Does consumer promotion relevance matter? : a cross FMCG category study

(1)

Does consumer promotion relevance matter?

A cross FMCG category study

J.M. van de Sande (10476016)

Amsterdam Business School (UvA)

Executive Programme in Management Studies Specialization: Marketing Management Supervisor: dr. K. Venetis

Second reader: dr. U. Konus Amsterdam, February 2015

(2)

Abstract

This study focuses on the measurement of the importance of promotions for consumer decision making, that is, promotion relevance in category, or PRIC-across multiple categories in the Dutch groceries market. Although sales promotions have attracted a large body of literature, especially regarding their effects in terms of sales, brand equity and reference prices, the question of how important promotions are for consumers within an product category and the extent to which PRIC differs across categories have been neglected.

We introduce the concept of PRIC. We develop a conceptual framework to validate the PRIC measure with the six promotion benefits of Chandon et al. (2000) as antecedents of PRIC and the three most reported immediate promotional effects as consequences of PRIC. We test our framework empirically with a sample of more than 1.000 consumers, and show how the construct varies across 10 product categories in the Dutch groceries market. Following literature, we research different category characteristics and consumer demographics which might explain differences in PRIC.

First, our results suggest a high validity of the proposed PRIC measure. Secondly, the PRIC-scale shows a stronger positive relationship towards promotional behavioral effects, like store- and brand switching than other promotion- and price constructs. Finally, we report substantial differences in PRIC scores between product categories and product category characteristics. Between the several consumer demographics we find very limited differences. In addition, we report a huge increase of the General deal proneness among Dutch consumers, which indicates a still continuing focus on sales promotions and consequently a lower loyalty both for brands and stores.

These findings can be of significant help to managers in making better investment decisions and increase the understanding of promotions in general and specifically of differences in promotional responses between product categories.

(3)

1. Introduction

Firms annually spend hundreds of billions of euros to implement their marketing strategy (Ataman et al., 2010) using different instruments of the marketing mix. Sales promotions are one of these instruments and have become a vital tool for marketers and its importance has been increasing significantly over the years. Figures 1a and 1b show the volume sales development, both total volume sales and volume sales on promotion, for several Western countries. In general, the total food and non-food volume sales are in decline at a total European level and a small increase in promotional activity in Europe which is higher in the non-food category, except for Spain and Germany (IrI, 2014).

Figure 1a and 1b. Volume sales evolution across countries (1a) and volume sales on promotion evolution across countries (1b). Source: IrI, July 2014

Figure 1c. Share of promotional volume in the Netherlands per product category (IrI, 2014)

The Netherlands show a high increase of promotional volumes and the share of sales on promotions increase for almost all product categories (figure 1c). For the Netherlands IrI (2014) stated: “The level of promotion has reached its highest level in 10 years but volume

(7)

sales have declined 2.9%, further proof that promotions don’t always have the desired impact.”

This statement of IrI contains an essential question regarding promotions: Do they have the desired impact? Ataman et al. (2010) researched the long term effects of the integrated marketing mix on sales and found disappointing total results (long plus short term) for sales promotions and advertising. In contrast, product (number of SKUs) and place (weighted distribution) showed much better total results. Nijs et al. (2001) found that sales promotions didn’t have any impact in the long run in 98% of the 560 researched product categories in the Dutch Supermarkets, although the short term effects are generally strong. Besides these sales effects, literature also shows that sales promotions have negative effects on brand equity (Yoo et al., 2000; Valette Florence et al., 2011; Buil et al., 2013), impact the formation of reference Prices (Lattin & Bucklin, 1989; Kalwani et al., 1990; Kalwani and Yim, 1992; Mayhew and Winer, 1992; Han et al., 2001) and expected future Prices (Zhang et al., 2012) which has effect on consumers’ purchase decisions.

Given the growing importance of sales promotions, its questionable total results on sales, its negative effects on brand equity and its indirect impact on consumers’ purchase decision (via reference prices), a question that needs to be answered is the relevance question: how important are sales promotions for consumers? Most studies regarding sales promotions are based on consumer scanner data and store-level scanner data (Gedenk et al., 2006) to research sales effects like purchase acceleration, brand switching and store switching. These studies measure actual behaviour but don’t measure consumers’ perception and relevance of sales promotions.

Several studies found differences in consumer responses to promotions between categories (Raju, 1992; Narsimhan et al., 1996; Bell et al., 1999; Chandon et al., 2000; Luijten, 2012). Bell et al. (1999) concluded that category factors are especially powerful in explaining variability compared to brand and consumers factors. Variables researched in these studies are: category penetration, storability, ability to stockpile, price, share of budget, expensiveness of category, competitive intensity, bulkiness of category, magnitude of discounts, deal depth, brand assortment, number of brands, necessity, purchase frequency, interpurchase times, taste sensation, consumption pleasure, frequency of discounts, deal frequency, perceived differentiation, impulse buying, private label share and assortment size. So, to research the promotion relevance for consumers, the product category is the main subject of this study to research sales promotions relevance, as it is recognized as a key driver of promotional response (Bell et al., 1999).

Study

In this study, we introduce a construct called “promotion relevance in category” (PRIC), which measures the overall role of sales promotions in customers’ decision making in a specific category. Under the assumption that the sales promotions offer additional benefits to the consumer, PRIC can be defined as the extent to which sales promotions influences customer decision making relative to other decision criteria in a particular product category.

(8)

As mentioned, we expect product category differences based on different variables derived from literature (Raju, 1992; Narsimhan et al., 1996; Bell et al., 1999; Chandon et al., 2000; Luijten, 2012).

This study makes several contributions to the literature. First, we introduce the concept of PRIC. Secondly, we develop and test a new scale to measure PRIC. We apply the scale to 10 product categories covering fast-moving consumer goods (FMCG) in the groceries sector in the Netherlands involving more than 1.000 consumers. The results offer insights into the differences of PRIC across categories. Finally, we study several consumer-specific and category-specific characteristics to analyse the observed differences and its consequences. We organize the remainder of this study as follows. We start the next chapter by a short introduction of PRIC, followed by a description of the context of PRIC. We continue chapter two by discussing the relevant literature concerning sales promotions. In paragraph 2.6 we discuss several existing price- and promotional constructs which are related to PRIC. The model of Churchill (1979) functions as the theoretical basis of the scale development and validation process of the PRIC-measure and is the main subject of paragraph 2.7. In addition we explain how we apply this model in our study. In the next paragraph we present the hypotheses which are part of this study. We end with a short summary of chapter two.

