unfavourable and deteriorating price image, and felt like it had to respond.
expensive”, its double-page colour advertisements in all national and local newspapers made clear that the chain was committed to decrease its prices systematically and permanently. First, several price reduction rounds of manufacturer brands where executed. Next on September 20, 2004, Albert Heijn reduced prices of 1000 private labels as well (Van Aalst et al. 2005; Holla and Koreman 2006). Next to the price decreases, Albert Heijn increased the assortment of generics. Generics, also called C-products, are low quality low priced products. Generics can be positioned with private labels and manufacturer brands (fancy labels). The first Dutch supermarkets introduced generics in the 1980s and most full service supermarkets followed in the 1990s.
A plain packaging design, limited advertising activities and cuts in quality yield a positioning in the lowest price tier (Yelkur, 2000; Harris and Strang, 1985). Generics are mostly offered next to a manufacturer brand (first price tier) and/or a copycat private label (PL) (second price tier) as a second or third price tier. Manufacturer brands are branded products from manufacturers and reach consumers through distributors and retailers (Kumar and Steenkamp, 2007). Private labels (PLs) are products owned and distributed by organizations whose primary economic commitment is distribution rather than production (Schutte, 1969). A copycat PL is a type of PL that is very close to manufacture brands in terms of packaging (Kumar and Steenkamp, 2007). The combination of manufacturer brands, private labels and generics is called the brand architecture of an assortment of a formula (Esjberg, et al., 2004).
Although full service supermarkets introduced or increased their generic assortment, they wanted to maintain their service image “quality and choice”. With the generic assortment they wanted to offer good value for money “valuable” as well. The price war changed the Dutch supermarket industry drastically. As Van Heerde et al. (2008) state “The price war initiator (Albert Heijn) managed to halt the slide in its market share, and its stock price improved. The losers were the rival mid-level and high-end chains. Unlike the initiator their price image did not improve, and they suffered from increased price image sensitivity.” Konmar (high-end chain) and Edah (mid-level chain) don’t exist any more. Other supermarkets improved their position, such as the upcoming formula Jumbo, offering high service and low price. Consumers respond very positive on this formula; Jumbo won the summer report of Elsevier for the eleventh time this year.1
The Dutch supermarket price war fits the trend that retail price competition has become increasingly vivid in recent years, reducing retailer profitability (Ailawadi, 2001). Discounters such as Wal-Mart, Aldi and Lidl are challenging traditional retail formats on both sides of the Atlantic (BusinessWeek, 2003). In almost all Western markets, grocery discounters have captured market share from traditional supermarkets and now occupy a prominent position (Cleeren et al., 2007). The reaction of traditional retailers has varied from focusing on quality and service, to engaging the challengers with substantial price reductions (Rogers, 2001). However, these price reductions may trigger price wars, which can last for a long time and strongly affect all market players (Rao et al., 2000). An alternative reaction could be to focus on the assortment composition. As stated above, the offer of low priced products (generics), influences price image as well. By having such a lowest price tier, traditional retailers can demonstrate that they have a basket that is price competitive against discounters.
However, few academic research exist on the effect of introducing a generic product or generic PL product to the brand choice set of a product group assortment. When a generic PL alternative is added to the assortment, this choice set changes. The choice set can change from a one alternative choice set (manufacturer brand or copycat private label)
1 http://www.elsevierretail.nl/1066352/Food/Food-nieuwsbericht/JumboWintWederomZomerrapport.htm The
to a two alternatives choice set (manufacturer brand or copycat PL and generic PL) or from a two alternatives choice set (manufacturer brand and copycat private label) to a three alternatives choice set (manufacturer brand, copycat private label and generic private label). In this study the consumer choice from a two alternatives choice set to a three alternatives choice set will be analysed, see figure 1.2.
Figure 1.2: Consumer choice from a two alternatives choice set to a three alternatives choice set
Two alternatives choice set Three alternatives choice set
Manufacturer brand Manufacturer brand
Copycat private label Copycat private label
Generic PL
What is the effect of introducing a generic PL to the choice set? Will consumer’s buy the new generic PL option or will the market share of the copycat PL or manufacturer brand increase? This research will try to gain more insight if and how the consumer choice changes and which main factors influence this change.
1.2
Problem statement and research questions
Based on the above background problem, the problem statement is as follows:
How does adding a generic private label product to the choice set influence consumer choice and consumer quality perceptions of a copycat private label?
To solve the problem statement the following research questions will be analysed: 1. What role does a generic private label play in retail assortment?
2. What is the influence of introducing a generic private label on consumer’s choice between generic private labels, copycat private labels and manufacturer brands? 3. What are the main factors influencing consumer choice and/or the relation between
introducing a generic PL and consumer choice, and to what extent?
This study consist of two parts: an academic literature study and a quantitative research. In chapter two, “theoretical framework”, an extensive literature study will discuss the research questions based on academic literature. Based on the literature study research hypotheses will be formulated. Next in chapter three, “Research design”, the research design for the empirical research based on a case study at Albert Heijn, a full service Dutch supermarket, will be discussed. In chapter four the analysis and results of the empirical research will be discussed. Finally, in chapter five the conclusions, managerial recommendations, research limitations and future research suggestions will be addressed.
1.3
Theoretical and social relevance
This study will add to current research in three ways. First, the effect of introducing a new generic PL product on consumer choice will be analysed. Second, the main factors influencing consumer choice and/or the relation between introducing a generic PL to the choice set and consumer choice will be identified and analysed on the extent of influence they perform. And third, insight will be gained in to what extent quality perceptions of copycat PLs change when a generic PL is introduced to the choice set.
2.
