Deviations From a Shopping Plan: When and Why Do Consumers Not Buy Items As Planned

Easwar S. Iyer, University of Massachusetts
Sucheta S. Ahlawat, University of Massachusetts
ABSTRACT - Past literature deals with deviations from a shopping plan only in terms of unplanned purchasing activity. However, shortfall, i.e., when a consumer does not buy items that s /he originally intended to buy, is an equally important variable. This study explores the possible effects of two mediators, i.e., Time Pressure and Prior knowledge of the Store Layout, on shortfall and also the reasons why it occurs.
[ to cite ]:
Easwar S. Iyer and Sucheta S. Ahlawat (1987) ,"Deviations From a Shopping Plan: When and Why Do Consumers Not Buy Items As Planned", in NA - Advances in Consumer Research Volume 14, eds. Melanie Wallendorf and Paul Anderson, Provo, UT : Association for Consumer Research, Pages: 246-250.

Advances in Consumer Research Volume 14, 1987      Pages 246-250

DEVIATIONS FROM A SHOPPING PLAN: WHEN AND WHY DO CONSUMERS NOT BUY ITEMS AS PLANNED

Easwar S. Iyer, University of Massachusetts

Sucheta S. Ahlawat, University of Massachusetts

ABSTRACT -

Past literature deals with deviations from a shopping plan only in terms of unplanned purchasing activity. However, shortfall, i.e., when a consumer does not buy items that s /he originally intended to buy, is an equally important variable. This study explores the possible effects of two mediators, i.e., Time Pressure and Prior knowledge of the Store Layout, on shortfall and also the reasons why it occurs.

INTRODUCTION

Grocery shopping can be best characterized as a routinized behavior. Even though consumers may enjoy the shopping experience, it is reasonable to state that grocery shopping is primarily viewed as a necessary chore that just needs to be done. To that extent, enhancing the efficiency levels of the information processing involved in executing that chore would be a desirable feature sought by consumers. In case of brand choice behavior, reliance on previously generated preferences, i.e., the affect referral heuristic (Wright 1975), would most likely increase efficiency. Similarly, the efficiency in executing the total shopping trip could be increased by following a well learned spatial script. Such scripts are products of enacting a behavior many times over. There are certain situational antecedents that would facilitate the retrieval and enactment of such spatial scripts. The purpose of this research is to define two such facilitators, i.e., Time Pressure and Prior Knowledge of the Store Layout, and study their effects on the deviations from a shopping plan.

THEORY OF SCRIPTED BEHAVIOR

The central theme in the theory of scripted behavior is that consumers carry many behavioral scripts in their cognitive system, and that such scripts determine the behaviors to be enacted in often-encountered environments (Abelson 1981, 1976). It is expected that individuals synthesize the essence of a behavioral setting in the form of a script which would enable subsequent reference to frequently repeated event sequences. Enactment of behaviors which are script driven require a minimal number of cues. Hence, it takes very few inputs to trigger off a scripted routine, thereby making it less effortful to enaCt such behaviors.

Various concepts such as plans (Miller, Galanter, and Pribram 1960), production systems (Anderson, Kline, and Beasley 1980), frames (Minsky 1975), schemata (Rumelhart and Ortony 1977), cognitive maps (Downs and Stea 1977), and mindlessness (Langer 1978; Langer, Balnk, and Chanowitz 1978) have been proposed in the cognitive science literature. All these share some meaning with each other to the extent that they all propose information-processing mechanisms which are selective and less effortful that guide behaviors. Although all of these are suggestive of habitual behaviors, enactment of scripted behavior is different from a habitual response. The key difference is that a script is a knowledge structure and not an automatic behavior response program. A script is like a network consisting of nodes and linkages. The strength of the linkages would vary from script to script as well as over time. Scripts probably get formed from the multiple learning experiences of individuals and hence familiarity is an important element in the process of script formation.