We continue with the data collection procedure and research method in chapter three. Then in chapter four, we describe the scale development and validation process of the PRIC scale. Afterwards, in the results section (chapter five) we discuss the findings of this study with respect to our hypotheses.

Finally, in chapter six, the most important conclusions and implications of the results of this study are discussed, supplemented with the most important limitations and a number of suggestions for further research.

(9)

2. Literature review

After a short introduction of the PRIC construct in paragraph one, we discuss the relevant literature regarding the PRIC construct. We start by describing the context of the PRIC construct in paragraph two. First, by giving a short overview of the definitions of sales promotions and secondly by describing the most important findings in literature concerning the strategic and permanent effects of sales promotions. These, more long term effects are discussed to show the broader context of this study and to strengthen the relevance question as raised in the previous chapter.

In paragraph two of this chapter we discuss the antecedents of the PRIC construct, which were derived from the sales promotions framework of Chandon et al. (2000). Afterwards we discuss the immediate sales promotions effects, which are heavily researched, especially by using panel and sales data. These, immediate effects are used as consequences of PRIC and are part of the validation model of the PRIC-measure, which will be explained in paragraph seven. In the next paragraph, the two groups contingency factors of the PRIC construct are discussed. The first and most important group in this study, are the product category characteristics (2.5.1 and 2.5.2) and the second group are the consumer demographics (2.5.3). In paragraph 2.6, we discuss four important and well known promotion- and price concepts, which are close related to PRIC and are also part of this study.

Finally, in the last paragraph we explain how the promotion benefit framework of Chandon et al. (2000), the immediate promotion consequences and the four price- and promotion constructs fit together in the validation model of the PRIC construct. Furthermore, we address the most important research questions and hypotheses of this study, which are driven by the contingency factors of PRIC (product category characteristics and consumer demographics).

2.1 The PRIC construct

Promotion relevance in category (PRIC) is a customer-oriented construct that measures the overall importance of promotions in customer decision making. Thus, it focuses on the category, not on promotions in general or individual promotions. For a given category, we define PRIC as the extent to which promotions influences customer decision making relative to other decision criteria e.g. purchase convenience and the brand.

Because PRIC does not measure psychographic characteristics, its concept is distinct from the meaning of widely used price- and promotions constructs, such as price sensitivity and general deal proneness (see paragraph 2.6). Because these constructs focus on consumer psychographics in general, they do not provide information on the relevance of promotions in consumer decision making on a category level.

(10)

Finally, regarding the construct itself, we emphasize that the proposed PRIC-metric is defined at the category level which implies it does not vary across brands and individual products but only across categories.

2.2 The context of PRIC

In this study we research the relevance of sales promotions, so we start by defining them in the first paragraph. We continue by discussing the findings in literature regarding the strategic and permanent effect of sales promotions, in order to show the broader context of this study and to strengthen the importance of the relevance question as represented by PRIC.

2.2.1 Sales promotions: definitions and scope of this study

Sales promotions are a marketing tool for manufacturers as well as for retailers Manufacturers use them to increase sales to retailers: trade promotions and consumers: consumer promotions (Gedenk et al., 2006). As discussed in chapter one sales promotions play an important role in the marketing programs of retailers. A large percentage of retailer sales are made on promotion, as illustrated by the numbers in figure 1c. In this paragraph we define sales promotions, promotion instruments and several promotion related concepts and explain the scope and definitions of this study.

Definitions of American Marketing Association

We start by the definitions which were used by Luijten (2012) in his use of related concepts which were derived from the American Marketing Association1.

Sales Promotion

The media and non-media marketing pressure applied for a predetermined, limited period of time at the level of consumer, retailer, or wholesaler in order to stimulate trial, increase consumer demand, or improve product availability.

Consumer sales promotion

Externally directed incentives offered to the ultimate consumer. These usually consist of offers such as coupons, premiums, rebates, etc., designed to gain one or more of the following: product trial; repeat usage of product; more frequent or multiple product purchases; introduce a new/improved product; introduce new packaging or different size packages; neutralize competitive advertising or sales promotions; capitalize on seasonal, geographic, or special events; encourage consumers to trade up to a larger size, more profitable line, or another product in the line.

Price promotion

The advertising of a price for a product or service. Usually, the price being promoted is a reduction from a previously

1

(11)

established price and may take the form of a lower price, a coupon to be redeemed, or a rebate to be received.

Promotional advertising

Advertising intended to inform prospective customers of special sales. It announces the arrival of new and seasonal goods, and it features, creates, and promotes a market for the merchandise items in regular stock.

Table l. Sales promotions and related concepts (American Marketing Association)

Promotion instruments

Different promotion instruments exist. A first distinction can be made between price and non-price promotions (Gedenk et al., 2006) . The non-price promotion instrument used most often is a temporary price reduction (TPR). However, other forms of price promotions are possible. Retailers can use promotion packs or multi-item promotions. Loyalty discounts also require the purchase of several units, but the consumer can do this over several purchase occasions. Retailers can also use coupons or rebates. With coupons, consumers have to bring the coupon to the store in order to get a discount (Gedenk et al., 2006).

“Supportive” non-price promotions are communication instruments used to alert the consumer to the product or to other promotion instruments (Gedenk et al., 2006). Very often they are used to draw attention to price promotions. For example, products on TPR are featured or displayed. Thus, the focus is not so much on the brand as on price. Note that they can also be used without a price promotion. For example, a feature can advertise an everyday low price policy or a new product (Gedenk et al., 2006).

Finally, retailers can use “true” non-price promotions, where the focus of the promotion is clearly on a brand or store, and not on a price cut. However, instruments such as sampling and premiums are mostly used by manufacturers, and not by retailers (Gedenk et al., 2006).

Scope of this study

To keep the discussion centred on a common theme, we focus in particular on the dynamic effects of price promotions (rather than non-price promotions) offered to consumers (rather than to the trade). In our study, price promotions are current-period temporary price reductions to enhance sales or loyalty (Gedenk et al., 2006).