Theoretical framework
In this chapter, the role of generic PL products in retail assortment will be addressed (research question one). Next, based on literature the expected effect of introducing a generic PL product on consumers’ choice will be discussed (research question two). Furthermore the main factors expected to influence the relation between introducing a generic PL and consumer choice, will be addressed (research question three). When discussing the main factors, the expected effect of introducing a generic PL on quality perceptions of copycat PL will be discussed as well (research question four).
2.1
Assortment
A key function of retailers is to provide ‘an assortment of products and services’ for consumers (Levy and Weitz, 2004). An early marketing definition views this ‘selling assortment’ as the ‘total number of items which can be sold by a firm in given transactions’ (Balderston, 1956). Hollander (1966) interprets assortment at the aggregate retail level of ‘number of lines’ on offer. Simonson (1999) expands the description to include ‘the total set of items offered by a retailer, reflecting both the breadth and depth of offered product lines’. Furthermore a subsequent level, termed the ‘brand assortment’, also called the brand architecture, suggests a subset of assortment whereby the mix of brands varies independently of the number of categories (Savitt, 1984).
So, product lines can be defined either vertically (depth) to capture differential consumer willingness to pay for quality, or horizontally (width) to capture different tastes within the same quality level (Draganska and Jain, 2006), see figure 2.1.
Figure 2.1. Visualisation of assortment width and depth Assortment width
Product lines Price tiers
Coke Orange soft
drink (sinas) Lemon-lime soft drink Raspberry soft drink (cassis) Coca-Cola,
Pepsi Fanta, Sisi, Orangina 7-Up, Sprite Hero, Fanta 1
st price tier
AH coke AH orange soft
drink AH lemon-lime soft drink
AH cassis 2nd price tier
A ss o rt m en t d ep th
The assortment width are the number of product lines (i.e. coke, sinas and cassis), the assortment depth are the number of price tiers, and the brand assortment is the combination of specific brands Manufacturer brands (Coca-Cola, Pepsi, Fanta, Sisi, 7-Up), copycat PLs (AH) and generic PLs (Euro Shopper).
The assortment width, assortment depth and brand assortment (also called brand architecture) will be discussed in more detail in the following paragraphs.
2.1.1 Assortment width
The width of an assortment is used to describe the number or choice of distinct product classes or categories carried by the store (Savitt, 1984). Width is often used synonymously with variety; for example, Risch (1991) defines it as ‘the number of different classifications carried by the store’. Thus, the selection of different categories or departments denotes the variety of the store assortment (Davies and Brooks, 1989). For example the width of the product category soft drinks can be: coke, orange flavoured soft drinks, lemon-lime flavoured soft drinks and other flavours. A wide assortment has various benefits. First, the greater the width of product assortment, the greater the range of different situations in which the retailer is recalled and considered by the consumer, and therefore the stronger its salience (Ailawadi and Keller, 2004). As noted by Keller (2008), salience is the most basic building block for a brand. Second, the one-stop shopping convenience that a wide product assortment enables, is becoming more important than ever for today’s time-constrained consumer (Messinger & Narasimhan 1997). Third, consumers regularly shop at more than one store, they may purchase a category in the store that they are visiting based on in-store assortment and marketing mix activities whereas they would otherwise have purchased it in another store (Ailawadi and Keller, 2004). Together with the fact that unplanned purchases comprise a significant portion of consumers’ total shopping basket, this gives an advantage to retailers with wider assortments (Ailawadi and Keller, 2004). Research from EFMI business school and CBL (Central Food Agency) validate the importance of assortment width. Dutch consumers (1501 respondents) validate large assortment as the third most important choice motive for supermarket choice, see table 2.1.
Table 2.1: Choice motivation indicator (top 10) (EFMI, CBL, 2008) Average importance of consumer choice motives on supermarket choice Top ten choice criteria
2. Good offer of fresh products 8.4 3. Large assortment 8.2 4. Attractive promotions 8.1 5. Low prices 8.1 6. Friendly staff 8.1 7. Supermarket is near by 7.9
8. Speed with the cash desk 7.8
9. Good offer of low priced products (generics) 7.8
10. Neatness of the store 7.8
Choices on assortment width have to be balanced and logic for the customer. As Danneels (2003) states “it is risky to extend too far, but, staying too tightly coupled to the current assortment and image may unnecessarily limit the retailer’s range of experimentation.” Generic products play a minimal role in the width of the total retail assortment. As generics are a copy/substitute of an existing manufacturer or private label brand, they are not adding to the assortment width in general. In the soft drinks example the generics offer a coke, orange soft drink and lemon soft drink substitute, but a cassis substitute is lacking. However, for consumers specifically interested in the lowest price tier, the width of generics can be very important. As a wide assortment can create customer value by offering convenience and ease of shopping.
2.1.2 Assortment depth
2.2
Brand architecture
Esjberg, et al. (2004) define a brand architecture as a combination of manufacturer brands, private labels and generic products in a food retailer’s assortment that results from a strategic process constrained by the retailer’s identity, desired image, policies, structures and technologies. The brand architecture has become particularly important in the last decade as a tool for retailers to influence their image and develop their own brand name. (Ailawadi & Keller, 2004) The relative importance of private labels and manufacturer brands differs between retailers (Esbjerg et al., 2004). Some retailers mainly sell private label products (e.g., Marks & Spencer) or exclusive brands, i.e. products that identify neither the manufacturer nor the retailer, but are only sold by a specific retailer (this is the case for discounters such as Aldi and Lidl) (Varley, 2003). Other retailers sell mainly manufacturer brands (e.g. Nettorama), while retailers such as Tesco, Albert Heijn and C1000 sell both manufacturer and retailer brands. Thus, retailers differ with regard to retailer brand penetration. Most retailers carry manufacturer brands, but, increasingly, they also offer private labels. In the next sections manufacturer brands and private labels will be discussed in more depth.
