Bawa & Shoemaker (2004) conducted a study on the impact of a sampling procedure. In their first experiment they mailed a sample one serving pouch of the product accompanied by a 75-cent coupon to 2,059 randomly selected households. The control group (N=1,994) received nothing. The results indicated a marked increase in product purchasing by the test group when compared to the control group. There was also increased buying in the test group 52 weeks after the promotion. The authors admit these long-term results are not typical when compared to other investigations, and the authors could not account for all extraneous variables, such as pre-experiment use of the product (Bawa & Shoemaker, 2004). In a second experiment the authors arranged four separate groups to test various aspects of product sampling. They arranged one control group, one group which received a solo $1.50 coupon, one group which received a $1.50 coupon on one side of a post card, and one final group which received a free-sample and a .50-cent coupon. Again, the researchers found that free samples produced a substantially higher increase in sales following the free sample campaign when compared to the control group. It should be noted that the effects, although still long lasting for this type of promotion, lasted for 22 weeks; this number is significantly less than results obtained from the first study. The results obtained from the Bawa and Shoemaker study provide recent compelling evidence of the positive effects of free-samples; however, there have been other studies that have come to alternative conclusions. Finally, Lammers, (1991) handed out free chocolate samples to random customers entering a candy store. The experimenter reported mixed results. First, while handing out chocolates did result in an increased purchasing of chocolates compared to the individuals who did not receive free samples, the purchasing was restricted to chocolates other than those handed out as a free sample. Also, the experimenter reported that the positive increase in purchasing of those given free samples was restricted to purchases under five dollars. The research examining sampling procedures has yielded mixed results. The current study attempted to extend the existing sampling research. More specifically, we examined the effects of sampling procedures across multiple sites. Experiment 1 Method Participants and Setting Participants included employees of a fast-food chain restaurant in the Southeastern United States. During the study, the employees were responsible for distributing free samples and providing hourly print-outs to the experimenters. The fast-food restaurant was located in the food court of a mall. Procedure A multi-element design was used to evaluate two separate sampling conditions. The two conditions included a sampling condition, in which an employee handed out small bites of chicken to individuals passing by the restaurant, and a no-sample condition, in which no samples were distributed. During data collection, observers informed the employees as to which of two randomly determined conditions were in place. Data collectors would then sit in an unobtrusive location during the two hour session. At the end of the session, data were collected from the employees in the form of an hourly customer printout. Data were collected on both a weekday (Fridays) and a weekend day (Saturdays). Interobserver agreement (IOA) data were collected during 33% of trials. Two observers independently interpreted the printout. IOA was calculated by dividing the number of agreements by the total number of agreements plus disagreements. The average IOA across trials was 100%. Independent variable (IV) integrity data were collected by recording whether or not the sampling conditions were being conducted as instructed. IV integrity data were collected during 33% of trials and calculated in the same manner as IOA data. The average IV integrity across trials was 100%. Results
Figure 1 Results from the study indicated that providing free samples to customers in this setting did not significantly increase the number of customers per hour. Figure 1 depicts the results from study 1. The data from Friday suggests that distributing free samples has little to no effect in drawing in additional customers when compared to non-sampling conditions. The data from Saturday show a slight increase in the number of customers-per-hour when distributing free samples. Experiment 2 Results from the first study suggest that free samples had little effect on increasing the numbers of customers. However, no measure of sales was examined, thus it is possible that food purchases (i.e., sales) may have been influenced by providing free samples. Study number 2 examined this possibility in another setting. Method Participants and Setting Participants included the employees of an independently owned coffee house in the Southeastern United States. During the study, the employees were responsible for providing samples. The coffee house was located in a small plaza along a busy road. Procedure A multi-element design was used to evaluate two separate sampling conditions. The two conditions included a sampling condition in which small pieces of biscotti were available by the register for customers to take, and a no-sampling condition in which no biscotti samples were available. During data collection, data collectors would inform the employees as to which of the two randomly determined conditions were in-place. Data collectors would then sit in an unobtrusive location during the two-hour session and collect data on number of sampled items purchased, and the number of samples taken. IOA data were collected during 38% of trials. IOA was assessed by having a second observer independently collect data on biscotti sold. IOA was calculated in the same manner as in experiment 1. IOA averaged 83% across all trials. IV integrity was assessed by recording whether or not samples were being distributed during data collection according to sampling schedule. IV integrity data were collected during 38% of trials and calculated in the same manner as IOA. IV integrity averaged 100% across all trials.
