Consumer perceptions on store brands : a comparison of private label products to nationally branded ones in an online grocery shopping environment

(1)

University of Amsterdam – Amsterdam Business School Faculty of Economics and Business

MSc in Business Administration – Marketing Track

Master Thesis

Consumer perceptions on store brands:

A comparison of private label products to nationally branded ones in an online

grocery shopping environment

Author: Annamária Kovács Student Number: 11374934 Supervisor: Kristopher Keller

(2)

Statement of originality

This document is written by Annamária Kovács who declares to take full

responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that

no sources other than those mentioned in the text and its references have been

used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision

of completion of the work, not for the contents.

(3)

Abstract

Consumer perceptions is a frequently studied area by researchers, as they can have a significant effect on people’s final purchase decision. These perceptions not just able to change over time, but can differ in case of private label or nationally branded products, shopping environments (offline vs online), product categories (hedonic vs utilitarian) and also in case of brand tiers (economy, standard, utilitarian) (Peter & Olson, 2001). This study considers all the above mentioned factors in its comprehensive conceptual model, and aims to investigate the perceptions of private label products compared to nationally branded ones, together with the specific features of the online shopping environment.

The main goal of this paper is to shed light on how consumers perceive private label products, and do these perceptions differ compared to nationally branded ones in an online grocery shopping environment. In order to investigate the aforementioned questions, a simplified theory of product placement (considering one row and the products are placed in the middle or on the right of this row), based on its importance in the in-store environment, has been taken as a baseline. Brand tier (economy, standard and premium) and product category (utilitarian and hedonic) were also included in the research to identify the underlying moderating effects.

The conducted online experiment reached over 260 people and 215 participants provided valid responses to the questions about a fictional online grocery shopping platform. Participants were assigned to one hedonic and one utilitarian condition, and were asked the same survey questions related to their product choice, the main motivation of their choice, the importance of brand name to them, and their product perceptions.

A linear mixed model, more specifically a random intercept model, whereby in addition to the fixed effects of product category, product placement, brand importance, purchase motivation and brand tier, random individual effects were also included (accounting for the fact

(4)

that each person was assigned to two conditions), was used to analyze the outcome of the online experiment and to test all four proposed hypotheses. The sample for the analysis was restricted to only those responses in which private label products were chosen, as all hypotheses referred to the perception of PLs. After the testing, no significant relationships and effects were found in case of hypothesis 1 (main effect between PL product placement and perceptions), hypothesis 2 (moderating effect between utilitarian PL product placement and perceptions), hypothesis 3a (moderating effect between economy tier PL product placement and perceptions), hypothesis 3b (moderating effect between premium tier PL product placement and perceptions), therefore all hypotheses were rejected.

The analysis was extended to the overall data, and then to only those responses in which nationally branded products were chosen originally, in order to investigate the general and the additional underlying effects. Results of the follow-up analysis indicated that the importance of brand name has a positive and significant underlying effect in case of NBs. In other words, brand name has a significant and positive effect on consumers’ perceptions of nationally branded products, indicating that brand name is more important in case of nationally branded than in case of private label products. During the follow-up analysis brand tiers were also found to have interesting effects on product placement. On the one hand, the lowest brand tier (economy) had no moderating effect on product placement neither in case of the overall data, nor in case of nationally branded products. On the other hand, the highest brand tier (premium) found to have a significant positive effect on product placement (middle placed products) in general (and in case of NBs as well, which was supported by the results of the overall data). Consequently, findings indicate that in general, premium products placed in the middle of the row has a significant effect on consumers’ product perceptions.

Managers can benefit from the above mentioned findings while managing their online grocery shopping platforms. The most important implications are as follows: 1. product

(5)

placement (in the middle or on the right of the row) does not have a significant effect on consumers’ perceptions and shopping behavior in an offline environment, under the tested circumstances, therefore, in general, the success of a product in an online shopping platform is not significantly dependent on where the product is placed in a row; 2. brand name is more important for nationally branded than for PL products, consequently, it is essential for managers to offer the best selling nationally branded products on their online shopping platforms; 3. products under the different tiers should be managed differently, and managers should consider placing premium tier NBs in the middle of a row instead of placing them at the right side.

Finally, considering the limitations of this study, future research could focus on using a bigger sample size for the analysis, including demographics related questions in the survey, using a more natural or real-life experiment and adding elements to the presented model in order to overcome the complexity of product placement.

(6)

List of Tables and Figures

Figure 1: Look and feel of the completed fictional online shopping platform……….…26

Table 1: Overview on all experiment conditions.……….27

Table 2: Results of linear mixed model and linear model, with out_stock as dependent variable………..………....34

(8)

1. Introduction

Private label (PL) products and also consumers’ perceptions of them have been under researchers’ scope in the past decades due to their wide spread and growing importance (Geyskens et al., 2010). Private label products are not only relevant because of their rapidly growing market share (Steenkamp et al., 2010), but their positive perception can result in more frequent store visits, stronger and favourable feelings towards a specific store or retailer, and even in higher store loyalty (Ailawadi et al., 2008). However, as the chief e-commerce and innovation officer at Ahold Delhaize emphasized in her 2016 ‘Strategy in a Digital Age’ presentation, the digital age and its new technologies are changing consumers’ shopping habits and also their perceptions on products (Faber, 2016).

Online shopping platforms and their benefits – such as convenience, speed, and time saving – has encouraged consumers to use the online shopping environment frequently in many areas of their everyday lives, including grocery shopping (Chu et al., 2010). But in order to enjoy the benefits offered by online shopping, people have to give up the physical examination and the quality check of the products before purchase (Anckar et al, 2002). Due to the aforementioned differences between the traditional offline and the modern online shopping environments, researchers focused their attention towards whether the same consumers behave differently in the online and offline shopping environments while doing grocery shopping. Despite the fact that several studies investigated consumer behaviour in case of different shopping environments, people’s perceptions – which can have a significant effect on their preferences and their final purchase decision – on brands in the different shopping environments has not received as much attention during the years (Peter & Olson, 2001).

The importance of consumers’ perceptions on products and brands lays in their impact on the final purchase decision (Peter & Olson, 2001). These perceptions, however, can change over time, just like in case of the judgement of private label product compared to nationally

(9)

branded ones (Rossi et al., 2015). Nowadays the perception gap between PLs and NBs is quite narrow, but underlying factors, such as product brand, product category, product placement and brand tier can still have an influential effect on consumers’ perceptions (Rossi et al., 2015). Although several studies investigated perception effects and differences, also in case of private label products and nationally branded ones, the knowledge on this topic is mainly based on researches conducted in an in-store environment, and little is known about whether the findings can be extended to other (e.g. online) shopping environments as well.