2.2.1 Manufacturer brands
Consumers moved from no-name products of inconsistent quality produced by local factories in the nineteenth century to branded products from global manufactures such as Coca-Cola, Procter & Gamble’s Ivory soap and Nestlé’s infant formula. The manufacturer brands message is focussing on smart shopping (brands are trustworthy, delivering quality, consistency, and innovation at a fair price. Manufacturer brands reach consumers through distributors and retailers. Consumers started buying manufacturer-endorsed brands as symbols of quality, trust and affluence. Manufacturer brands have evolved to brands bought as symbols of aspirations, images, and lifestyles.
For most of the 20th century retailers were relatively, small, compared with their
2.2.2 Private labels
Private labels, often also referred to as own labels, store brands, or distributor-owned brands, are products owned and branded by organizations whose primary economic commitment is distribution rather than production (Schutte, 1969). One motivation for offering PLs is the higher percent margins that they provide to retailers (Hoch & Banerji 1993); another is the negotiating leverage they provide over manufacturers, as discussed above (Narasimhan & Wilcox 1998); and a third is that a private label brand can engender loyalty to the retailer (Ailawadi et al. 2008). In addition, since private labels are exclusively distributed products, the retailer avoids direct price competition, enhances store differentiation, and creates traffic (Davies, 1990). Private labels are growing rapidly. In the U.S., PLs have outperformed manufacturer brands in the last ten years (ACNielsen, 2003). They now account for 20 percent of U.S. sales in supermarkets and mass merchandisers as well as a healthy share of sales in department stores, category killers, specialty stores and convenience stores.
The three best-selling PL categories in food and non-food may still be predictable; milk, eggs, and bread in food; food storage and trash bags, cups and plates and toilet tissue in non-food. (Kumar & Steenkamp, 2007) However, today’s large and sophisticated retailers are able to develop credible PL offerings for categories where traditionally customers were more wary of staying from their favourite manufacturer brand names. The PL phenomenon is not restricted to consumer packaged goods (CPG) and grocery retailers. Best-in-class retailers and distributors such as Best Buy, Boots, Decathlon, Federated, Gap, IKEA, HEMA, Lowe’s, Office Depot, Staples, Target, Toys “R” Us, Victoria’s Secret, and Zara carry a large percentage of, or in some cases exclusively, private labels. The number and types of retailers and distributors that fall under the PL spell continue to increase. (Kumar & Steenkamp, 2007)
Therefore Kumar and Steenkamp (2007), describe four types of private labels: generic private labels, copycat private labels, premium private labels and value innovators. The next sections are based on Kumar and Steenkamp (2007) and will discuss the four types of PLs in more detail.
Premium private labels and value innovators
to the manufacture brands. There is no attempt to confuse customers and make them think that these are manufacturer brands, as is the case with copycat private labels.
Value innovators aim for objective product quality on a par with manufacturer brands at unbeatable prices, with a constant search to lower the prices continually. They achieve lower prices through rigorous system processing cost savings (e.g., low overheads, limited assortment, minimal displays, low staff intensity, efficient logistics) rather than by compromising on product quality. The champions of this type of value innovators have been the hard discount chains like Aldi, Lidl, and Netto in Europe.
Generic private labels
PLs started as cheap, inferior products. Historically, they did not even carry the name of the store and were therefore called ‘generics’. Usually, the package with black letters on a white background simply identified the product, like paper towels or dog food. These cheap, shoddy products, however, did offer lower-income and price-sensitive customers a purchase option, and as a result enabled the retailer to expand its customer base. Typically, generics did not account for a large proportion of the retailer’s volume, and as a result they were not strategically important to the retailer. Over time, generics have lost shelf space and importance to copycat store brands, premium store brands and value innovator own labels.
However, generics have become more important during the last years. To respond to the intense price pressure from hard discounters like Aldi and Lidl, mainstream retailers have been forced to develop a dedicated private label that identifies the lowest price at which a product is available in the store. By having such a lowest-price private label line, traditional retailers such as Albert Heijn, Carrefour, Delhaize, and Sainsbury demonstrate that they have a basket that is price competitive against the hard discounters. However, it is not entirely clear that these generics generate the required profitability to justify their shelf space. Furthermore, they may end cannibalizing the retailer’s own higher-priced and higher-margin private label range. Yet, the hope for retailers is that such a line attracts price-conscious customers, whose shopping basket ultimately ends up with a mix of low-margin generics as well as some higher-margin non generic products. Thus, while the lowest-price range may not be very profitable in its own right, it may still attract profitable shoppers.
Copycat private labels
encourages both brand comparison and brand confusion on the part of shoppers. To ensure quality, retailers analyze the contents of a leading manufacturer brand and then re-create the product step by step, a process called reverse engineering. In this sense, they are free riding on the manufacturer’s innovation, research, product development, and image-building efforts for its brand. Since there are few research and development or sales and marketing expenses for the retailer, and the products are aggressively outsourced for low-cost manufacturing, the price on such copycat PL-branded products is considerably lower than the referent manufacturer brand while still delivering high margins to the retailer, at least in percentage terms. The fight between copycat PLs and manufacturer brands can be very intense. In the Netherlands, Unilever placed full colour advertisements in national newspapers where they compare their products with the copycat PL of Albert Heijn. These advertisements were a reaction on the promotion campaign of Albert Heijn, where they stated to offer the same quality as manufacturer brands, for a lower price. For the advertisement see introduction, (figure 1.1).
Evolution of private labels
Over time, generics have lost shelf space and importance to copycat PLs, premium PLs, and value innovators. However generics still play an essential role in retail assortments by offering a price substitute for full service supermarkets against the value innovators of the discounters, see figure 2.2.