Thus, in this study we define sales promotions as monetary promotions, also called price deals or price promotions. Within the research context (Dutch Supermarkets), price promotions are most frequently offered. Further, within a groceries context the Dutch term “Aanbiedingen” is understood as a price promotion.

2.2.2 Strategic effects of sales promotions

Strategic effects are categorized by Neslin and Van Heerde (2009) in their comprehensive overview of the current literature regarding the various effects of promotions (Neslin and Van

(12)

Heerde, 2009). We added the effect on brand equity to this category due to the wide body of research and interest for this particular topic in relation to sales promotions.

State dependence

State dependence is the degree to which current choice aﬀects future utility (Neslin and Van Heerde, 2009). Choosing a brand in this period may increase the probability it will be purchased again. There are several reasons why consumers may exhibit state dependence (Neslin and Van Heerde, 2009): (1) learning: the customer learns more about the attributes and/or performance of a brand after purchasing it, (2) inertia: it may be cognitively easier to stay with the brand last purchased; the consumer may not see the value of investigating another brand considering the eﬀort involved, (3) switching costs: even if another brand appears to be better, it isn’t worth the hassle costs of switching to it, and (4) variety seeking: consumers may innately derive pleasure from switching brands, just for the sake of trying something new.

Inertia and switching costs result in positive state dependence. Learning may increase or decrease the likelihood of subsequent purchase, depending on what is learned. Variety seeking by deﬁnition decreases the likelihood of purchasing Brand A at the next purchase occasion. Although state dependence could have either a positive or negative impact on future purchasing, the almost universal ﬁnding is that state dependence is positive. The predominance of positive state dependence suggests either that consumers learn positive things about the products they buy, or through either inertia or switching costs, become creatures of habit (Neslin and Van Heerde, 2009).

Reference Prices

The reference price is the standard with which consumers compare the currently available price of an item in order to assess the attractiveness of the available price (Winer, 1986). If the comparison is highly favourable, i.e., the available price is lower than the reference price, the consumer is more likely to purchase the product. If the comparison is not favourable, the consumer will be less likely to purchase the product (Neslin and Van Heerde, 2009).

The reference price phenomenon is important for several reasons (Neslin and Van Heerde, 2009). First, it can serve as a strategic argument for the existence of price promotions. Second, the reference price effect influences the profitability of promotions, because it increases the size of the promotion bump and has a negative repeat-purchase effect after the promotion. A third implication of the reference price phenomenon is that manufacturers should avoid promoting too frequently. Frequent promotions can render promotions less effective in the long-run (because reference price is lowered so a given discount doesn’t seem very attractive) and decrease baseline sales (Neslin and Van Heerde, 2009).

Price sensitivity

Frequent exposure to sales promotions may aﬀect consumer perceptions of promotional activity (Krishna et al., 1991) and change their response to promotions (Raju, 1992). Discounting policies are typically found to decrease price elasticities (i.e. make them more negative) by focusing consumers’ attention to price-oriented cues (Boulding et al., 1994;

(13)

Foekens et al., 1999; Mela et al., 1997; Papatla and Krishnamurthi, 1996; Pauwels et al., 2002).

Summarized, there is ample evidence that an extended exposure to price promotions make consumers more price sensitive (Neslin and Van Heerde, 2009).

Brand equity

Focusing on the direct effect on brand equity dimensions, monetary promotions are likely to have a negative influence on perceived quality and brand associations (Buil et al., 2013). The reduction in the internal reference price is one of the main reasons why monetary promotions have a negative influence on perceived quality. Consumers use price as an extrinsic cue to infer product quality (Agarwall and Teas, 2002.) As such the influence of price discounting on consumers’ reference price can lead to unfavourable quality evaluations (Mela et al., 1998; Delvecchio et al., 2006). Similarly, monetary price promotions can erode brand associations. Martinez et al. (2007) and Monetar and Pina (2008) reported that monetary promotions have a negative impact on brand image. In addition monetary promotions campaigns are too short to establish long-term brand associations and can create uncertainty about brand quality, which results in more negative brand perceptions (Buil et al., 2013). Regarding brand awareness the results are inconsistent. Yoo et al. (2000) found a negative relationship between brand awareness and price promotions. However Srinivasan et al. (2008) identify a positive relationship. Huang and Sarigöllü (2012) stated this may be due to the use of different research contexts (durable goods versus convenience goods and jointly measurement of brand awareness and brand associations versus pure awareness). In their own research they conclude the relationship is positive and managers should use price promotions to create brand awareness and induce brand usage. Specifically, price promotions encourages brand switching and provides consumers with an incentive to try those brands which they would not purchase at the full price (Huang and Sarigöllü, 2012).

Although some discussion about the effect of price promotions on brand equity still exists Palazón-Vidal and Delgado-Ballester, 2005; Joseph and Sivakumaran, 2008), the empirical findings indicate that the frequent use of price promotions has a negative impact on perceived quality and brand association dimensions, because this tool leads consumers think primarily about price and not about the brand (Yoo et al. 2000). Regarding brand awareness the empirical evidence suggests the relationship is positive, especially for convenience goods (Huang and Sarigöllü, 2012).

2.2.3 Permanent effects of sales promotions

In this paragraph, we discuss some important literature regarding the permanent effects of promotions, especially on long term category demand and long term brand sales.

(14)

Long term effects on category demand

The focus of the research of Nijs et al. (2001) is on the short and long-run category demand effects of price promotions instead of their impact on brand sales. They find that category demand is predominately stationary. Although they report a generally strong short term effect of price promotions on category demand, in the long term they rarely find persistent effects. In fact, the positive effects of price promotions dissipate after a period of about 10 weeks, and their long-term impact decreases to zero in 98% of the 560 product categories researched in Dutch supermarkets (Nijs et al., 2001).

Some other interesting results of this study are: the less oligopolistic the category, the smaller the short term effects of price promotions and both the short- and long term promotion effectiveness is higher in perishable product categories (Nijs et al., 2001).

Permanent eﬀect on brand sales

The picture that emerges from the research from Neslin and Van Heerde (2009) is that there is quite a strong difference in the permanent effect of promotions for new and small brands versus mature and large brands. For new and small brands, promotions generate trial and can lead to a long-term increase in sales. For mature and large brands, promotions tend to have no or very small permanent effects on sales. Their customer base is established and there is little scope to attract new users. Hence, for these brands, discounting plays a largely tactical role by generating strong bumps in the short run, but it has no or little value as a strategic long-term marketing instrument (Neslin and Van Heerde, 2009).