As the aim of this research is to study whether consumers perceive private label products compared to nationally branded ones differ in an online grocery shopping environment, my research framework includes the aforementioned research gaps and the factors, which can have a significant effect on people’s perceptions. On the one hand, my study sheds light on how consumers perceive private label products, and whether their perception differences compared to national brands are relevant in a purely online environment as well. On the other hand, my study investigates whether the perception differences related to product placement can be extended to the online environment too. Finally, the study offers insights on what products should be offered in an online grocery shopping platform (PLs vs NBs) taking into consideration the influence of product category and brand tier.

This research contributes to the academic knowledge with a comprehensive study on product perceptions and their differences. The developed research framework considers several meaningful factors (product placement, product category, and brand tiers), and places the study into a new, and different environment (online grocery shopping). Consequently, the study combines the investigation of perception differences between PLs and NBs, and its underlying factors (brand tiers and product category), and includes the specific features of an online shopping environment as well. For managers the results offer help to form effective online

(10)

marketing strategies depending on whether they would like to increase the sales of PLs or NBs, and to manage their online shopping platforms accordingly.

The literature review first uses a broader perspective in order to introduce the differences between the online and offline shopping environments, then it focuses on how people perceive private label products and national brands. The second part of this section gives a closer look on PL brand tiers, the differences between hedonic and utilitarian products and the importance of product placement. The literature review ends with the identified research gap, the proposed research question and the research’s contribution to the academic knowledge and managerial applications. In the conceptual framework part, the conceptual model and the proposed hypotheses for the experiment are introduced, which is followed by an overview on the data collection (including research design) and method. The following analysis section gives a detailed picture about the performed analytical work, starting with the data preparation, and then turning to the methodology (model description and results table), the results of the hypothesis testing, and finally to the results of the follow-up analysis (overall data and NB subsample). Then the conclusion section summarizes the main findings, and offers for managers based on the. Finally, the limitations section discusses the limited aspects of the study and recommends possible extensions and directions for future research.

2. Literature review

In this section, the main focus is on those areas of the studied literature which are closely related to the research question. First, the major differences between private label products and national brands is investigated, and from this broader perspective the focus is narrowed down to the perception of the different private label brand tiers (economy, standard and premium). Then the differences between the online and the offline shopping environments is introduced. The review continues with the differences between hedonic and utilitarian products, and the importance of product placement. Then the literature review closes with the identified research

(11)

gaps in order to emphasize the relevance of the research. Finally, the expected contributions to the academic literature and the managerial implications are indicated.

2.1. Perception differences between private label products and national brands

Store or private label brands (PLs) are the ones, which represent the retailers’ own label within a product category (Ailawadi et al., 2008). Almost every major grocery retailer has already introduced its private label offerings to the public, and these ‘standard’ or ‘regular’ private label products are positioned and seen in most cases as mid-quality alternatives (Geyskens et al., 2010). The general perception on private label products is that they are good deals, and most consumers choose PLs when they are looking for an alternative to save money on groceries (Garretson et al., 2002). Private label products are perceived and handled differently from the so-called national brands, which are offered by several different retailers (Ailawadi & Keller, 2004), and are typically priced below non-price promoted nationally branded goods (Garretson et al., 2002).

The base of the aforementioned distinction is that the price-quality construct addresses issues towards the perception of PLs by stating that the low price of PL brands can be attributed to some problematic aspects of the product, therefore these products can be perceived as inferior in the overall level of quality compared to national brands (Garretson et al., 2002). However, a recent paper found evidence on that private label products becoming more and more accepted by consumers over time (Steenkamp et al., 2010), which questions the existence of the huge perception gap between PLs and NBs. Furthermore, the rapidly growing sales of private label products across the globe (Sethuraman & Gielens, 2014), and the results of blind taste tests indicate that people has high purchase intentions towards PLs, supporting the assumption that ‘consumers increasingly consider private labels to be as good as national brands’ (Rossi et al., 2015, p. 74). Although different factors, such as product brand, can still have a significant impact on people’s perceptions and eventually on their final purchase decisions, recent studies

(12)

show that people’s perceptions towards PLs have changed over the years, and differs from the common belief that private labels are lower-quality and lower-priced options compared to national brands (Rossi et al., 2015).

Consumers’ perceptions on products not just change over time, but it can vary depending on the environment within the decisions are made (Richardson et al., 1996). Findings support that store aesthetics – the interior design, the store layout, the music, the lighting, the colours and the overall tidiness – can have a significant effect on consumers’ perceptions of store brands (Richardson et al., 1996). Consequently, store atmosphere can lower risk perceptions, have a favourable effect on the positioning of store brands, and increase people’s willingness to purchase private label products (Delgado-Ballester et al., 2014).

Based on the above mentioned reasoning, the offline environment has a significant (positive or negative) effect on consumer perceptions and decisions, as it can trigger emotional states in consumers’ minds and eventually can influence consumer response (Richardson et al.,1996). One can argue that as private label products’ perception can be increased or decreased in an in-store environment, will the same effect be present in case of other, for instance online shopping environments as well.

2.2. Private label brand tiers

The general perceptions on private label products and nationally branded ones have changed over the past years, but some factors, for instance goods categories and countries, can influence the success of PLs (Steenkamp & Geyskens, 2014). The former factor has a significantly important role in how consumers perceive private label products, as within the same product category retailers tend to introduce not just one, but two or even more store brands (Palmeira & Thomas, 2011).

Retailers can shift from an undifferentiated single private label product line to a differentiated, three-tiered PL portfolio by following the ‘good, better, best’ approach

(13)

(Geyskens et al., 2010). In order to accommodate a wider range of consumer preferences, they offer a low-end, a standard or medium and a premium option to consumers within the same product category (ter Braak et al., 2013). On the one hand, introducing multi-tier store brands enabled retailers to reach a wider customer base and to compete with the high quality NBs with their premium offerings (ter Braak et al., 2014). On the other hand, one tier can succeed at the expense of the others: economy PLs found to cannibalize the standard (higher-margin) private label products (Geyskens et al., 2010).