Next to the PL evolution, the figure gives a good overview of the positioning of the different types of PLs. The types of PLs are compared on price and quality versus manufacturer brands.
Example three alternatives choice set
As described in the introduction, in this study the effect of introducing a generic PL to a current two alternatives choice set (manufacturer brand and copycat PL) will be studied. In table 2.2 an overview of the brands in the new three alternatives choice set is presented for the product peanut butter.
Table 2.2: example brand architecture peanut butter at full service supermarket Albert heijn
Brand Albert Heijn Example Price Price index Price tier
Manufact urer brand € 2.12 125.4 First Copycat private label € 1.69 100 Second Generic private label € 0.94 57.3 Third
Note: source www.ah.nl, date: 30-06-2009
Next to the logo and product, the price and price index are presented. For the price index the copycat PL is the reference price and has an index of 100. The manufacturer brand is priced higher (125.4) and the generic PL is priced lower (57.3) compared to the copycat PL. Leading to a first- (manufacturer brand), second- (copycat PL) and third (generic PL) price tier.
2.3
Factors influencing consumer choice
will be introduced briefly. First the compromise effect. When a generic PL is added to the choice set a consumer’s shopping decision could change. Previous research in decision making and marketing indicates that consumer preferences for options in a choice set are influenced by the decision context, leading to the widely studied attraction and compromise effects (Burton & Zankhan, 1987; Simonson, 1989; Simonson & Tversky, 1992). Context effects imply that, when evaluating a focal option, individuals take into consideration characteristics of other comparative alternatives rather than only the features of that focal alternative, complicating the decision-making process (Sheng, et al., 2005). This means that when a consumer is evaluating the three options (manufacturer brand, copycat PL and generic PL) the respondents take the characteristics of all three options into consideration.
When a generic PL is added to the choice set a consumer can choose among two private labels (i.e. copycat PL and generic PL) and one manufacturer brand. Previous research of Zielke and Dobbelstein, 2007 indicates that consumer willingness to buy a new store brand (i.e. generic PL) instead of the manufacturer brand, is influenced by various factors. The main factors who are expected to influence this choice are: quality perceptions, decision involvement, product risks and brand experience. All factors will be discussed in more detail in the following sections.
2.3.1 Compromise and attractiveness effect
The essence of the compromise effect is that “an alternative would tend to gain market share when it becomes a compromise or middle option in the choice set” (Simonson, 1989). That is, a brand in a two-alternative choice set can gain more market share after the addition of an adjacent competitor that makes one brand a compromise choice in the choice set. The new entrant is relatively superior to the pre-existing options on one dimension and inferior on the other, effectively making on of the pre-existing options a compromise. In addition, Simonson and Tversky (1992) explain compromise effects by an individual’s “extremeness aversion,” due to the fact that disadvantages of an alternative are weighted more heavily than the corresponding advantages. Empirical studies have demonstrated that compromise effects are not exceptional, but both common and robust (Simonson, 1989; Simonson & Tversky, 1992).
The distinction between the compromise and attractiveness effect is that the compromise effect does not share the dominance relationship of the attractiveness effect. The new alternative in the compromise effect is superior on attribute 1 and inferior on attribute 2, while for the attractiveness effect the new alternative is inferior on both attributes. The distinction between compromise and attractiveness effects is presented in figure 2.3.
Figure 2.3: comparison of attraction and compromise effects (Sheng, et al., 2005)
In the above figure the letters stand for a brand positioning, based on two attributes. The attributes can be interpreted as quality (attribute 1) and good price (attribute 2). For example: A scores high on “quality” and low on “good price”. The essential distinction between the compromise and attractiveness effect is that the decoy brand X is completely dominated by the brand B in the attraction-effect situation, whereas brand C is perceived as equivalent to brand B in the compromise-effect situation.
In this research the compromise effect is applicable. A generic PL is superior on the attribute “price” and inferior on the attribute “quality” compared to the copycat PL and manufacturer brand.
The results of the study of Sheng et al., (2005) show that the factors decision uncertainty and product familiarity affect the compromise effect. This study will analyse if these factors influence the compromise effect in this study as well. Decision uncertainty and product familiarity will be discussed in more depth in the following sections.
Decision uncertainty
A consumer’s shopping decision almost always involves uncertainty. Uncertainty in a shopping decision is a psychological state that results primarily from a lack of adequate information or knowledge concerning the outcome of a purchase situation (Duncan, 1972). Consumers can be uncertain as to which option will give them the greatest value. In the compromise-effect decision context, a consumer makes a decision to maximize the expected gain, equal to minimizing the expected loss (Sheng et al., 2005). Yates (1990) speak of a loss when it results in an outcome that falls short of specified reference point. The expected loss is minimized when choosing for the compromise option (i.e. the copycat PL) instead of the more extreme manufacturer or generic PL options. In conclusion a consumer with higher decision involvement is more likely to avoid risks and choose the compromise option:
H2: the higher an individual’s decision uncertainty, the more likely he or she will choose the compromise option (i.e. copycat PL) in the choice set.
Product familiarity
H3: the more familiar an individual with the product, the less likely he or she will choose the compromise option (i.e. copycat PL) in the choice set.
2.3.2
Private labels versus manufacturer brands
When a generic PL is added to the choice set a consumer can choose among two private labels (i.e. copycat PL and generic PL) and one manufacturer brand. In this paragraph the factors that are expected to influence this choice will be discussed.