Long term effects on brand sales

In their study Ataman et al. (2010) consider the relative role of the integrated marketing mix (advertising, price promotion, product, and place) on the long-term performance of mature brands. They state “their results stand in marked contrast to the previous emphasis in the literature on price promotions and advertising”. This conclusion is based on the finding that when all the elements of the marketing mix are combined, the effects indicate that product (60%) and distribution (32%) have a substantially larger relative effect on brand sales over the long run than discounting (2%) or advertising (6%). Product is in their study operationalized by line length (number of stock keeping units), discounting by discount depth, distribution by % ACV(=all commodity volume)-weighted distribution and advertising by expenditure (Ataman et al., 2010).

In addition, they report that over the long run base sales are positively affected by advertising but negatively affected by discounting. Finally they confirm the negative effect of discounting on regular price elasticities (Ataman et al., 2010).

These outcomes allow Ataman et al. (2010) to state that “discounting plays a largely tactical role by generating strong bumps in the short run, but it has adverse effects as a strategic long-term marketing instrument”. In this way they are aligned with the conclusion of Neslin and Van Heerde (2009) concerning mature and large brands.

(15)

2.3 Antecedents of PRIC

When customers believe that promotions are important in their buying decision, they expect promotions to provide benefits. The notion of promotion functions has been widely discussed in the literature both from an economic benefit (or purchase cost) perspective and from a hedonic benefit perspective (Martinez and Montaner, 2006). Chandon et al. (2000) integrate both perspectives and consider that consumers respond to sales promotions due to the positive experience provided both from utilitarian (economic) and hedonic benefits.

In this study, we follow this view, which we discuss in this paragraph. Thus, we consider the utilitarian (economic) and hedonic benefits as important antecedents of the PRIC construct (see figure 2).

Figure 2. PRIC and its antecedents

The promotion benefits model

Chandon et al. (2000) built a framework to embrace the full scope of benefits of sales promotions and so go beyond the obvious saving benefit. They define two categories of benefits, which consist of 3 benefits each (see figure 3).

Figure 3. The multi-benefit framework of sales promotions (Chandon et al., 2000)

Utilitarian benefits Hedonic benefits

(16)

The classification starts with the distinction between utilitarian (extrinsic) and hedonic (intrinsic) benefits (Holbrook, 1994). Utilitarian benefits are primarily instrumental, functional, and cognitive; they provide customer value by being a means to an end. Hedonic benefits are non-instrumental, experiential, and affective; they are appreciated for their own sake, without further regard to their practical purposes (Hirschman and Holbrook, 1982). Babin et al. (1994) show that this distinction also applies to shopping, because this activity provides utilitarian benefits (by helping consumers efficiently find and buy the best products) and hedonic benefits (by creating entertainment and raising self-esteem). Similarly, the benefits of sales promotions can be classified as utilitarian when they help consumers maximize the utility, efficiency, and economy of their shopping and buying and as hedonic when they provide intrinsic stimulation, fun, and self-esteem. Through these definitions, the savings, quality, and convenience benefits of sales promotions can be tentatively classified as utilitarian, because they help consumers increase the acquisition utility of their purchase and enhance the efficiency of the shopping experience (Chandon et al., 2000).

In contrast, the entertainment and exploration benefits of sales promotions can be tentatively classified as hedonic, because they are intrinsically rewarding and related to experiential emotions, pleasure, and self-esteem (Chandon et al., 2000). As figure 3 shows, the value expression benefit of sales promotions is different, because it entails both hedonic and utilitarian dimensions. On the one hand, buying a promoted product can provide shoppers with the moral satisfaction of behaving according to their principles and values (e.g., being good) an intrinsic or hedonic benefit. On the other hand, buying a promoted product can be a means of increasing shoppers' prestige and achieving higher social status or group affiliation (e.g., becoming recognized as a smart shopper), an extrinsic or utilitarian benefit (Chandon et al., 2000).

In their research Chandon et al. (2000) conduct several measurement studies and conclude that the mean value and explanatory power of each benefit are significantly different between monetary and nonmonetary promotions. Nonmonetary promotions provide stronger hedonic benefits and weaker utilitarian benefits than monetary promotions, and nonmonetary promotions are evaluated primarily on the basis of their hedonic benefits, whereas monetary promotions are evaluated primarily on their utilitarian benefits. With the exception of value expression, which is a universal predictor because of its dual utilitarian and hedonic nature (Chandon et al., 2000).

Regarding our study, we expect that the six benefits of the framework of Chandon et al. (2000) are important determinants of PRIC. The more important sales promotions are for gaining these benefits in a given category, the higher PRIC should be. But, as these researchers found no significant impact of the exploration and entertainment benefits for monetary promotions and in the Dutch supermarket context “aanbiedingen” will be interpreted as monetary promotions, we expect that the four other benefits will be more important determinants in this study.

(17)

2.4 Consequences of PRIC

Price promotions are used extensively in marketing for one simple reason: consumers respond. So one of the purposes is to elicit a direct impact on the purchase behaviour of the firm’s customers (Blattberg and Neslin, 1990). This implies, promotion relevance in category is associated with several utilitarian (economic) consequences at the customer level. In categories with higher promotion relevance, customers will have a greater demand for promotion benefits, such as monetary savings and increased quality. As a consequence, when promotions are more relevant to customers, customers should be more willing to switch brands, switch stores and accelerate their purchases both in time and quantity. These and the other immediate consequences (effects), derived from literature, will be discussed in this paragraph.

Immediate sales effects

Sales promotion effect measurement has been a key research area in marketing for decades (Blattberg and Neslin, 1990; Blattberg, et al., 1995; Neslin, 2002). Sales promotions are documented to have strong impact on own item sales (Neslin, 2002). In addition, numerous insights have been generated on the effects of promotions on brand choice (Guadagni and Little, 1983; Tellis and Zufryden, 1995), on purchase timing (Jain and Vilcassim, 1991), on purchase quantity (Neslin, et al., 1985), on store switching (Kumar and Leone, 1988; Walters 1991), on increases in consumption (Wansink, 1996; Ailawadi and Neslin, 1998), and on anticipatory responses (Krishna, 1994a and 1994b; Mela et al., 1998; Sun, et al., 2003).