The products under the different PL brand tiers are positioned to compete with different NBs: while standard private label products compete with mainstream manufacturer brands, the premium ones are positioned at the top end of the market, and they challenge the premium NBs (ter Braak et al., 2014). The differences in case of competing brands and positioning indicates that consumers tend to perceive the brand tiers and also the products related to them in a different light. Consequently, one can argue that brand tiers can also have a significant effect on people’s perceptions, and this factor should not be excluded from the investigation on consumers’ perceptions of PLs.

2.3. Online vs offline shopping environments

Several retailers have already been incorporated the Internet into their multichannel strategy besides the existing offline ones, as both the traditional brick-and-mortar stores and the online shopping platforms have their unique features which significantly differentiate the two from each other (Chu et al., 2010). On the one hand shopping online is a highly convenient way of purchasing products, because consumers can avoid standing in ques, the orders are delivered to people’s doors, and consumers can shop from the comfort of their homes. On the other hand, in order to enjoy the benefits (e.g. rapidity, convenience and simplicity) of online shopping, consumers have to give up the opportunity to physically examine the quality of the offered products (Anckar et al, 2002). The lacking possibility for quality assessment in online shopping

(14)

platforms is more significant in case of groceries, as a large sum of these products are perishable goods which consumers in general like to examine personally prior to shopping (Anckar et al, 2002).

The aforementioned ‘health risk’ – consumers perceive a higher quality uncertainty when buying food products than in case of non-food products – serves as a possible explanation that the perceived risk is increased in the online shopping environment, especially in case of grocery products (Chu et al., 2010). Consequently, brands with a high market share enjoy a loyalty advantage in the online shopping platforms, as consumers tend to stick to well-known brand names, over low market share brands, which tend to be more successful in the offline stores (Chu et al., 2010).

As consumers’ motivation regarding the usage of online vs offline shopping environments vary, mostly due to the different offered benefits, one can argue that people’s perceptions on products differ between the shopping environments as well. The above mentioned ‘health risk’ effect supports this assumption, therefore consumers’ product choices and brand perceptions can vary in the different shopping environments, especially in case of groceries. The higher, health related risk explains the perceived superiority of well-known brands in online grocery shopping environments (Chu et al., 2010). However, it is unknown to which extent is this assumption still valid, and whether consumers’ perceptions remain the same or differ when the health risk is not the primary factor in the decision making process.

2.4. Hedonic vs utilitarian products

Consumers can categorize goods based on their nature and the benefits they offer: in that sense, utilitarian products provide practical functionality, while hedonic ones promise experiential enjoyment (Okada, 2005). Despite the fact that most products involve both hedonic and utilitarian aspects to some extent, people tend to characterize certain products as primarily hedonic – ‘whose consumptions is primarily characterized by an affective and sensory

(15)

experience of aesthetic or sensual pleasure, fantasy and fun’, and others as primarily utilitarian – ‘whose consumption is more cognitively driven, instrumental, and goal oriented and accomplishes a functional or practical task’ (Wertenbroch & Dhar, 2000, p.61).

The perceived product characteristics can underlie consumer choices, and as consumer perceptions and preferences in most cases have both hedonic and utilitarian dimensions, the consumption of these products can vary in typical purchase situations, depending on the decision context (Wertenbroch & Dhar, 2000). Furthermore, consumers not only perceive the benefits and the nature of hedonic and utilitarian good differently, but they also differ in the terms of what they presume as hedonic and as utilitarian goods (Okada, 2005).

The assumption that the purchase decision context has an effect on the consumption of hedonic and utilitarian products is in line with the findings of a recent study, in which the author emphasizes that people in general prefer to pay in time for hedonic goods (willing to go the distance for a good bargain – offline purchases), and in money for utilitarian ones (willing to pay a premium for convenience – online purchases) (Okada, 2005). Based on this reasoning one can argue that the product category (hedonic or utilitarian) also has an effect on product and brand perceptions, therefore this aspect is included in this research as well.

2.5. Product placement

In-store product placement can also have a significant effect on people’s perceptions of brands and products (Sigurdsson et al., 2009). As consumers tend to rely on environmental cues (e.g. accessibility and visibility) in order to guide their purchase decisions, placing products to more visible and accessible locations in stores can increase the likelihood that buyers will purchase these brands (Wong et al., 2015). Furthermore, in-store product placement is perceived to be one of the underlying factors of impulsive food purchasing (Nakamura et al., 2014). Findings of a recent study indicate that not just in-store product placement in general has a significant role in shaping people’s perceptions, but shelf placement in general (brands

(16)

placed on the middle shelf increased sales compared to the ones placed on the high or low shelves) and within shelf placement also affects people’s buying behaviour (Sigurdsson et al., 2009). The effect of within shelf placement on consumers’ perceptions and buying behavior, is further supported by another article, which states that those products which are placed at the start of an aisle do not sell well, as people need time to adjust to their environment, and consequently middle placement on shelves is a more preferred option (Kendall, 2014).

By manipulating the placement of items on store shelves, the aforementioned study used one of the methods of assessing the effects of shelf placement on consumer behavior. As the findings demonstrate that product placement has a significant effect on consumer behavior and consumer perceptions in an in-store environment, one can argue that a similar effect can be observed in case of an online shopping environment as well.

2.6. Identified research gaps

The reviewed literature highlighted that several studies focused on the different shopping environments and their effects on consumer choices (Chu et al., 2010; Anckar et al, 2002). These studies mainly investigate the final purchase decision, which is the last stage of the shopping process, and the why behind it. In this thesis, the focus takes ‘a step back’ and the attention is on the perception of the different products and brands, primarily on the private label ones. PLs are present in several product categories, and they have the ability to mature over time. In other words, people tend to accept these products as alternatives to national brands (Steenkamp et al., 2010). Furthermore, researchers are becoming less certain whether consumers will have the motivation to pay extra money for nationally branded products in the future or not (Steenkamp et al., 2010).

As how consumers process information and perceive products and brands can have a significant effect on the final purchase decision (Petty & Cacioppo, 1983), it is important to further investigate this topic. Peter and Olson’s model, the Three Elements for Consumer

(17)

Analysis, further supports the motivation to dedicate attention to people’s perceptions (Peter & Olson, 2001). These aforementioned elements – Consumer Affect and Cognition, Consumer Environment and Consumer Behaviour – interact with each other, therefore a change in one element can affect the others, including the consumer’s final decision (Peter & Olson, 2001). Based on this presumption, if consumers’ perceptions on brands change, their preferences and eventually their final purchase decision can also be different (Peter & Olson, 2001). Consequently, this study concentrates on investigating consumers’ perceptions, rather than their final purchase decision, as perceptions can change, and have a critical role in the shopping process.