Quality perceptions
Introducing a generic PL to the choice set has already been reported in 1981 to have an effect on consumer quality perceptions of both private labels and manufacturers brands (Wheatley, 1981). Monroe (1979) has suggested that price perceptions of particular products are influenced by the prices of similar products. This is interesting for this study as the effect of adding a generic PL to the choice set of similar products is analysed. The study of Wheatley (1981) demonstrated that a new brand selling at a lower price than that prevailing prior to its introduction seemed to have the effect of altering consumer perceptions of the older brands within the product category. The new low priced product (generic PL) had the effect of lowering the “average” or reference price of the affected product category for consumers. Resulting that the previously low priced copycat PL, lose its “bargain” image and become “moderately” priced. This suggests that, since price is often used as a cue for product quality (Malhotra, 2004), some previous manufacturer brand buyers might switch to copycat PLs because of their enhanced quality image when generic PLs are added to the choice set. In conclusion consumer’s quality perceptions of the copycat PL are expected to increase when a generic PL is added to the choice set:
H4: when a generic PL product is added to the choice set, consumer’s quality perceptions of copycat PLs will increase.
Decision involvement
the lowest quality and thereby the exact opposite of the “optimal manufacturer brand”. Consumers who are highly involved are expected to be less likely to choose the generic PL, leading to the following hypothesis:
H5: the higher a customer’s decision involvement, the more likely he or she will choose the generic PL in the choice set.
Product risks
Results from previous studies indicate that risks related to a product group have an influence on consumer willingness to buy new store brands. (Semeijn et al., 2004; Batra and Sinha, 2000). and Semeijn et al., (2004) distinguish financial, functional and social risks. Financial risk is defined as the potential financial loss resulting from a bad purchase. So, the financial risk depends on the process level of the product group. It will be higher for products like laundry detergent or sparkling wine than for butter. Functional risks are defined as the potential loss resulting from an inadequate product quality. These risks are relevant when the function of a product is important. For example, there could be a certain risk that a laundry detergent will not clean the washing or that a shampoo desiccates the hair. Finally, social risks affect a possible loss of image or prestige resulting from the purchase or use of certain products. These risks mainly exist with products that are consumed in public or offered to guests. In conclusion consumer choice is expected to differ between product groups, depending on certain risks which are typical for the respective product group:
H6a: the higher the financial, functional and social risk of a product group, the lower the preference for the generic PL.
H6b: the higher the social risk of a product group, the higher the preference for the manufacturer brand.
Brand experience
3.
Research design
In this chapter the research design for the descriptive/causal study will be described in which the hypotheses will be tested by using an online survey. First, I will briefly discuss the industry and company, i.e. Albert heijn, where the research is executed. Next, I will discuss the method, population & sampling, reliability, validity & representativeness and plan of analysis.
3.1
The Dutch supermarket industry and company
The hypotheses described above are empirically tested in the Dutch market for the full service supermarket Albert Heijn. Albert Heijn is founded in 1887 by Mr. Albert Heijn. Headquarters of Albert Heijn is established in Zaandam. With more than 750 stores and various formulas (AH wijkwinkel, AH XL, AH to go, AH doorbraak and web shop ah.nl) Albert Heijn is the biggest supermarket chain in the Netherlands. Albert Heijn is part of Royal Ahold. The mission of Albert Heijn is: “the daily affordable, the special within reach”. (het dagelijkse betaalbaar, het bijzondere bereikbaar) The slogan of Albert Heijn is currently: ‘Gewoon bij Albert Heijn’.
3.2
Method
Data for this study was obtained by means of an online questionnaire that consisted of two groups of respondents. Five product groups were used in the study. In this paragraph the two groups, the product selection and measurement scales will be discussed.
3.2.1 Groups of respondents
generic PL option) than in a two alternatives choice set (hypothesis 4). Figure 3.1 gives an overview of the research design for all hypotheses.
Figure 3.1: research design
GA: X1,4 O1
GB: X1,4 O2
GA: X2,3,5,6a,6b,7 O1
The variables compromise effect (X1) and quality perceptions (X4) will be exposed to the three alternatives choice set group (group A, GA) and the two alternatives choice set group (group B, GB). Leading to two observations (O1 and O2).
The variables decision uncertainty, product familiarity, decision involvement and brand experience will be exposed to only the three alternatives choice set group (GA). Leading to one observation (O1).
3.2.2 Product selection
To select the five product groups a preliminary research has been executed. I selected the products based on a field research and the research of Zielke and Dobbelstein (2007). The field research compared the current generics assortment of Albert Heijn with four major competitors. In total 300 different products were identified that were sold at one or more of the competitors but not sold at Albert Heijn. These products are discussed with management of the price, presentation and choice department. This discussion resulted in a list of 90 potentially interesting products. This list was the starting point of the final product selection. Next, I have used Zielke and Dobbelstein (2007), in their study they identify factors influencing customers’ willingness to purchase new store brands. One of the main factors expected to influence customers’ willingness to buy new store brands is the extent of financial/functional/social product risks. They have made a categorization of five different product groups based on associated financial, functional and social risks. As the level of risk associated with a new generic PL is expected to impact consumer choice, see hypothesis 6a, this categorization helps to select the product groups for testing this hypothesis, see fig. 3.1.
Table 3.1 product groups categorized by risks (Zielke and Dobbelstein, 2007)
Product group Financial risk Functional risk Social risk
Shampoo Low High Low
Potato chips Low Low High
Sparkling wine High Low High
Butter Low Low Low
Unfortunately, not every product group was listed in the 90 potentially interesting products, I could not use the exact same product groups. However my goal was to analyze five product groups which had a similar risk categorization as displayed in table 3.1. Therefore, I have executed a small research into the perceived financial/functional/social risk of the 90 potential product groups. Data was obtained by means of an online questionnaire. Based on a discussion with some retail experts and based on own estimation ten product groups were selected to include in the online questionnaire. I selected the following ten product groups: diapers, shaving foam, roasted peanuts, white wine, paprika powder (spices), laundry detergent, mouth wash, cherry chocolates, “Ossenworst” and bake butter. For every product group, 40 respondents answered questions concerning financial, functional and social risks on a 5- points Likert scale, based on Sridhar (2007), (see table 3.2). (See appendix 1 for the questionnaire.)