Figure 4. Consumer responses to promotions (Van Heerde and Gupta, 2005)

The seven direct promotion (decomposition) effects identified by Van Heerde and Gupta (2005), contributing to the sales promotion bump are: increased consumption, cannibalization, brand switching, stockpiling, anticipation effects and store switching effects. See figure 4 for an overview, which also provides a selective overview of literature (Van Heerde and Gupta, 2005). These responses are also used by Luijten (2012) in his research to sales promotions effectiveness in the Dutch FMCG-market. In table 2 we provide a short definition of these seven immediate promotions effects.

Gupta (1988) Chiang (1991) Chintagunt a (1993 Buklin and Lattin (1992) Tellis and Zufryen (1995) Ailawadi and Neslin (1998) Sun et al. (2003) Ailawadi et al. (2005) Van Heerde and Gupta (2005) Store switching √ √ Brand switching √ √ √ √ √ √ √ √ √ Cannibalization √ Timing acceleration √ √ √ √ √ √ √ √ √ Quantity acceleration √ √ √ √ √ √ √ Increased consumption √ √ √ Anticipation √ √ Stockpiling

(18)

1 Store switching

Sales for the promoted brand can increase during the promotion by attracting customers from other stores. Sales promotions can induce consumers to change stores, to take advantage of offered promotions.

2 Brand switching

Sales promotions can induce consumer to switch their choice from other brands.

3 Cannibalization

Effects of sales promotions on the demand for items of the same brand have the label “cannibalization,”

4 Stockpiling by timing acceleration

Sales being shifted from the future to the current period, so promotions can induce consumers to purchase earlier than usual.

5 Stockpiling by quantity acceleration

Sales being shifted from the future to the current period, so promotions can induce consumers to purchase earlier than usual.

6 Increased consumption

Sales promotions might be able to increase consumption, which is equal to the total (net) observed demand increase across all products, stores, and time periods.

7 Anticipation = deceleration

Promotions can induce consumers to defer purchase until a promotion is available (purchase deceleration). So anticipation" involves shifts from past sales to the current period.

Table 2. Definitions of immediate consumer responses to promotions (Neslin and Van Heerde, 2008)

The size of the immediate sales effects

Several researchers have decomposed sales promotion elasticities based on household scanner-panel data into the described immediate effects (Van Heerde et al., 2003). A key result is, that the majority of the sales promotion elasticity, approximately 74% on average, is attributed to secondary demand effects i.e. brand switching and the remainder is attributed to primary demand effects i.e. timing acceleration and quantity increases (Van Heerde et al., 2003).

In the coffee category, Gupta (1988) finds that the percentage of own-brand sales elasticity with respect to a particular promotion that is due to brand-switching elasticity is 84%, that is due to purchase acceleration elasticity is 14%, and that is due to quantity elasticity is 2%. Gupta (1988) notes that such decomposition may be used to compare the effectiveness of alternative promotional offerings and to determine the most suitable promotion for a brand. As mentioned, Van Heerde et al. (2003) find that, on average, secondary demand effects (brand switching) account for the vast majority (approximately 74%) of total elasticity, which leaves 26% for primary demand effects (purchase acceleration and quantity increases).

(19)

Van Heerde et al. (2003) also demonstrate that this result does not imply that if a brand gains 100 units in sales during a promotion, the other brands in the category lose 74 units. The authors offer a complementary decomposition measure based on unit sales. This measure shows the ratio of the current cross-brand unit sales loss to the current own-brand unit sales gain during promotion. They find that approximately 33% of the unit sales increase is attributable to losses incurred by other brands in the same category, that is, the other brands together lose approximately 33 units (Van Heerde et al., 2003).

The much higher percentage of secondary demand effects indicated by elasticity decomposition results in other studies arises because it focuses on the gross change in sales for the non-promoted brands, when category volume is held constant (Van Heerde et al., 2003). In contrast, Van Heerde et al. (2003) focus on the net change, accounting for increasing category volume, which partly benefits the non-promoted brands. Another main finding is that the primary demand effects of promotions are greater than what has been assumed so far: 66% in unit sales rather than 25% in terms of elasticities (Van Heerde et al., 2003).

The relationship between size and frequency of sales promotions

If a brand is promoted very infrequently, consumers are likely to use these opportunities to stock-up for future consumption (Raju, 1992). On the other hand, if a brand is on promotion most of the time, consumers are not likely to benefit much by stocking up on any given promotional occasion (Krishna et al., 1990). This decrease in stockpiling as promotional frequency increases has been shown by using simulations by Helsen and Schmittlein (1989) and Assuncao and Meyer (1990). Furthermore, Winer (1986) and Lattin and Bucklin (1989) conclude that more frequent discounts may lower the reference price of the promoted brand, which in turn may negate the effect of promotions. Consequently, if most brands in a product category are promoted very often, the consumer is not likely to indulge in stockpiling behaviour for any of the brands. Hence, category sales are not likely to be affected as much. Therefore, product categories in which the brands are promoted very often may exhibit lower variability (Raju, 1992).

Immediate effects of sales promotions in this study

In this study, we will consider brand switching, store switching and purchase acceleration as important consequences of PRIC. In this way, we follow Luijten (2012) in his choice for the scope of effects of sales promotions.

Concerning the other short term effects, cannibalization is not feasible due to the fact we conduct a category study and not a brand study. Increased consumption is an effect which is, according to Luijten (2012), relatively not important due to the fact the FMCG-market in the Netherlands is a saturated market, although Bell et al. (1999) reports increased consumption for some categories. Further we leave anticipation out of scope due to the small sales effects as found by Van Heerde and Gupta (2005).

(20)

2.5 Contingency factors of PRIC

Several studies (Raju, 1992; Narasimhan et al., 1996; Bell et al., 1999 ; Chandon et al., 2000; Luijten, 2012) have suggested that the type of product is an important variable in explaining differences in promotional responses and therefore might be an important factor in relation to PRIC. So, consumers may perceive the importance of promotions as being different across categories because of differences in product-market characteristics, such as the hedonic/utilitarian nature of the category, purchase frequency, bulkiness, category penetration, perishability, stockability and share of budget. Moreover, we assume that consumer demographics, such as age, gender, income, education and family size will impact PRIC.