Findings of a research might not be generally applicable to other shopping contexts or customers, therefore it is advisable to study a topic from several different angles, as this allows researchers to explore the specific features of the different shopping contexts (Delgado-Ballester et al., 2014). Online and offline shopping environments significantly differ in case of brand loyalty, brand perceptions, and in the perceptions of the potential risks and benefits, which indicates that consumers tend to perceive and use these channels differently (Chu et al., 2010). However, most papers focused on whether the same consumers exhibit different shopping behaviour (and final purchase decision) in online/offline environments, but to the extent of this study’s author, there is little investigation on the perception differences between PLs and NBs in a purely online shopping environment. Furthermore, those literatures which examined the perceptions on private label products conducted their research in a physical store environment. Consequently, this research is placed in the online shopping environment, and focuses on the perceptions of private label products together with the chosen factors – brand tiers, product category – that can influence it, as an important research gap could be filled in this way.

(18)

Based on the identified gaps, the following research question is proposed for further studying:

‘How do consumers perceive ‘economy’, ‘standard’ and ‘premium’ tier, hedonic and utilitarian private label products compared to national brands while doing online grocery

shopping?’

2.7. Contribution to academic literature and managerial implications

Investigating the above mentioned research question will contribute to different areas of the academic literature. Research on consumers’ perceptions about private label products will add insights to the consumer behaviour literature (consumer perceptions), and to the branding and the retailing literature (private label products). By adding factors to the research which can influence the perceptions of PLs (brand tier and product category) this study provides a comprehensive overview on the chosen topic. However, placing the research into a new context (studying online grocery shopping platforms) is considered as the most impactful contribution to academic literature. This study is expected to shed light on consumers’ perceptions of private label products in an online grocery shopping environment and contribute to the existing knowledge as follows:

1. Insights on whether product placement in an online shopping platform has an effect on people’s brand perceptions and their choice decisions

2. Insights on what products should be offered in an online shopping platform – including product category and brand tier influence

This study and its findings has relevance for managers as well. Referring back to Ahold Delhaize’s chief e-commerce and innovation officer’s presentation (Strategy in a Digital Age), the food retail industry has not been disrupted yet, and the digital age and its new technologies are about to change people’s shopping habits (Faber, 2016). Food e-commerce is expected to grow, and more and more food retailers launch and continuously expand their online shopping platforms (Faber, 2016). As online grocery shopping is becoming part of people’s everyday life

(19)

and it is more common among retailers to set up their online platforms, conducting research on consumers’ perceptions by using this growing and expanding shopping context will offer valuable insights to managers. In general, this study helps managers to operate online shopping platforms more efficiently. More explicitly, the findings can offer valuable insight to managers on how to form effective online marketing strategies based on their goals. Whether they would like to increase the sales of PLs or NBs, the insights on product placement and on the online product portfolio offers them possible solutions on how to manage the online shopping platforms. Furthermore, the perception part of the study which focuses on consumers’ perceptions on private label products compared to national brands, can advise managers which product options should be offered in the online platforms, and under what circumstances do perception differences matter in case of PLs and NBs, and which approaches should be followed accordingly.

3. Conceptual framework

In this section the conceptual model, which is the backbone of the research design and approach, is introduced. Then the independent and dependent variables of the model, the two chosen moderators and their relationship to the variables are defined. Finally, based on the conceptual model, the formulated hypotheses and their theoretical background are presented.

3.1. Conceptual model

Based on the formulated research question, ‘How do consumers perceive ‘economy’,

‘standard’ and ‘premium’ tier, hedonic and utilitarian private label products compared to national brands while doing online grocery shopping?’, the following conceptual model is used

(20)

3.2. Independent variable

The study’s main question is focusing on consumers’ perceptions on PL products compared to national brands. In order to study this topic, the placement of private label products – placed in the middle or on the right of the row in a fictional online grocery shopping platform – is used as independent variable. The chosen independent variable is manipulated in the experiment by creating different images of a fictional online grocery shopping platform, where the order of the private label products (placed in the middle or on the right of the row) differs. The participants are asked to choose between the three (one private label and two nationally branded) available options and to justify their choice (i.e. indicate the main motivation) in order to control for any underlying effects.

3.3. Dependent variable

The perception of private label products compared to national brands served as

dependent variable in the experiment. In order to measure the aforementioned perceptions, the

last part of the survey focused on how participants perceive their initially chosen product compared to the other available options in an out of stock situation.

3.4. Moderators

Finally, the experiment uses two moderators: product category (hedonic and utilitarian) and brand tier (economy, standard and premium). As there are several differences between the

(21)

perception of hedonic and utilitarian products (e.g. people tend to buy utilitarian product via the Internet), and between the perception of the different brand tiers (e.g. standard PLs in general are seen as mid-quality alternatives), the chosen moderators is expected to add meaningful insights to the study (Okada, 2005; Geyskens et al., 2010). Similarly, to the independent variables, the moderators are manipulated during the creation of the images of the fictional online grocery shopping platforms.

3.5. Hypotheses

3.5.1. Brand perceptions and product placement (H1)

Standard private label products are perceived as good deals in general, and consumers tend to consider and choose PLs when they intend to save money on groceries (Garretson et al., 2002). However, this assumption that PLs are more associated with product value (good deal for your money), while national brands are more about the absolute quality perceptions is not entirely valid nowadays (Garretson et al., 2002). Private label products have become more and more accepted by consumers over time (Steenkamp et al., 2010). Considering that PLs tend to mature, their sales is rapidly growing and that blind taste tests proved that there are little or no quality differences between private label products and nationally branded ones, the perceptions of PLs are shifting to ‘as good as national brands’ (Rossi et al., 2015, p. 74). Although different factors, such as product brand and the environment within the decisions are made, can still have a significant impact on people’s perceptions and eventually on their final purchase decisions (Rossi et al., 2015; Richardson et al., 1996).

As the offline environment has a significant (positive or negative) effect on consumer perceptions and decisions (Richardson et al.,1996), despite the fact that there are several differences between the online and offline shopping environments (e.g. different benefits and drawbacks), perception differences between products can extend to the online shopping environment as well and influence consumer response (Chu et al, 2010).

