EMPIRICAL INVESTIGATION OF IN-STORE SAMPLING PROMOTIONS
Abstract
Purpose -- This paper investigates the impact of in-store sample promotions of food products on consumer trial and purchasing behavior. We investigate differences in the trial rate for free samples across different products and consumer types, as well as the impact of sampling on product and category purchase incidence. The results of our study are relevant for retailers and manufacturers who invest in in-store free sample promotions.
Design/methodology/approach– We use data from a field study, which leveraged an actual free-sample program implemented by a US grocery store chain. Data was collected on six different products promoted by in-store free samples over six different weekends. The data collected included consumers’ trial and purchasing behavior with respect to the free sample, as well as their attitudes towards the free sample that day and free sample promotions in general.
Findings– Free sampling is very effective in inducing trial, especially among lower educated consumers. For consumers who were planning to buy the product in promoted category, free sampling can encourage switching from the planned to the promoted brand. For consumers who did not have such previous plans, free sampling can “draw“ them into the category and encourage category purchase.Samplers’ interactions with the person distributing the sample or with other samplers at the scene also seem to boost post-sample purchase incidence.
Originality/value– Despite the importance of free samples as a promotional tool, few studies have examined consumer trial and purchasing behavior with respect to in-store free samples. We present one of the first known field studies that examines this topic.
Key words: in-store samples, point-of-purchase promotions, trial behavior, purchase incidence
Paper type:Research paper
INTRODUCTION
Consumers today are bombarded with all different types of promotions. Vlachvei et al. (2009) show that for some food products like wine, promotions are even more important than advertising in ensuring profitability of the product. Most attention in the academic research has been devoted to the impact of promotions such asprice promotions and coupons(Gadenk and Neslin, 1999; Colombo et al., 2003; Swaminathan and Bawa, 2005) and the impact of loyalty schemes (Mauri, 2003; Meyer-Waarden, 2007; Demoulin and Zidda, 2008). However, the research on promotions such as product trials, free samples and free gifts have largely been overlooked (Peattie, 1998).
Among various types of promotions, few are as effective as free samples for generating trial and purchase (e.g., Rossiter and Percy, 1987; Belch and Belch, 1990). Studies in the popular press on in-store sampling have shown that 92 percent of consumers would rather be presented with a free sample over a cents-off coupon while in the store (Fitzgerald, 1996), nearly 70 percent of shoppers will try an in-store sample if approached, 37 percent will buy a product once sampled (Lindstedt, 1999), and in-store samples can increase sales of the sampled product by as much as 300 percent on the day of the promotion (Moses, 2005). Given these benefits, it is not surprising that expenditures on sampling programs increased to approximately $2B in 2004, a 50 percent increase from 2003 (Zwiebach, 2005). However, in the academic marketing literature, sampling remains one of the most under-researched areas of promotions (Heiman et al. 2001).
In-store sampling promotions have unique characteristics that require special consideration and therefore studies of their own. First, in contrast to at-home sampling which has been studied extensively in the literature (Rothschild and Gaidis, 1981; Gadenk and Neslin, 1999), in-store sampling occurs in a public setting at the point-of-purchase and therefore may be affected by a variety of environmental or situational factors, such as the manner in which the sample is presented, the presence of others at the sampling promotion site, or whether the consumer was planning a purchase in the category that day. Any of these could affect a consumer’s decision of whether to take a free sample and/or to purchase the promoted product. Another unique characteristic of in-store sampling is a marketer’s lack of control over who ultimately receives the promotion. This is because the distribution of in-store free samples is determined by consumers’ willingness to take the sample. Therefore, it is advantageous for marketers to understand consumers’ motives for taking, or rejecting, an in-store free sample if they wish to increase the trial rate of such a promotion. Despite these unique characteristics of in-store sampling, and the importance of this promotional tool to retailers and consumer packaged goods manufacturers, no research has investigated the impact of free sample promotions on trial and purchase in its natural environment to address such issues.
The research examining in-store free samples is sparse and in many cases not easily generalized. Lammers (1991) provided samples to consumers in a chocolate store and found the sample increased the immediate sales of chocolate. However, due to the unique characteristics of chocolate (e.g., impulse/gift purchase) and the retail format studied (e.g., a specialty shop with limited products and brand alternatives), the results here are not easily generalized to other food types and to sampling in a grocery store setting. Furthermore, because all patrons were given a free sample, the study does not probe the factors that affect a consumer’s decision to take or decline a free sample. Steinberg and Yalch (1978) also examined the effects of in-store free samples on shopping behavior, but focused on the differences in shopping behavior of obese and non-obese consumers. They found that consuming a food sample caused obese consumers to increase their purchasing in the store that day. Again, while interesting, the results here do not have broad application.