Table 3.2 questions, component and scale for preliminary research product groups selection
Question Component Scale
I do not incur financial loss if my choice of product is not going to work properly
Financial risk Sridhar (2007)
My family member/friends/relatives
laugh if I purchase wrong product Social risk Sridhar (2007)
Post purchase, if product does not function properly it does not bother me a lot
Functional risk Sridhar (2007)
Figure 3.2: product groups categorised by risks 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5 Sha ving foam Dia pers Roa sted pea nuts Bak e bu tter Whi te w ine Laun dry dete rgen t Pap rika pow der Mou th w ash Che rry choc olat es Oss enw orst Product groups A v e ra g e r is k Functional Social Financial 3 line
Every outcome is tested with a one sample t-test to test if they score significantly higher or lower than 3. Three product groups did not scored significantly higher or lower than 3: laundry detergent financial- (.06) and functional risk (.09), mouth wash financial- (.17) and functional risk (.77) and “ossenworst” social risk (.19). Four suitable product groups could be selected: diapers, shaving foam, wine and bake butter. Only one product group could not be selected, i.e. a product group who scored low on financial and functional risk and high on social risk. “Ossenworst” and “roasted peanuts” did not fit the profile. Therefore, a new product group had to be selected. After another look at the 90 potential products, I selected prawn salad. After discussing the option with experts and by asking 10 respondents to fill in the questionnaire again (fin. risk 2.4 functional risk 2.6 and social risk 3.4), prawn salad was selected as the fifth product.
3.2.3 Measurement scales
Table 3.3: questions, component and scale final research
Question Component Scale
I’m sure that my first choice is more desirable to me than the other two options
Decision uncertainty Sheng et al (2005) I am very familiar with <product> Product familiarity Sheng et al (2005) I estimate the quality of <brand
and product> is high Quality perceptions Wheatley et al. (1982)
When I buy <product> I attach great importance to make a good choice.
Decision involvement Zielke and Dobbelstein (2007)
Please divide 100 points among below product groups (more points, means you buy more products in this group).
Brand experience Zielke and Dobbelstein (2007)
I have made two changes to the constructs. First, I lowered the number of items per construct and second I changed the question for measuring brand experience. I will discuss both changes in more depth. By lowering the number of items per construct for decision uncertainty, product familiarity and decision involvement, I tried to make the questionnaire more accessible to affect the response rate positively. To establish validity of the constructs, I consulted a number of experts in marketing and market research concerning the downsizing of the items. There was a common believe it would not significantly effect the results. Furthermore, I have changed the question to measure brand experience to create more research options. For measuring brand experience I used the question of Zielke and Dobbelstein (2007) as a starting point. Zielke and Dobbelstein (2007) use a ordinal 5 points scale (1= <20 percent and 5 = >80 percent), where I have chosen a sum scale, which is a ratio scale. By using a ratio scale all research options are possible: objects can be classified or identified, objects can be ranked, and intervals or differences can be compared (Malhotra, 2007). The four remaining constructs decision uncertainty, product familiarity, quality perceptions and decision involvement were measured with a 7-item Likert scale, where 1= was “strongly disagree” and 7= “strongly agree”. All questions have been translated to Dutch, as all respondents were Dutch. Possible negative effect could be translation issues. (For the complete Dutch questionnaire see appendix 3.)
3.3
Population and sampling
To determine the sample size I used a statistical approach. Several factors are important to calculate the minimal sample size:
- The confidence level is the probability that a confidence interval will include the population parameter (Malhotra, 2007). In this research a confidence level of 95 percent will be applied, the corresponding z-value is 1.96
- The precision level is the desired size of the estimating interval. This is the maximum permissible difference between the sample statistic and the population parameter (Malhotra, 2007). In this research a minimal precision level of +/- 7,5% will be used. - The population proportion. Reasonable estimates for the population proportions are
not available. Therefore I use the maximum possible population variation = 0.5 (Malhotra, 2007).
- The population consist of Albert Heijn customers between 18 and 80 years old, who are responsible for the groceries within their household and were reachable by email. As the population is large enough (more than 20.000) a population correction is not applicable (Malhotra, 2007).
Based on the above factors the minimal sample size can be calculated (Malhotra, 2004):
Sampling technique
3.4
Reliability, validity and representativeness
To judge the quality of the research various criteria can be used: construct validity, internal validity, external validity, face validity and reliability (Malhotra, 2004). To increase reliability I have executed a pre test. I have sent the first version of the questionnaire to 10 respondents. After the ten respondents filled in the questionnaire, I discussed the questionnaire with them. Based on the pre test some questions were formulated differently and one question was added to the questionnaire. To realize construct validity multiple theoretic sources have been used to specify the theoretical relationships, to make sure each construct will measure what it should measure. Furthermore various experts have been asked their opinion concerning the constructs, hereby taking face validity into consideration.
3.4.1 Sample characteristics
Important for the reliability of the research is that both groups are comparable. By comparing the sample characteristics I gain insight in possible differences between both groups. In appendix 4 an overview is given of all sample characteristics. In figure 3.3 the average score on various characteristics of both groups is displayed.
Figure 3.3: graphical representation of sample characteristics
In general the distribution is quite similar. For both groups the majority of respondents were women (respectively 66,9 and 59,6 percent) and most respondents are single. However one difference does appear. Group B contains more primary AH customers than group A (58.3 against 42.9 percent). Group A contains more secondary AH customers than group B (46.7 against 29.7 percent). To test if the difference in secondary and primary customers between the two groups influence results I have compared both groups on the variable brand experience. I have compared both groups with a independent sample T-test, see table 3.4 for the results.