In this paragraph we discuss the relevant literature regarding the impact of both consumer demographics and category differences on promotional response. Further, we discuss the scope of this study for these contingency factors and our expectations regarding their relation to PRIC.

2.5.1 Hedonic and utilitarian products

Many product-related variables exist to distinguish between types of products. A first product category variable is the type of product: utilitarian or hedonic. Sloot (2006) defines this product-variable as the basic benefits that a product provide to consumers.

Batra and Ahtola (1990) describe the origin of this bi-dimensional consumer attitude: “…because consumers purchase goods and services and perform consumption behaviours for two basic reasons: (1) consummatory affective (hedonic) gratification (from sensory attributes), and (2) instrumental, utilitarian reasons concerned with "expectations of consequences" (of a means-ends variety, from functional and non-sensory attributes)”. Hedonic products, such as ice cream and salty snacks, provide more experiential consumption (i.e., fun, taste sensation, pleasure, excitement), whereas utilitarian products, such as detergents and toilet paper, provide primarily instrumental and functional benefits (Hirschman and Holbrook, 1982). Some products may offer both utilitarian and hedonic benefits to consumers (Sloot, 2006). Shampoo, for example, combines a utilitarian benefit (cleaning hair) with a hedonic benefit (nice smell). Moreover, even products that are bought mainly for utilitarian reasons may provide some hedonic benefits (Sloot, 2006). For example, consumers may perceive a product such as milk, which is often bought for its nutritional value (utilitarian benefit), as very tasty (hedonic benefit).

The different nature of utilitarian and hedonic products may affect the buying process, in that the buying process of utilitarian products may be driven mainly by rational buying motives, whereas that of hedonic products includes emotional motives as well (Sloot, 2006). Thus, in the promotional buying process of hedonic products emotional motives might play an important role, which may affect promotional responses.

(21)

In our study we use the scale of Sloot and Verhoef (2008) which was derived from the scale of Batra and Ahtola (1990). We asked respondents to score the products involved in this study on three items using a 7-points Likert- scale (1=low, 7=high): (1) attractive, (2) nice and (3) enjoyment. See appendix 8.1 and 8.2 for the questions in both English and Dutch. Note that we selected typical utilitarian or typical hedonic categories for our research. In table 3, the results of Sloot and Verhoef (2008) are shown. Sloot (2006) mentioned beer, chips, cigarettes and cola as typical hedonic products and eggs, milk, margarine and detergent as typical utilitarian products.

Table 3. Utilitarian and Hedonic products (Sloot and Verhoef, 2008) based on the scale Batra and Ahtola (1990) which is also use in this study

2.5.2 Promotions and product category characteristics

In this paragraph we will present the most relevant literature concerning promotions and category characteristics (other than their hedonic/utilitarian nature) which mainly origin from the nineties. We end with some more recent studies which confirm some results of the earlier studies.

Market characteristics and promotional price elasticities

Bolton (1989) was among the first to investigate the market characteristics associated with differences in promotional responses. She developed a model which take into account factors such as brand market share, manufacturer and retailer advertising levels and promotional activity at the brand and category level. Her results indicate that the researched brand and market/category variables explain a substantial amount of the variation in promotional responses (26%). A main finding is that brands with smaller market shares, lower levels of category and brand display activity and higher levels of category and brand couponing are more elastic. She also reports that the effects of category display and feature activity on promotional responses are much larger than the effects of brand prices, display and feature activity (Bolton, 1989).

Product category Cronbach's alpha (three-item scale) Average hedonic level (1-7) SD M argarine .83 3.40 1.21 Rice .84 3.86 1.21 Detergent .80 3.90 1.45 Toilet paper .86 3.92 1.54 Frozen vegetables .74 4.05 1.06 Sauces .83 4.54 1.22 Cola .82 5.03 1.22 Coffee .54 5.38 0.81 Beer .73 5.55 0.97 Cigarettes .88 5.78 0.94

(22)

Category structure and promotional activity for grocery products

Fader and Lodish (1990) report systematic relationships between category characteristics and promotional policies (the frequency and types of promotions used by retailers and manufacturers). To investigate these differences, they analysed a large number of product categories. The following variables were part of their research: household penetration, purchases per household (separate amount of shopping trips), the average interpurchase time (days), the private label share in the category and price (average spending per purchase occasion). Of these structural variables they conclude that household penetration is the single most informative measure of category movement and price demonstrates surprisingly weak associations with cross category differences (Fader and Lodish, 1990).

The effect of price promotions on variability in product category sales

The study of Raju (1992) differs from the one of Fader and Lodish (1990) in the sense they focus on the relationship between category characteristics and the use of promotional instruments (coupon offers, price cuts etc.) but their research does not address how promotional activity in a product category affects category sales. The category variables researched in the study of Raju (1992) are: bulkiness (size) of the products in the category, the expensiveness of the products in the category (weighted average of regular prices of various brands in the category) and the intensity of competition in the category (a combination of the market share pattern and the number of brands in a product category). Another category characteristic which may affect variability in category sales but was not integrated in this study is perishability (Litvack et al., 1985). Perishability is important if product categories have different shelf lives. However, perishability might be important when comparing categories such as milk and meat with cereal and ketchup (Raju, 1992).

Raju’s results suggest that categories which are bulky, or categories in which the degree of competitiveness is high, exhibit significantly lower variability in sales. This means these categories show lower promotional responses. The expectation that expensive categories

exhibit lower variability in sales was not supported. (Raju, 1992).

Promotional elasticities and category characteristics

Narasimhan et al. (1996)) performed an extensive study to the relationships between product category characteristics and average promotional elasticity within the category. They studied 108 product categories and used data compiled from weekly scanner data, scanner panel data, and own survey data. They present a framework for understanding these relationships and use it to test their hypotheses. Their research focuses on how the average increase in brand sales resulting from three different types of promotions (featured price cuts, displayed price cuts and pure price cuts) can be explained by the following seven category characteristics:

 Category penetration;  Interpurchase time;  Price;

 Private label share;  Number of brands;  Impulse buying;

(23)

The first five category characteristics were obtained from panel data (IrI). Impulse buying and stockability data were collected via survey data. Their results indicate that promotional elasticities are higher for categories with relatively fewer numbers of brands, higher category penetration, shorter interpurchase times, and a higher stockability scores (Narasimhan et al., 1996). The authors find no statistically significant relationship between promotional elasticity and either impulse buying or private label market share.