Figure 2
Figure 3 Results Figure 2 shows the number of biscotti sold in both conditions. Figure 3 shows the total number of samples taken during the sampling condition. The data suggests that providing free samples to customers in the store had little effect on the number of sampled items sold. Experiment 3 Although sampling seemed to have little effect on the number of items sold in experiment 2, few of the sampled items were sold in general. In addition, previous studies have demonstrated that discounts can be effective in increasing the number of sales in sampled items. The purpose of experiment 3 was to replicate these findings and address the problem in study 2 of too few sampled items being purchased. Method Participants and Setting Participants included the employees of a chain ice cream shop in the Southeastern United States. During the study, the employees provided the samples and contingent discount described below. The ice cream store was located in a large plaza next to a busy road. Procedure One of three flavors of popular ice cream were designated as target flavors. An observer sat in an unobtrusive table and took data on sales of these three flavors (i.e., chocolate, mint chocolate chip, and vanilla). The dependent variable for this study was the total percent of these three target flavors sold during each session (i.e., a two-hour observation period). Three conditions were assessed in this study. First, a baseline condition of no sampling was conducted in which no samples of the target flavor were given unless a customer specifically asked for a sample of that flavor. A customer asked for a specific target flavor sample twice during baseline. In the experimental phase, two conditions were compared. The first experimental condition was a sampling condition in which signs were prominently displayed which conveyed to customers that the three target flavors were featured flavors, and were available to sample. The second condition involved signs featuring the three targeted flavors, as in the previous condition, as well as a description of a “discount”. While the signs were on display, a free topping (.75 value) was added to the purchase of any of the three target flavors. IOA data were collected during 36% of trials and calculated in the same manner as study 1. IOA was assessed by having a second observer independently record the number of target flavors sold. IOA averaged 100% across all trials. IV integrity was assessed by collecting data on whether the correct sampling condition was in-place. IV integrity was collected during 36% of trials and averaged 100% across all trials. Results
Figure 4 Figure 4 presents the results for experiment 3. The data suggest that neither the sampling sign, nor the discount condition increased target flavor sales above baseline. General Discussion
In the second study, distribution of free samples was examined to increase the number of a particular item sold once customers entered the store. Although customers took samples that were offered, the results were lackluster. However, due to a low number of total sampled item sales, results were inconclusive. The third study evaluated the effectiveness of sampling procedures by offering discounts in the form of free toppings. The discount procedure was ineffective in increasing sales of the target flavors. The discount did not involve a direct decrease in price of the target flavor, but rather a free addition (i.e., the overall cost of the target flavor remained constant across conditions). Future research could evaluate the degree to which direct discounts in item cost affect purchasing behavior.
Future research on sampling should examine the conditions under which a sampling procedure is effective. Future research should also evaluate whether or not the cost of giving away free merchandise and paying for the distribution of samples is cost effective. References Amir, H., McWilliams, B., Shen, Z, & Zilberman, D. 2001. Learning and forgetting: modeling optimal product sampling over time. Management Science, 47(4) 532-546. Bawa, K. & Shoemaker, R. (2004). The effects of free samples promotions on incremental brand sales. Marketing Science, 23, 345-363. Donnely Marketing Inc. 1994. Annual survey of promotional practices. Donnely Marketing Inc., Oaksbrook Terrace, IL. Gedenk, K., & Scott A. N. 1999. The role of retail promotion in determining future brand loyalty: Its effect on purchase event feedback. Journal of Retailing, 75(4) 433-459. Holmes, J. H., & Lett, J., D., Jr. 1977. Product sampling and word of mouth. Journal of Advertising Resources, 17(5) 35-40. Jain, D., Vijay, M., & Eitan, M. 1995. An approach for determining optimal product sampling for the diffusion of a new product. Journal of Product Innovation Management, 12 124-135. Lammers, H. B. (1991). The effect of free samples on immediate consumer purchase. Consumer Marketing, 8, 31-37. Peter, J. P., & Nord, W., R. (1982). A clarification and extension of operant conditioning principles in Marketing. Journal of Marketing, 46, 102-107. Schultz, D. R. & Petrison, W. A. (1998). Sales promotion essentials. Lincolnwood, IL: NTC Business Books. Scott, C. A. (1976). The effects of trial and incentives on repeat purchase behavior. Marketing Research, 13, 263-269. Byron Wine received his M.S. in Applied Behavior Analysis from the Florida Institute of Technology. He works as a behavior analyst for Advoserv of New Jersey. He can be reached at wineb@advoserv.com Amanda Fixsen, a graduate of the Florida Institute of Technology, is currently a graduate student at the University of Oregon. Carelle Harris is a graduate student at the Florida Institute of Technology in Melbourne, FL. David A. Wilder is an associate professor of psychology at the Florida Institute of Technology in Melbourne, FL and serves as an associate editor for the Journal of Organizational Behavior Management. He can be reached at dawilder@fit.edu.
|
||||||
The field of Organizational Behavior Management (OBM) has constructed a solid technology for affecting change in organizations. However, little OBM research has focused on how organizational practices affect customer behavior. Handing out free samples (also known as product sampling, or sampling) to customers is a common practice among businesses (Schultz et al., 1998). Many businesses use free samples to draw in costumers, and entice them to buy the product—Donnely (1994) indicated that 78% of surveyed packaged goods manufacturers distributed free samples. With such a large financial commitment to sampling, it is surprising to find that although some research has been conducted, there is not a solid base to support the practice.
In the first study, distribution of free samples was used to draw potential customers away from competitors. The data suggested, however, that this procedure was generally not effective. Nevertheless, a slight increase in customers was observed when samples were presented on Saturdays. Increased mall traffic, which occurs on the weekends, may have enhanced the difference between these conditions. Future research should compare sampling versus no sampling procedures in very high traffic shopping malls.
The failure of sampling to produce notable results in the current series of experiments calls this procedure into question. It is clear that more research is necessary to determine why some research has found sampling to be effective and other studies, including the current one, have found the procedure to be ineffective. However, it is impossible to conclude from these data whether receiving samples has an effect on future purchasing behavior. |
|