There are some studies that, while they do not research the impact of in-store free samples explicitly, include in-store sampling in their broader study of different types of in-store promotions (Rothschild and Gaidis, 1981; Gadenk and Neslin, 1999). Because in-store sampling is only one type of the many promotions studied in these papers, these papers do not examine in-store sampling at the level of detail as we, for example, by investigating the factors that affect consumers’ decision to take an in-store free sample or the impact of free samples on purchasing behavior and store perceptions.
Our study focuses on in-store free samples of food products presented within a grocery or club store. The purpose of our study is to examine the factors that affect consumers’ decision to take such free samples and the impact of in-store free samples on purchasing behavior and store perceptions. We present empirical results from an in-store field study that intercepted consumers during a free sample promotion. Data was collected on six different weekends for six different sampled products. Our study provides insights into this important type of in-store promotion and is useful for retailers and manufacturers of food products who invest in such sampling promotions.
The remainder of the paper is organized as follows. In the next section we provide the details of our in-store field study and the resulting data collected. We then present the results of our empirical analysis, the insights gained from our findings, and the relevance of our results for marketers. We conclude with directions for future research.
IN-STORE FIELD STUDY
To collect our data we leveraged the existing free-sample program implemented by a Midwestern U.S. grocery store chain. On any given weekend the grocery chain ran sample promotions for the same products in all of its stores over the entire weekend (Friday-Sunday). For ease of the facilitation of the data collection, each weekend we focused exclusively on one promoted product only and data was collected for that promoted product in three different stores within the chain on that weekend. Different outlets within the grocery chain were chosen to ensure variance in the demographics and socio-economic status of the consumers in our sample.
Data collection commenced for six consecutive weekends resulting in a dataset consisting of six different products promoted by in-store samples. The products investigated were chosen to allow for a cross-section of product characteristics among those we were interested in investigating and the six consecutive weekends were chosen to avoid issues of seasonality and holiday periods that might have influenced shopping behavior, and as such, we were able to ensure as much consistency as possible in the shopping conditions across the six weekends investigated. While the duration of the data collection was consistent each day (three hours), the starting times differed each day in order to vary the mix of customers intercepted and to eliminate any possible “time of day” effects. Finally, in all cases the free sample was presented in close proximity to the shelf where the product was normally displayed and packages of the sampled product were also available for purchase at the sampling station.
Consumers were intercepted while exiting the store and screened for having seen the free sample of interest. Among those who had seen the free sample we surveyed both “samplers” (those who took the sample) and “non-samplers” (those who saw the sample but declined to take it) about their behavior and attitudes with respect to the free sample promotion that day and in-store free sample promotions in general. Upon completing the survey all participants were offered $5 in exchange for their grocery receipt and participation in a lottery drawing to win one of three $100 cash prizes. All survey participants accepted this offer.
Of the 380 intercepted consumers who had seen the free sample, 354 (93 percent) agreed to participate in our study, and of those 259 (73 percent) were “samplers” and 95 (27 percent) were “non-samplers”. Table 1 provides information about the products studied, the sample rate of those products (e.g., the percentage of consumers who took the sample given they saw it), and the purchase incidence for the promoted product among “samplers” and “non-samplers”.
[Insert Table 1]
We note that the products in our study varied on many characteristics that should be of interest to practitioners and academics. These include the price of the sampled product, whether the sample was for a new product (e.g., a line extension in the form of a new flavor or new form, but not a new brand or a brand extension into a new category), whether it was a store brand, whether it required cooking preparation at the site of the promotion, and whether the product was perceived as offering consumers hedonic (e.g., fun or more “treat like”) and/or utilitarian (e.g., useful and practical) benefits. Regarding this last characteristics, because there was virtually no overlap in the products in our study and those prior to ours that have labeled products as “hedonic” or “utilitarian” (Chandon, Wansink and Laurent 2000; Simonson, Carmon and O’Curry 1994; Voss, Spangenberg, and Grohmann 2003), the procedure used to label the products in our study as either “high”, “medium” or “low” on the hedonic/utilitarian scale,and as reflected in Table 1,is provided in the Appendix.