(22)

Finally, another important factor which can have a significant effect on product and brand perceptions besides the different environments, is the placement of the products (Sigurdsson et al., 2009). Researchers found evidence in a field experiment that not just in-store product placement in general (Wong et al., 2015), but the placement of products on shelves (placed in the middle or on the left/right side) is also a very important underlying factor of consumers’ buying behavior (Sigurdsson et al., 2009). A recent article further supports the effect of within shelf placement on consumers’ perceptions and buying behavior, by stating that those products which are placed at the start of an aisle do not sell well, as people need time to adjust to their environment, and consequently middle placement on shelves is a more preferred option (Kendall, 2014).

Despite the fact that online and offline shopping environments differ from each other in several aspects, researchers agree that the design of online shopping platforms is ‘the virtual equivalent of retail atmospherics’ (Wolfinbarger & Gilly, 2002, p. 25). Based on the aforementioned structural similarities, one can argue that the rows in an online shopping platform and the shelves in an in-store environment not just look and feel, but can also fulfil similar purposes. Following this line of reasoning suggests that the significant effect of within shelf product placement on perceptions and sales in an offline shopping environment can be extended to the online environment as well.

However, the effect of product placement on consumers’ perceptions might not be equally strong in case of private label and nationally branded products. Despite the fact that private label products can mature and become more accepted over time (Steenkamp et al., 2010), consumers in general tend to be loyal to their favorite brands and they look for them in any store where they shop, which indicates that brand name has a key role in shaping people’s perceptions and shopping behavior towards NBs (Garretson et al., 2002). Because of the aforementioned importance of brand name in case of nationally branded products, one can argue

(23)

that other factors, such as the placement of the products, have a somewhat stronger or weaker effect on consumers’ perceptions in case of different products. Therefore, product placement in case of private label products is expected to have a slightly stronger effect on people’s perceptions, due to the lack of brand name importance.

Consequently, based on the above mentioned reasoning, which includes the importance of product placement on perceptions in an offline environment, the structural similarities between online shopping platforms (rows) and offline stores (shelves), and the possible differences of the strength of product placement on consumers’ perceptions in case of private label and nationally branded products (due to brand importance in case of NBs), the following hypothesis is proposed:

H1. When PL products are placed in the middle of the row in an online shopping platform, consumers’ perceptions on them are more favorable, than when they are placed on the right of the row.

3.5.2. Utilitarian and hedonic products (H2)

In case of the upcoming two hypotheses, the focus is on the moderating effect of product category and brand tier. In other words, the main attention is on under which circumstances is the strength of the relationship between product placement and consumers’ perceptions in case of private label products increase or decrease. As standard PL products in general are still positioned (e.g. priced below non-price promoted nationally branded goods) and in several cases also seen as mid-quality alternatives (Geyskens et al., 2010; Garretson et al., 2002), it is expected that consumers will prefer utilitarian (primarily instrumental and functional) private label products over hedonic (more experiential consumption, fun, pleasure and excitement) ones (Wertenbroch & Dhar, 2000). A recent study emphasized that due to the reasonable price, consumers acquire utilitarian benefits when buying a private label product, which provides

(24)

further support that PLs are perceived as utilitarian products in nature (İpek et al., 2016). Moreover, when consumers seek for hedonic products they tend to spend time on discovering the different alternatives in stores, but when they buy utilitarian products they are more likely to purchase them on the Internet (Okada, 2005). Therefore, the following hypothesis is proposed:

H2. When PL products are placed in the middle of the row in an online shopping platform, consumers’ perceptions on them are more favorable, than when they are placed on the right of the row, but only in case of utilitarian products.

3.5.3. Private label brand tiers (H3a, H3b)

Similarly, to the product category, the moderating effect was in the focus while formulating the final two hypotheses. Several retailers embraced the three-tier (economy, standard and premium) private label portfolio, in order to sufficiently compete with nationally branded products (Geyskens et al., 2010; ter Braak et al., 2014). The lowest tier (economy) represents the inexpensive offers, and due to their favorable price and moderate quality, these product options found to cannibalize the higher-margin standard private label ones (Geyskens et al., 2010). As products under the economy tier possess a relevant strength thanks to their affordable price, compared to the other tiers, one can argue that product placement might have a stronger effect on these products by strengthening their advantageous position, and making them even more appealing. Based on this line of reasoning, the following hypothesis is formulated to further investigation:

H3a. When PL products are placed in the middle of the row in an online shopping platform, consumers’ perceptions on them are more favorable, than when they are placed on the right of the row, but only in case of products under the economy brand tier.

(25)

Products under the premium tier are also differ in some aspects from the ones under the economy and standard tiers. These products are ‘at the top end of the market and deliver higher quality’, and they also offer unique features, for instance in the form of flavors, ingredients, or unique packaging (ter Braak et al., 2013, p. 89). Because of their higher quality and exclusivity premium PL products stand out from the other private label offerings, increase consumers’ willingness to pay, and enables them to compete with even the higher quality NBs (ter Braak et al., 2014). Following the thinking described above, one can argue that product placement might have a stronger effect on premium tier PL products, by increasing their already positive perceptions in the eyes of the consumers. Consequently, I propose the following hypothesis:

H3b. When PL products are placed in the middle of the row in an online shopping platform, consumers’ perceptions on them are more favorable, than when they are placed on the right of the row, but only in case of products under the premium brand tier.

4. Research design and method

In order to investigate consumers’ perceptions on private label products compared to national brands while doing online grocery shopping, an online experiment was conducted. Qualtrics was used to randomly allocate participants to the different conditions. The experiment applied a mixed design, where participants were assigned to two conditions: one hedonic and one utilitarian. In other words, each participant saw two pictures of a fictional online shopping platform during the experiment: one with utilitarian and one with hedonic products. Consequently, a 2x2x3 (PL product placed in the middle of the row/PL product placed on the right of the row; utilitarian/hedonic product category; economy/standard/premium brand tier) mixed experiment design was used.