Thus, category penetration, interpurchase time and stockability have an important and consistent relationship with promotional elasticity. Therefore promotions for brands in easily stockpiled high-penetration categories with short interpurchase times should produce large sales increases (Narasimhan et al., 1996). Examples of such categories are products such as towels, toilet tissues, canned products (like seafood and vegetables) and pasta. Promotions for brands in categories that are relatively less readily stockpiled, have lower penetration and longer interpurchase times should show lower sales increases (Narasimhan et al., 1996). Examples, are categories as health and beauty aids (like deodorant, razors, shaving cream, and bath products) and miscellaneous products (like spices, floor cleaners and bottled water).

The decomposition of promotional response

Bell et al. (1999) included category, brand and consumer factors in one study and researched their impact on promotional response and distinguished this response into primary (purchase acceleration both in timing and quantity) and secondary (brand switching) demand effects. They conclude that category-specific factors, brand-specific factors and consumer demographics explain a significant amount of the variance in promotional response for a brand at both the primary and secondary demand levels (Bell et al., 1999). Category-specific factors explain most of the variability in promotional response. Further brand-specific factors play a modest role and consumer characteristics have a relatively little explanatory power, see table 4 (Bell et al., 1999).

Table 4. Factors explaining differences in promotional response (Bell et al., 1999)

The following seven product category factors or characteristics were part of this study:

 Share of budget: share of the budget for the consumers shopping list, comparable and operationalized in line with the expensiveness concept of Raju (1992);

 Brand assortment: this is the breadth of the variety and in line with the number of brands concept of Narasimhan et al. (1996);

 Size assortment: refers to the depth of variety, meaning the variety of size assortments within a brand;

Secondary demand

Class of factor Brand choice Purchase incidence Purchase quantity

Category factors Strong M oderate Strong

Brand factors M oderate M oderate M oderate

Consumer factors Weak Weak No effect

(24)

 Storability - Stockability: the extent to which the products in a category facilitates stockpiling and therefore intertemporal purchasing, which is comparable with the ability to stockpile concept of Narasimhan et al. (1996);

 Perceived differentiation: the perception of assortment by the consumer, so not measured by the number of brands but by the experience and perception of the consumer;

 Necessity: this concept is related to the impulse buying concept of Narasimhan et al. (1996) where necessity is defined as non-impulse buying because consumers have little flexibility to adjust their buying behaviour;

 Purchase frequency: in line with Fader and Lodish (1990) and thus defined as the amount of (separate) shopping trips for a product category.

Except for size of assortment all these product category factors are found to have a significant impact on primary demand effects, secondary demand effects or both. Bell et al. (1999) especially mention share of budget and storability as two category factors that (see table 5 for the supported and not supported hypotheses) play a large role. Share of budget decreases the switching elasticity, but increases the quantity elasticity, while storability increase both. An interesting result concerning storability is that refrigerated products (margarine, yoghurt, ice-cream, bacon and butter) have much higher proportions for the incidence effect (buying and buying earlier) than for the quantity effect and storable products have just the opposite pattern. This result confirms that perishability (Litvack et al., 1985) influences consumer buying behaviour, which was also mentioned by Nijs et al. (2001) as a factor which influences both short- and long term promotion effectiveness (see paragraph 2.3.3).

Table 5. Category factor hypotheses of Bell et al (1999). In black the empirical supported hypotheses and in grey the not supported hypotheses

More recent studies confirming the impact of category differences on promotional response

Foxall and al. (2013) conclude in their study to brand related and situational influences on demand elasticity that price elasticity varies across both products and brands. Thus, the nature of the product influences the used independent variables in this study differently. They suggest product class and type as further area for exploration. Furthermore, Luijten (2012) confirms the influence of categories on consumer responses on promotions. He concludes

Category factors Secondary demand Primary demand

Share of budget -/- +/+ Brand assortment -/- +/+ Size assortment +/+ +/+ Storability - Stockability +/+ +/+ Perceived differentiation -/- +/+ Necessity +/+ -/-Purchase frequency -/-

(25)

-/-that all four immediate effects of sales promotions (brand switching, store switching, purchase acceleration by timing and purchase acceleration by quantity), researched by him occur, but they differ in magnitude for the investigated product categories, brands and stores where the differences found in his study are more across category than within category (Luijten, 2012).

2.5.3 Promotions and consumer demographics

In this paragraph we discuss three different promotional studies in which consumer demographics and psychographics were included. First, the already discussed studies of Bell et al. (1999) and Luijten (2012). And furthermore, the extensive study of Ailawadi et al. (2001) who research the impact of consumer psychographics and demographics on promotional usage.

Consumer factors in the study of Bell et al. (1999)

Bell et al. (1999) take three consumer factors into consideration namely age, education and income. Regarding income they expect that higher income consumers should be less sensitive to price but conversely also that stockpiling might occur more frequently and on a larger scale because these consumers have more ability to take advantage of deals when the opportunity arises. Concerning age, Bell et al. (1999) expect (based on the research of Ainslie and Rossi, 1998) that older consumers are less price sensitive although older households should have more time available to shop and search for deals. Finally, Bell et al. (1999) expect that more educated consumers are more diligent in taking advantage of price variability. Bell et al. (1999), conclude that consumer factors have relatively little explanatory power but nevertheless, they observe two small but significant effects (for age and education, see table 6, in which the expected and supported effects are shown). Furthermore, they support earlier literature in stating that variables that describe consumer shopping patterns might be more valuable in capturing price response than demographics alone (Ainslie and Rossi, 1998; Bell and Lattin, 1998).

Table 6. Consumer factor hypotheses of Bell et al. (1999), in black the empirical supported hypotheses and in grey the not supported hypotheses

The impact of consumer psychographics and demographics on promotional usage

Ailawadi et al. (2001) have investigated the extent to which store brands and promotions attract the same consumers. To do so, they have employed a structural model to study the characteristics of consumers who buy store brands and promotions. Then, they clustered

Consumer factors Secondary demand Primary demand

Income -/- +/+

Age -/-

(26)

consumers into segments on the basis of their use of store brand and promotions and distinguished between the segments using psychographic characteristics (15 constructs like price consciousness, need for cognition, loyalty and shopping enjoyment) and demographic characteristics (6 demographics like age, gender and income).