EMPIRICAL FINDINGS
The following presents our insights about trial behavior with respect to in-store free samples, consumers’ motives for taking or declining a free sample, and consumers’ attitudes towards in-store free samples in general.
Trial Behavior
We begin by noting that of the 354 consumers randomly intercepted in the store and who had seen the free sample, 259 (73 percent) took the free sample. This high “trial given exposure” rate is in line with the popular press statistics reported earlier and is good news for manufacturers who invest in such a promotions, especially if their main goal is to spur trial.
Akin to “coupon-proneness” which can vary by consumers’ individual characteristics and some category specific variables (Colombo et al, 2003; Swaminathan and Bawa, 2005), we investigate whether we can differentiate customers according to their “sample-proneness”. In particular, we were interested in whether there were any personal characteristics that differentiated those who took a free sample and those who declined to take one despite seeing it and whether certain products were more likely to be sampled than others. To investigate the former we compared “samplers” and “non-samplers” based on their demographic and shopping behavior information shown in Table 2. The only demographic characteristic that differentiated samplers from non-samplers was whether they had a college degree. Specifically, we found that samplers were less likely than non-samplers to have a college degree (p < 0.05). This suggests that free sample promotions have the potential to be more effective for generating trial in markets with a less-educated population.
[Insert Table 2]
We note there is no difference between samplers and non-samplers in the size of their shopping baskets or the amount they spent on their shopping trip. This dispels the idea that sampling is less likely to occur on “filler” trips when consumers wish to get through the store quickly. However, we do find that those who took the sample of interest sampled more products in the store that day (1.78 samples) than those who did not take the sample (0.60) (t=11.11, p < 0.001). This suggests that “sampling” may be a consumer characteristic rather than something a consumer does if the product is of interest to them.
Next, we were interested in whether the trial rate varied across products. To test this we compared the trial rate (e.g., the percentage of consumers who sampled the product given they saw it) across the six product categories. We see in Table 1 that the trial rate for ice cream and bread is statistically greater than the trial rate for the rest of the products (p < 0.01). We also found the trial rate for bagels (0.62) and yogurt (0.74) is not statistically different from each other (p > 0.15), but is statistically (p < 0.06) greater than the trial rate for ground turkey (0.43) and frozen pizza (0.40). Examining these results in light of the products’ characteristics in Table 1 it seems that consumers are most likely to sample products perceived as highly hedonic (ice cream) or highly utilitarian (bread). We note that the two products that were the least likely to be sampled were those that required cooking at the point of promotion. Perhaps consumers’ fears of consuming undercooked food or food that isn’t prepared to their tastes deters them from taking samples of such products. Price did not seem to impact the likelihood that a product will be sampled. These results are highly relevant for managers as they consider which products within their portfolio to promote with an in-store free sample promotion.
Consumer Motives for Taking or Declining Free Samples
In addition to observing sampling behavior, we were interested in the reasons why consumers take or decline free samples after seeing them. To investigate the former we asked samplers to indicate their agreement with the list of motives for taking a free sample shown in Table 3 and which were derived from exploratory research prior to this study. We see that the leading reason consumers take free samples is tofind out how the product tastes. This provides some initial support for one of our earlier findings, that new products have a high trial rate, perhaps because presumably consumers have no knowledge of how the new product tastes. It also suggests that free samples may be effective for inducing trial of existing products among consumers who have little experience in the category or with the promoted product, or for “follower brands” that simply do not have the same consumer “exposure” as a market leader, because in these cases consumers presumably have limited knowledge about the taste of the promoted product.
While these insights may seem obvious, one might have expected the opposite behavior – that due to risk aversion consumers would be more likely to try products with which they were familiar in order to avoid the potentially bad experience of sampling a product they did not like. In light of our findings, it is promising that free samples do not seem to be wasted on consumers who already know how the promoted product tastes, but rather samples seem to be attracting consumers who are truly interested in trying the product because they are unfamiliar with it.
[Insert Table 3]
Another point we were interested in was whether there existed a relationship between consumers’ motives to sample a product and their likelihood to purchase the product after sampling it. To investigate whether post-sample purchase was correlated with sample motives we began by reducing the sampling motives to a more meaningful and manageable number using exploratory factor analysis with principal components method and Varimax rotation. A three-factor solution captured 62 percent of the variation and was maintained based on the fact that the Eigen values for the first three factors were greater than one and those of the remaining factors were less than one. The three-vector solution was also supported by the scree plot. Cronbach’s alpha ranged from .66 to .75.