(26)

4.1. Survey and experiment design

4.1.1. Experiment design and conditions

The design process started with the creation of a fictional online shopping platform. As the aim was to create a simple, well-designed shopping platform, Albert Heijn’s online platform was chosen as inspiration for the design. The basic features of Albert Heijn’s shopping platform were included into the design, for instance the search field, the different special options, and the layout in order to keep the ‘feeling of a real’ online grocery shopping platform. At the same, the design was simplified by removing those options (e.g. recipes) which were not essential to create the illusion that the subjects are looking at an existing shopping platform. The Albert Heijn logo was also excluded from the design, and ah.nl as a source was not added eventually at the bottom of the fictional platform, in order to avoid biasing the subjects (i.e. positive or negative associations towards Albert Heijn) during the experiment. Taking into consideration all the above mentioned factors and features, the fictional online shopping platform was created in Photoshop. Product placement (i.e. independent variable) was included into the design by manipulating the position of the private label products. As the reviewed literature suggests that consumers’ perception in an in-store environment differed when products were placed in the middle or in one of the two sides of the shelves (consequently, the left and right sides were considered as equal), product placement was manipulated in the experiment as follows: one row was added to the fictional shopping platform with three product options (one private label and two nationally branded ones), and the PL products were placed either in the middle of the row or on the right of the row. In other words, besides placing a PL product in the middle of the row, which had a significant effect on consumers’ perception in an offline environment, as the two side of the row is considered equal, the right side was chosen for the manipulation. Figure 1 presents the look and feel of the completed fictional online shopping platform. 1

1_{Although simplified product placement manipulation approach was used in the experiment, it is important to note that product placement is very}

(27)

Figure 1: Look and feel of the completed fictional online shopping platform

Different private label and nationally branded toilet papers were chosen as utilitarian product alternatives in the experiment. Based on research on Albert Heijn’s online grocery shopping platform (ah.nl), three Albert Heijn private label toilet papers from different brand tiers (economy tier – ‘AH Zacht & sterk toiletpapier’, standard tier – ‘AH Extra zacht toiletpapier’ and premium tier – ‘AH Wellness toiletpapier’), and two mainstream-quality nationally branded toilet papers (‘Edet family comfort toiletpapier’ and ‘Page Origineel Schoon toiletpapier’) were chosen. The price and the quantity (rolls) of some products was adjusted in order not to bias the participants. Consequently, all product options offered 6 rolls of toilet papers and the prices were calculated based on the price of one roll. The brand name and the product specification (i.e. number of layers) were the only differentiating factors.

In case of the hedonic conditions, private label and nationally branded chocolates were chosen as different product alternatives in the experiment. Similarly, to the utilitarian conditions, research was conducted on Albert Heijn’s online grocery shopping platform (ah.nl) in order to choose three Albert Heijn private label chocolates from different brand tiers. However, as there were not any product options, which could fulfil the requirements of a

(28)

premium tier one, an alternative option was used. Organic products are positioned higher than ‘average’ products, as consumers’ tend to perceive their quality and beneficial effects more positively, and therefore they are willing to pay a higher price for these products (Bezawada & Pauwels, 2013). Following this line of reasoning, the scope of the premium brand tier in case of hedonic products was extended to the organic ones as well. Consequently, the following products were chosen for the experiment from the different brand tiers: economy tier – ‘AH Baic Tablet Melk Chocolade’, standard tier – ‘AH Delicata Reep Melk’ and premium tier – ‘AH Biologische Melkchocolade Tablet’. Besides the private label products, two mainstream-quality nationally branded chocolates (‘Milka Tablet Alpenmelk’ and ‘Côte d’Or Tablet Melk’) were added to the chosen product. The price and quantity adjustments were implemented in the same way as in the utilitarian conditions. Therefore, respondents could choose from product options offering 100 g chocolates, and the brand name and the product specification (e.g. Alpenmelk) were the only differentiating factors.

As mentioned above, in the experiment the position of the private label products – placed in the middle of the row or placed on the right of the row – was manipulated. Besides the placement of the private label product, the two moderators (product category and brand tier) were also considered during the preparation of the overall twelve conditions. Table 1 presents a detailed overview on all experiment conditions.

(29)

4.1.2. Survey design

The survey related to the experiment contained four questions, each was a multiple choice one. The questions did not vary neither per condition, nor per product category. As such, only the product choice options differed per condition. Each question measured different insights, as follows:

1. Purchase decision

2. Main motivation behind purchase decision 3. Importance of brand name

4. Product perceptions

The first question asked the participants to choose one product from the three available options (one PL and two NBs). The main goal of this initial question was to capture the respondents’ initial perceptions on the products and their purchase behaviour. Furthermore, it is closely connected to the dependent variable, which aims to capture participants’ perception (differences) on private label and nationally branded products. To do so, the final question asked the respondents to indicate on a 5-point scale how they perceived their product choice compared to the other options in an out-of-stock situation, when their chosen product was not available. In order to further stimulate participants’ perceptions, the picture with the available product options appeared again, but at this time, the respondents’ original choice was grey, and an ‘Out of Stock’ sign appeared next to the products name.

The other two questions, the motivation behind the purchase decision and the importance of the brand name, were acted as control variables, to identify any underlying effects.

4.2. Experiment procedure

First, the participants were shown an image of a fictional online grocery shopping platform (one of the total six conditions) with three utilitarian products – one private label and two nationally branded ones. After showing them the first picture, participants were asked to

(30)

choose between the three possible options and to justify their decision, by stating the most important underlying motivation of their choice: 1=’Price’; 2= ‘Brand’; 3= ‘Value for money’; 4= ‘Previous purchase’; 5= ‘Other’. Then the importance of brand name was tested by asking the following question: ‘In this product choice situation how important is the brand name of the product to you? – 0= ‘Not important at all’; 1= ‘Of little importance’; 2= ‘Of average importance’; 3= ‘Very important’; 4= ‘Absolutely essential’. In order to test participants’ product perceptions, the picture of the fictional online shopping platform with the three utilitarian products was shown to them again, but in this case their original choice became grey, and an ‘Out of Stock’ sign appeared next to the chosen product’s picture. The last question of the utilitarian condition asked how do respondents consider their originally chosen product compared to the currently available options, by indicating their answer on a 5-point scale, where 1= ‘Much worse choice’; 2= ‘Somewhat worse choice’; 3= ‘Equally good choice’; 4= ‘Somewhat better choice’; and 5= ‘Much better choice’.

After completing the utilitarian conditions, participants were randomly assigned to one of the total six hedonic conditions. The questions in case of the hedonic conditions and their order were the same as in the utilitarian conditions, and the experiment was conducted in the exact same way – all steps were carried out – as in case of the utilitarian conditions.

4.3. Participants and data collection

Online channels, such as student e-mail, social media platforms, were the only channels used for the distribution of the surveys. The distribution process contained three stages: 11th April – first launch of the survey on Facebook (including own wall, university groups); 12th April – distribution of the survey via student mail; 20th April – reminder/second launch of the survey on Facebook (university groups). In order to reach the high number of respondents (over 200), the duration of the survey was three weeks, with a closing date on 2nd_{May. The conducted}

(31)

undergraduate students, between age 20-25. Although, due to anonymity reasons no demographical data were collected from the respondents (e.g. nationality, age, gender), therefore, the age and status of the participants cannot be reported. Regarding the number of respondents, the identified 47 partial responses were excluded from the overall dataset, leaving 215 valid answers for the analysis.