Table 3.4: comparison group A and B on brand experience
Variable Choice set Mean Std.
Deviation Significance (2-tailed)
Two alternatives 45.54 28.20 .18
Experience manufacturer
brand Three alternatives 41.63 22.7
Two alternatives 47.66 73.36 .08
Experience copy cat PL
Three alternatives 36.63 16.71
Two alternatives 18.41 26.92 .22
Experience generic PL
Three alternatives 21.67 26.91
Based on a significance level of .05 the two groups do not significantly differ on brand experience of manufacturer brand, copy cat PL and generic PL. However based on a significance level of .10 the two alternatives choice set group scores significantly higher on copycat PL experience than the three alternatives choice set. This higher level of copycat PL experience will be taken into consideration in chapter 5, “conclusions and recommendations”.
3.4.2 Response statistics
Table 3.5 gives an overview of the response statistics.
Table 3.5: response statistics
Group A (incl. Generic) Group B (excl. Generic)
Started 174 182
Drop-outs 12 18
Completed 162 164
Control characteristics 10 8
Usable 152 156
The low number of drop-outs can be explained by the short and convenient questionnaire and the relation of the respondents with the researcher. After the control characteristics were applied, another 18 respondents where not usable for analysis. Resulting in 152 usable respondents for Group A and 156 usable respondents for Group B.
The minimal sample size calculated in paragraph 3.3 was 170 respondents. As two groups are compared, each group has to contain 170 respondents. However due to limits of time (the collection process lasted for 2.5 weeks) the sample size used for analysis is for group A 152 respondents and for group B 156 respondents. The consequences of this lower response rate are that the results will be less reliable when applying the results to the entire population.
3.5
Plan of analysis
In this paragraph I will describe for every hypothesis how it will be statistically tested and analysed. In the next chapter the analysis and results will be discussed.
H1: when a generic PL is added tot the choice set the market share of the copycat PL will increase.
The compromise effect suggests that the market share of the compromise option (copycat PL) will increase as a result of the addition of the third brand (generic PL) to the choice set. To test if the compromise effect is applicable, I will make an overview of the market shares for each product for the two groups (the two alternatives choice set group and the three alternative choice set group). When the three alternative choice set has a significantly larger market share for the products the above hypothesis is supported.
H2: the higher an individual’s decision uncertainty, the more likely he or she will choose the compromise option (i.e. a copycat PL) in the choice set.
computing variable I will run the independent samples T-test. First, I recode for every product the choice variable to: copycat PL = 1, manufacturer brand and generic PL = 0. Next, I perform the independent sample T-test analysis for each product. The new variable “choice binary” is the grouping variable and “decision uncertainty” is the test variable. Based on the output I can conclude for every product if consumers scoring higher on decision uncertainty are significantly more likely to choose the compromise option (i.e. copycat PL).
H3: the more familiar an individual with the product, the less likely he or she will choose the compromise option (copycat PL) in the choice set.
The hypothesis suggests that respondents more familiar with an product are more likely to choose the middle option/compromise option (i.e. copycat PL). To test this, I will perform a independent sample T-test. I use the recoded choice variable from hypotheses 2: copycat PL = 1, manufacturer brand and generic PL = 0. Next, I perform the independent sample T-test analysis for each product. The new variable “choice binary” is the grouping variable and “product familiarity” is the test variable. Based on the output I can conclude for every product if consumers more familiar with the product are significantly less likely to choose the compromise option (i.e. copycat PL).
H4: when a generic PL is added to the choice set, consumer’s quality perceptions of copycat PLs will increase.
Tot test if the three alternatives choice set group (including generic PL option) score significantly higher on quality perception of the copycat PL than the two alternatives choice set group, a independent sample T-test is executed for each product group. The grouping variable used in the analysis was “Group” (two alternatives choice set=0, and three alternatives choice set=1. The test variable used was “quality copycat PL <product>”. Based on the output I can conclude for every product if consumers in the three alternatives choice set perceive higher quality for the copycat PLs.
H5: the higher a customer’s decision involvement, the less likely he or she will choose the generic PL option in the choice set.
First I create a new binary variable by recoding choice: generic PL=1, and manufacturer brand and copycat PL=0. Next I perform the independent sample T-test analysis for each product. The new variable “choice binary” is the grouping variable and “decision involvement” is the test variable. Based on the output I can conclude for every product if consumers higher involved in the decision are significantly less likely to choose the generic PL option.
H6a: The higher the financial, functional or social risk of a product group, the less likely a consumer will choose the generic PL.
H6b: The higher the financial, functional or social risk of a product group, the more likely a consumer will choose manufacturer brand.
The above hypotheses suggests that consumer choice between the three alternatives (manufacturer brand, copycat PL and generic PL) is influenced by the associated risk (financial, functional and social) of the product group. To test the hypotheses 6a and 6b, I will calculate the averages of the products that are classified as having high financial/functional/social risk versus low financial/functional/social risk. Next, I perform a Chi-square test, to test if low financial/functional/social risk product groups score significantly different than high financial/functional/social risk groups. Based on the output I can conclude if consumers, for product groups associated with high financial/functional or social risk, are more likely to choose manufacturer brands and less likely to choose generic PL products.
H7: the higher a customer’s manufacturer brand/copycat PL or generic PL experience, the more likely he or she will choose the manufacturer brand/copycat PL or generic PL in the choice set.
Other analysis
4.
Analysis and results
Based on the plan of analysis in chapter three, all hypotheses will be tested in this chapter. First, the hypotheses in relation with the compromise effect will be analysed (h1, h2 and h3). Next, the hypotheses concerning the factors expected to influence the choice between manufacturer brands, copycat PLs and generic PLs private labels will be analysed (h4, h5a, h5b, h6 and h7). Finally consumers’ switching behaviour will be analysed.