SCRIPT RETRIEVAL AND ENACTMENT

Mere presence of information (script availability) would not necessarily lead to its accessibility (automatic retrieval), and the latter would be affected by many factors (Lynch and Srull 1982). One such important factor may be the motivation to retrieve the script. It has been shown that the motivational state will not only affect the information availability through increased cognitive processing (Krugman 1965; Petty and Cacioppo 1979), but also its retrieval (Mitchell 1981). Therefore, the effect of a grocery shopping script on a consumer's grocery shopping plan would have to be examined in the context of facilitators that motivate the retrieval of a script. Time pressure is one such factor that would act as a catalyst in the process of script retrieval and enactment.

Time pressure can be conceptualized as a perceived constriction of the time available to an individual to perform a task. This constriction may facilitate the retrieval of information. Bronner (1982) found that the demand for additional information was drastically reduced if the decision were to be made under time pressure. Thus, it seems that individuals rely on available information, limited as it may be, in order to make a decision, rather than invest more time in gathering additional information which may or may not enhance the "correctness" of the decision itself. Such a trade-off reflects the need for efficient and quick processing when under time pressure. Based on these arguments, one could expect time pressure to be a facilitator in the retrieval of scripts.

However, the extent to which a script can be enacted would be a function of other environmental factors such as the degree of applicability of the encoded script to the problem-solving task on hand. For example, using a script that has been learned in a familiar environment while shopping in an unfamiliar environment would not increase the processing efficiency. Incompatibility between the incoming information and the encoded script inhibits the enactment of the script (Pitchert and Aderson 1977; Kozminsky 1977). This suggests that a situation that calls for efficient processing, i.e., time pressure, would facilitate the retrieval of a script and a familiar setting, i.e., prior knowledge of the store layout, would allow for its enactment (Iyer and Park 1986).

SCRIPTS AND PURCHASING BEHAVIOR

The relevant question, which forms the core of this study, pertains to the implications of a script-driven shopping behavior on purchasing behavior. Behavior that is driven by scripts is very likely to be in conformity with the norm. This is because scripts are most useful under conditions when efficient processing is required and very little attention needs to be paid to other cues in the environment. Thus, one can expect the deviations from a shopping plan to be minimal when purchasing behavior is driven by scripts. Invariably, deviation from a shopping plan has implied unplanned purchasing activity, i.e., items not on the plan that are purchased. This is probably because most of the earlier literature has been devoted to unplanned purchasing activity only (Kollat and Willet 1967; Raju 1980; Iyer and Park 1986). However, little or no attention has been devoted to the concept of shortfall. This is possibly because of the perception that influencing the amount of unplanned purchasing would lead to an increase in sales. However, shortfall is an equally important variable that influences sales. Shortfall is de- fined as the count of the items that a shopper intended to purchase but did not.

A trivial case of shortfall would be when the item intended for purchase was not available in the store. This is because the shopper has no option except not to purchase the item. There could be a variety of other reasons for shortfall, such as the price of the product being too high or the quality of the product being not acceptable, and so on. These may be justifiable reasons and to the extent they can be identified, the solutions available to store management in order to prevent a shortfall are also obvious. From a management point of view, the real crux of the matter lies in identifying other external situational factors that may influence a shopper not purchasing an item that s/he had originally planned to. In other words, if one were to eliminate those cases of shortfall caused due to the non-availability of products (since it is a trivial case), and further assume that the other "justifiable" reasons were randomly distributed across different situations, then the interesting question would be to study the varying frequency of occurrence of shortfall across different situations.

HYPOTHESES

Given sufficient time, a shopper is highly likely to fulfill his/her intended purchases. However, when shopping under time pressure, the shopper's behavior is expected to be script driven. Enactment of scripts requires very few informational inputs from the external environment and the dominant consideration is one of efficiency. In the process, a shopper is likely not to purchase all the items on her/his original list. Hence, the first hypothesis is proposed.