5. Data Analysis

The upcoming section focuses on the analysis of the gathered data. The first part includes general information on the dataset and introduces the steps prior to the analysis (i.e. data preparation). Then the chosen model and the applied processes to test each hypothesis are briefly discussed. Finally, the reporting on the results of the analysis and the discussion on whether the previously proposed hypotheses are supported or rejected closes the section.

5.1. Preparation of the dataset

The Qualtrics online survey tool recorded the answers of the participants for three consecutive weeks, and it closed the survey with 262 responses. As a first step, those variables (e.g. IP address, recorded date, location latitude etc.) were removed from the dataset which had no added value to the analysis. The “clean” dataset contained the following variables: progress, finished, responseid, and all survey questions for the total 12 conditions. Then the aforementioned dataset was used to create two separate ones: the utilitarian (including data for conditions 1-6), and the hedonic (including data for conditions 7-12) one. The combined version of the two separately prepared and formatted datasets was used during the analysis.

The formatting of the utilitarian dataset was started with the recoding of the variables (utilitarian=1) in order to have one variable per question for all conditions as opposed to per condition. Looking into the frequencies showed that the number of the valid responses were 215, and the number of the partial responses were 47 in the utilitarian conditions. As a valid response must have answers for all questions, the 47 partial or invalid responses were excluded

(32)

from the dataset and therefore, from the analysis as well. The recoding of the variables continued as follows: product placement (right=1, middle=0), brand tier (economic=1 – c1, c2; standard=0 – c3, c4; premium=1 – c5, c6), question 1 c6)=Purchase_dec., question 2 (c1-c6)=Motivation, question 3 (c1-c6)=Brand name_1, question 4 (c1-c6)=Out_of_stock_1/2/3 (depending on the respondents’ original purchase decision). After finishing the recoding, all the unnecessary question related data were removed from the utilitarian dataset and the above mentioned recoded variables were used during the analysis.

The formatting and the recoding of the variables (hedonic=0) of the second, the hedonic dataset followed the exact same steps as in case of the utilitarian one described above. The hedonic conditions were ranged from 7 to 12, consequently the brand tier variable was recorded as follows: economic=1 – c7, c8; standard=0 – c9, c10; premium=1 – c11, c12. 2

The utilitarian and the hedonic datasets were combined into one overall dataset, which contained all relevant data and was used for the analysis. As all participants were assigned to two conditions, one utilitarian and one hedonic, frequency analysis was run on the responseid variable in order to see whether the IDs match. The results showed that the frequency equals 2 in case of all 262 response IDs, therefore it is confirmed that all participants were assigned to two conditions regardless of whether they completed the survey or not.

As the last part of the dataset preparation for the analysis, dummy variables were created. First, a dummy variable which determined those conditions where the participants chose a nationally branded product over a private label one was created. Then one out of stock variable was created, which did not depend on the selected product. As a next step, the dependent variable (i.e. out_stock), a dummy which takes the value of 1 if NB was chosen and 0 if PL was chosen, was created. Finally, the categories were recoded into dummy variables in

(33)

order to prepare all independent variables: motivation (category variable) was recoded into n-1 dummies, but brand was already an ordinal variable which did not need any recoding.

5.2. Methodology

The analysis of this study is based on a linear mixed model and more specifically a random intercept model, whereby in addition to the fixed effects of product category, product placement, brand importance, purchase motivation and brand tier, random individual effects are also included. Such a model is used to control for individual specific effects as each individual was assigned to two conditions, and therefore, the standard errors are correlated for the answers of the same individual. In this model it is assumed that there were some baseline differences among individuals with regards to their perceptions, but the above mentioned effects, such as product type, product placement, brand tier etc., are the same for each individual (i.e. the slope is the same but the intercept varies by individual). A random intercept model seems appropriate, as it is safe to assume that the individuals are a random sample from an overall population.

The same linear mixed model, including the main fixed effects of product category, product placement, brand importance, purchase motivation and brand tier; random individual effects; and interaction effects between product category (utilitarian or hedonic) and product placement, and between brand tier (H3a – economy; H3b – premium) and product placement was used to test all hypotheses.

The analysis restricts the sample to only those cases/responses in which private label products were chosen originally, as all hypotheses refer to the perception of PLs. Therefore, the outcome variable has to refer to the reaction when the initially chosen private label product was unavailable, and those responses in which NBs were chosen do not provide the adequate information.

(34)

The (base) linear mixed model used in the analysis takes the following form: 𝑷𝒆𝒓𝒄𝒆𝒑𝒕𝒊𝒐𝒏𝒊𝒋= (𝒊𝒏𝒕𝒆𝒓𝒄𝒆𝒑𝒕 + 𝒊𝒏𝒅𝒊𝒗𝒊𝒅𝒖𝒂𝒍 𝒆𝒇𝒇𝒆𝒄𝒕𝒔𝒊) + 𝒖𝒕𝒊𝒍𝒊𝒕𝒂𝒓𝒊𝒂𝒏𝒊𝒋+ 𝒓𝒊𝒈𝒉𝒕𝒊𝒋+ 𝒆𝒄𝒐𝒏𝒐𝒎𝒊𝒄𝒊𝒋 + 𝒑𝒓𝒆𝒎𝒊𝒖𝒎𝒊𝒋+ 𝒎𝒐𝒕𝒊𝒗𝒂𝒕𝒊𝒐𝒏 (𝒑𝒓𝒊𝒄𝒆)𝒊𝒋+ 𝒎𝒐𝒕𝒊𝒗𝒂𝒕𝒊𝒐𝒏 (𝒃𝒓𝒂𝒏𝒅)𝒊𝒋 + 𝒎𝒐𝒕𝒊𝒗𝒂𝒕𝒊𝒐𝒏 (𝒗𝒂𝒍𝒖𝒆 𝒇𝒐𝒓 𝒎𝒐𝒏𝒆𝒚)𝒊𝒋+ 𝒎𝒐𝒕𝒊𝒗𝒂𝒕𝒊𝒐𝒏 (𝒑𝒓𝒆𝒗𝒊𝒐𝒖𝒔 𝒑𝒖𝒓𝒄𝒉𝒂𝒔𝒆)𝒊𝒋 + 𝒃𝒓𝒂𝒏𝒅𝒊𝒋+ 𝒖𝒕𝒊𝒍𝒊𝒕𝒂𝒓𝒊𝒂𝒏𝒊𝒋 𝐗 𝒓𝒊𝒈𝒉𝒕𝒊𝒋+ 𝒓𝒊𝒈𝒉𝒕𝒊𝒋 𝐗 𝒆𝒄𝒐𝒏𝒐𝒎𝒊𝒄𝒊𝒋 + 𝒓𝒊𝒈𝒉𝒕𝒊𝒋 𝐗 𝒑𝒓𝒆𝒎𝒊𝒖𝒎𝒊𝒋