4.1
Compromise effect
In this paragraph the hypotheses related to the compromise effect will be analysed. First, the analysis and results for the compromise effect will be discussed. Next, analysis and results for the two factors (i.e. decision uncertainty and product familiarity) expected to influence the compromise effect will be addressed.
The compromise effect suggests that the market share of the compromise option (i.e. copy cat PL) will increase as a result of the addition of the third brand (i.e. generic PL) to the choice set (Simonson, 1989). To test if the compromise effect is applicable for the five product groups the market share for manufacturer brand, copycat PL and generic PL for every product group per choice set is calculated, see table 4.1.
Table 4.1: compromise effect across product groups Product
group
Choice Set Manuf. Brand
(share) Copycat PL (share) Generic PL (share) Missing Values Two alternatives 46.2 51.9 1.9 Diapers Three alternatives 35.5 38.8 23.7 2 Two alternatives 26.3 72.4 1.3 Bake butter Three alternatives 20.4 39.5 38.8 1.3 Two alternatives 27.6 68.6 3.8 Prawn salad Three alternatives 22.4 48 27.6 2 Two alternatives 45.5 50 4.5 Shaving
foam Three alternatives 47.4 28.9 22.40 1.3
Two alternatives 48.7 48.1 3.2 White wine Three alternatives 38.2 52.6 5.9 3.3 Two alternatives 38.9 58.2 2.9 Average score
5 products Three alternatives 32.8 41.6 23.7 1.9
Note: the bold numbers represent the highest market share on copycat PL between the two choice sets.
choose the manufacturer brand and 51.9 percent the copycat PL. In the three alternatives choice set group, 35.5 percent choose the manufacturer brand, 38.8 percent copycat PL and 23.7 percent the generic PL. For the product group diapers this means that the two alternatives choice set group score higher on copycat PL market share (i.e. the compromise option) than the three alternatives choice set group.
Next to diapers the product groups bake butter, prawn salad and shaving foam have a larger market share for the compromise option (i.e. copycat PL), in the two alternatives choice set than in the three alternatives choice set. For these four products the compromise effect is not applicable. The market share for copycat PL white wine does increase when a generic PL is added to the choice set. For the product white wine the compromise effect is applicable. Based on the average score of the five products it seems the market share of the generic PL option (+23.7%) is mainly responsible for the market share decrease of the copy cat PL option of (-16.6%).
To test if these differences between the two groups are significant, an independent sample T-test is executed, see appendix 5 for the results. The significance level is tested 1-tailed as a one sided positive relation is expected; the market share of the copycat PL will increase when the generic PL is added to the choice set. All products except white wine were significant based on a significance level of .05 (white wine had a significance level of .213). So, although white wine was the only product with a higher market share in the three alternatives choice set, this higher market share is not significant. Leading to the result that, contrary to expectations, all products do not have a significant higher copycat PL market share in the three alternatives choice set compared to the copycat PL market share in the two alternatives choice set.
Although the compromise effect is not applicable the factors expected to influence the compromise effect (i.e. decision uncertainty and product familiarity) will be studied. As it can provide insights in the factors influencing consumer choice when a generic PL is added to the choice set.
4.1.1 Decision uncertainty
T-test is executed. The average score on decision uncertainty for respondents choosing the manufacturer brand, copycat PL or generic PL per product group, is presented in figure 4.1.
Figure 4.1: score on decision uncertainty per choice, per product group
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5
Diapers Bake butter Prawn Salad Shaving foam White wine Product groups D e c is io n u n c e rt a in ty Manufacturer Brand Copycat PL Generic PL
Figure 4.1 shows that on average respondents who chose the manufacturer brand score lowest on decision uncertainty, followed by copycat PLs. Respondents who have chosen the generic PL are most uncertain if their choice was the best option for him or her. In table 4.2 the results of the one sample T-test (i.e. the mean, standard deviation and significance) are stated per product group per choice. For the complete output of the independent sample T-test, see appendix 6.
Table 4.2: decision uncertainty affecting consumer choice
Product group Choice Mean Std. Deviation Significance
generic PL option
Copycat PL option 1.86 1.43 .29
White wine
Man. brand or
generic PL option 1.72 1.65
Note: the bold numbers represent the highest decision involvement between the two choices per product group.
For the product group diapers, respondents who chose the copycat PL scored on average higher (2.54) on decision uncertainty than respondents who chose the manufacturer brand or generic PL option (2.09), although this difference is not significant at the .05 level.
The significance level is tested 1-tailed, as hypothesis 2 suggests a one sided positive relation; the higher an individual’s decision uncertainty, the more likely he or she will choose the compromise option (i.e. copycat PL) in the choice set. For four product groups (i.e. diapers, prawn salad, shaving foam and white wine), respondents score higher on decision uncertainty when they have chosen the copycat PL, but this was not significant at the .05 level. For the product group bake butter, respondents score, contrary to expectations, lower on decision uncertainty when they have chosen the copycat PL, nevertheless this is not significant at the .05 level.
4.1.2 Product familiarity
In the three alternatives condition, a respondent’s product familiarity was expected to influence his/her choice. More specifically hypothesis 3 expected the more familiar an individual with the product, the less likely he or she will choose the compromise option (i.e. copycat PL) in the choice set. A consumer familiar with the product is less likely to simply choose the middle option (i.e. copycat PL), because it is a compromise. The average score on product familiarity for respondents who chose the manufacturer brand or copycat PL or generic PL, is presented in figure 4.2 per product group.
Figure 4.2: average score on product familiarity per product per choice