The authors find that the impact of demographics on these behaviours (the usage of store brands and promotions) is funnelled through psychographics instead of being direct. So demographics do not influence these behaviours directly. See table 7 for an overview of the psychographics which (statistically significant) impact the promotional usage of in-store or out-store promotions or both, as found by Ailawadi et al. (2001). Concerning Price consciousness, Quality consciousness, Innovativeness, Variety seeking and Time pressure no significant effect was found by these researchers. A similar study was performed by Martinez and Montaner (2006). They found similar but also some different results regarding the psychographics but didn’t take the impact of demographics into account.

Table 7. Consumer psychographics which impact the usage of national brand promotions (Ailawadi et al., 2001)

The indirect impact of demographics on consumer psychographics

As mentioned Ailawadi et al. (2001) found several demographics which impact promotional usage indirectly:

 Consumers with higher incomes are less price conscious and less financially constrained.  People who are employed full time and those with children in the household are more

pressured for time.

 People who live in a house perceive that they have more storage space.

 Age, gender and education are associated with many psychographic variables. Older consumers are more likely to be shopping mavens, have higher motivation to conform to the expectations of others, are less pressured for time, and have more storage space. Women are more likely than men to be innovative, impulsive, shopping mavens, and planners. They are also more store loyal (supported by Sloot and Verhoef, 2008) and have

Psychographic characteristic In-store promotion usage Out-store promotion usage

Financial constraint +/+ +/+ Shopping enjoyment +/+ Impulsiveness +/+ M avenism +/+ M otivation to conform -/-Brand loyalty +/+ Store loyalty -/-Planning +/+ +/+

Need for cognition

(27)

lower need for cognition. Finally, more educated consumers are less financially constrained, more quality conscious, more innovative, and have higher need for cognition. They do not enjoy shopping and have more storage space.

Thus, although the direct effect of demographic variables on promotional usage is weak, these significant indirect effects show that demographics might play a role in determining promotional shopping behaviour.

Demographics in the Dutch groceries market

Luijten (2012) found a significant negative effect of education on purchase incidence which indicates that people with a higher level of education are less triggered to buy promotions than people with a lower education level. For family size, he concludes that purchase incidence seems to increase when family size increases (a positive impact) which seems reasonable as the consumption needs of larger families is higher and probably the purchase frequency in a certain category as well. Luijten (2012) was not able to find a consistent effect for age as for some categories promotions effects increased when age increases and for other categories he found just the opposite effect.

2.5.4 Category characteristics and consumer demographics in this study

In this study we will include the hedonic/utilitarian nature of the category, purchase frequency, category penetration, perishability, stockability and share of budget, as category characteristics. As consumer demographics we will consider age, gender, education, family size and district. In this paragraph we discuss our expectation of the impact of these contingency factors on PRIC.

Product category characteristics

First, concerning the hedonic/utilitarian nature of products, we expect PRIC to be higher for more utilitarian products due to the research of Verhoef and Sloot (2008) who state that the buying process of utilitarian products may be driven mainly by rational buying motives. Further, Luijten (2012) found that people are more brand loyal towards hedonic (taste) products, which avoids brand switching. In addition, also from a price sensitivity perspective we expect that consumers are less sensitive to price in categories that are perceived as primarily hedonic in nature (Wakefield and Inman, 2003). So it seems people tend to find promotions more relevant for utilitarian products than for hedonic ones.

Secondly, we predict that the relevance of promotions is higher for categories which are easy storable, because this effect was found both by Narasimhan et al. (1996) and Bell et al. (1999). In our study, we follow the definition suggested by Litvack et al. (1985) to classify the stockability of our product categories. Apart from that, in our study the terms stockability, storability and ability to stockpile are interchangeably. Next to stockability, we also take perishability into consideration as logically this factor impacts the stockability of a product category. Further, Nijs et al. (2001) found a higher short- and long term promotion effectiveness for perishable categories.

(28)

Thirdly, we expect a higher PRIC for categories with a higher category penetration due to the findings of Fader and Lodish (1990) who report that category penetration is the best predictor of the level of usage of promotions by retailers and manufacturers, and so perhaps also for the usage by consumers. This presumption was confirmed by Narasimhan et al. (1996) who found that promotional elasticities are higher for categories with a higher category penetration.

Fourthly, we have no explicit expectation with regard to share of budget as Raju (1992) find no significant effect for price and Bell et al. (1999) report a negative effect for primary demand and a positive effect for secondary demand. In our study we operationalize share of budget by trip spend (per category) and we will use these terms interchangeably. Intuitively, we expect that PRIC will be higher for more expensive categories (higher trip spend).

Finally, regarding purchase frequency we don’t have a clear prediction of its effect on PRIC as Narasimhan et al. (1996) and Bell et al. (1999) found opposite effects on its impact on promotional responses. Intuitively, we expect that PRIC will be lower for frequently bought products.

Consumer demographics

Although, as described in this paragraph, literature tends to conclude that consumer demographics have no clear or no direct effect on consumer responses, we take some demographics into account. Thus, we we will test for some conclusions from literature in relation to promotional response and demographics. Luijten (2012) found some significant effects for education and family size with a negative effect for education (more educated people show lower promotional responses) and a positive effect for family size (larger families respond more to promotions than smaller families do).

Bell et al. (1996) found a negative effect for age and a positive effect for education, which is the opposite effect in comparison to Luijten (2012). Ainslie and Rossi (1998) found support for the idea that older consumers are less price sensitive, although often is argued that they should be more elastic because they have more time available to shop and search for deals. Bell et al. (1996) didn’t find an effect for income. However, Ainslie and Rossi (1998) stated that higher income consumers should be less sensitive to price.

Summarized, literature doesn’t provide a clear prediction of the effects of consumer characteristics on PRIC. We expect that the findings of Luijten (2012) regarding family size and education make sense and as a result of that, households with higher education will show a lower PRIC and larger households will consider promotions as more relevant. In our study we don’t have access to the income of respondents, so this variable will not be taken into consideration.

2.6 Price- and promotional constructs related to PRIC

In this paragraph we discuss four other promotions and price constructs which are often used and discussed in literature. These constructs are General deal proneness, Value

Does consumer promotion relevance matter? : a cross FMCG category study