The dependent variable (i.e. out_stock) is the perception of PL products, for each respondent i and scenario j (which takes up two levels, as each individual was assigned to two scenarios). The included intercept is the same for every respondent, and the individual random effects only vary per individual, and not by scenario. The independent variables are the following: a utilitarian dummy (which equals 1 if the product is utilitarian, and 0 if it is hedonic); a product placement dummy (i.e. rightij; which equals 1 if the PL product is placed

on the right of the row, and 0 if it is in the middle of the row); two brand tier dummies (i.e. economic and premium; and the third tier, standard, is acting as the reference category); four purchase motivation dummies (i.e. price, brand, value for money and previous purchase; where other, the fifth possible motivation, is acting as the reference category); and a brand importance dummy (which is an ordinal variable with five levels – not important at all, of little importance, of average importance, very important and absolutely essential). Finally, the following interaction effects were added to the model: interaction between the type of products (utilitarian or hedonic) and product placement (i.e. utilitarian*right); interaction between the economy brand tier and product placement (i.e. right*economic); and interaction between the premium brand tier and product placement (i.e. right*premium).

5.3. Results – Hypotheses testing

As all hypotheses refer to the perception of private label products, the analysis for testing them restricts the sample to only those cases/responses in which PLs were chosen originally

(35)

(N=52). The exact same model, which was introduced in the methodology section, was used on the PL subsample. As mentioned above, all main effects and interactions were estimated in the same model.

Table 2 summarizes the results of the analysis. The first two columns present the results of the analysis with the linear mixed model and the last two from a linear model without the random intercepts. The findings indicate, that the random effects are significant (as the coefficient on the random effect intercept is significant), and therefore they need to be included in the model. Furthermore, the Akaike’s Information Criterion (AIC) and the Schwarz’s Bayesian Criterion (BIC) values, which show the goodness of fit in case of the two models, are lower for the mixed linear model, and thus confirm that the linear mixed model is the more suitable option for the analysis. Overall, though, as the results in Table 2 shows, both the random intercept model and the linear model lead to the same conclusions with regards to the insignificance of the effects of the independent variables on the perceptions.

Table 2: Results of linear mixed model and linear model, with out_stock as dependent variable

Variables

Model 1 – random intercept model Model 2 – linear model with fixed effects only β t β t Intercept 3,392*** 8,617 3,320*** 6,517 Right -,303 -,392 -,558 -,693 Utilitarian -,489 -1,320 -,159 -,361 Economic -,021 -,071 ,138 ,436 Premium ,522 1,163 ,042 ,079 Motivation 1 ,233 ,352 1,680 1,475 Motivation 2 ,844 1,169 -1,258 -,974 Motivation 3 -,181 -,272 -1,442 -1,272 Motivation 4 ,288 ,381 -1,095 -,941 Brand -,154 -1,301 -,081 -,653 Utilitarian*right ,351 ,664 ,046 ,072 Right*economic -,175 -,394 -,337 -,755 Right*premium -,054 ,064 ,742 ,846 AIC 309,294 312,596 BIC 348,221 348,928

Random effect intercept ,920***

(36)

5.3.1. Hypothesis 1

The first hypothesis (H1) tested whether the position of PL products (placed in the middle or on the right of the row) matters for perceptions. Results indicate that, there is no significant relationship between PL product placement (middle of the row or right of the row) and consumers’ perceptions on PL products (all p> .10). Due to the lack of the aforementioned significance between the DV and the IV, H1 was rejected.

The moderating effect of utilitarian products on the position of PL products (placed in the middle or on the right of the row) was tested in case of the second hypothesis (H2). In order to test the second hypothesis, the interaction between the type of products (utilitarian or hedonic) and product placement was included in the model (i.e. utilitarian*right). Findings indicate, that utilitarian products do not have a significant moderating effect on PL product placement (middle of the row or right of the row) (all p> .10). As no significant moderating effect was found between the DV and the first moderator (product category), H2 was rejected.

The third hypothesis tested the moderating effect of brand tier on the position of PL products (placed in the middle or on the right of the row), and included two parts: H3a tested the moderating effect in case of products under the economy brand tier, while H3b tested the same in case of premium ones. In order to test the last two hypotheses, an interaction between brand tier (H3a – economy, H3b – premium) and product placement was added to the model (i.e. right*economic; right*premium).

First, H3a was tested in order to investigate whether there are any moderating effects between products under the economy brand tier and PL product placement (middle of the row or right of the row). Results show, that the lowest brand tier (economy) does not have a

(37)

significant moderating effect on PL product placement (middle of the row or right of the row) (all p> .10). Because of the lack of significant moderating effects, H3a was rejected.

Then, H3b tested the moderating effect between products under the premium brand tier and PL product placement (middle of the row or right of the row). The results did not indicate any significant moderating effect between the highest brand tier (premium) and PL product placement (middle of the row or right of the row) (all p> .10). Due to the fact that there is a lack of significant moderating effects, H3b was rejected.

5.4. Results – Additional analysis

Testing all hypotheses based on the restricted sample (only in cases when respondents chose PL products) did not show any significant relationship or moderating effects on the dependent variable. However, one can argue whether there are any possible underlying effects in case of nationally branded products. In order to get a comprehensive overview on the results, follow-up analyses have been run. First, in order to see whether product placement in general has a significant effect on people’s perceptions of products or not, the analysis was performed on the overall data set (including PLs and NBs, N=215). Then, to test whether product placement has any significant effects on consumers’ perceptions in case of nationally branded products (N=163) another analysis was performed. The model for the extended analysis was the exact same one which was used for the hypothesis testing, only the (sub)sample – first the two subsamples together (overall data), then only those responses where participants initially chose NBs – differed.

5.4.1. Follow-up analysis on the overall data

The overall data (PL and NB subsamples together) was tested for H1. Results indicate that, there is no significant relationship between product placement (middle of the row or right of the row) and consumers’ perceptions on products (all p> .10) in general. Therefore, none of