H1: The amount of shortfall will be a direct function of time pressure.

When shopping in a familiar environment, there is a high congruity between the encoded information about the store layout and the actual store layout. This facilitates the routine completion of a planned shopping sequence. However, the same encoded information will not be applicable to an unfamiliar shopping environment, which may introduce inefficiencies in the execution of a well-planned sequence thereby leading to the shopper not being able to buy all the items on her/his original list. Based on this discussion, the second hypothesis is proposed.

H2: The amount of shortfall will be an inverse function of prior knowledge of the store layout.

The above two hypotheses are consistent with the principles of resource allocation. The amount of time available for shopping and the cognitive capacity can be viewed as two of the many resources a consumer brings to a grocery shopping context. When a constraint is imposed on the time resource (e.g., shopping under time pressure) or when a demand is placed on the capacity resource (e.g., shopping in an unfamiliar store), the shopper is likely to adjust her/his purchasing strategy. As a result. some shortfall is likely to occur. Further, combining the above two hypotheses, it is clear that the following two hypotheses can be proposed.

H3a: The amount of shortfall will be highest for consumers shopping in an unfamiliar store under time pressure.

H3b: The amount of shortfall will be lowest for consumers shopping in a familiar store under no time pressure.

METHODOLOGY

Design

The results described in this paper came from a larger experiment consisting of a 2x2 factorial design. The two in dependent variables were time pressure and prior knowledge of the store layout. Time pressure was treated as either absent (A) or present (P), whereas prior knowledge of the store layout was treated as either low (L) or high (H).

Subjects

Subjects were randomly assigned to each of the experimental conditions. Most of the subjects belonged to a consumer panel affiliated to a large northeastern University. Certain others were recruited from the staff and students of the same university.

There were more females (78%) than males in the sample. The median income ($24,900) of the sample is somewhat higher than the national median ($18,000). in part, reflecting the relatively high educational level of the sampleC 23.5 percent had earned graduate degrees and 16.2 percent had completed college.

Experimental Task

The experimental task consisted of conducting a full replenishment grocery shopping trip. Subjects were picked up from their residence, escorted to the grocery store, and then dropped back at home. They paid for their purchases, but were Riven $10 for their participation.

Before departing to the store, the interviewer made note of the items that the subject planned to purchase. This was done by asking subjects to verbalize the items that they would purchase. By not asking the subjects to write town a shopping list and by posing the question almost as an afterthought, its salience was reduced. The subtle manner in which this question was handled avoided sensitizing the subjects. Subjects were also clearly informed that they were not bound by the list and could shop in their own manner and make any decisions. They were asked to estimate the time they required to complete the shopping for the items they indicated. The time estimate included the period from the time they entered the store to the time they came to the checkout line. This estimate formed the basis for the subsequent treatment of time pressure.

Even at this stage subjects were unaware of any potential experimental manipulation. Depending on the experimental condition, they were taken to their regular "favorite" store or to some other store they had never shopped at before. If it was a new store, it was always another store from the same chain. This ensured that the store layout was different but the range of products was similar. For those who had to shop under time constraints, instructions were provided to achieve the desired level of the treatment. The interviewer then provided a revised estimate of the reduced time that was available. The reduced time estimate was derived by halving the original time estimate already provided by the subject. An earlier pilot study had revealed that halving the time estimate induced time pressure without creating a panic situation.

At the end of the shopping trip, when the subject was ready to check out, s/he was asked to respond to scaled questions on the amount of time pressure s/he felt and her/his perceived familiarity with the store layout. Then, an item by item inventory of the shopping cart was conducted and compared to the original list provided by the subject. This inventory count was used to identify shortfall, and in all cases of deviations, the subject was asked to provide a reason.

After the task was completed, those who were under time pressure were offered extra time to go on their own and add/delete to/from their existing purchases. No record was kept of this activity since the subject was allowed to do this by her/himself. Not many subjects acted on this offer. Similarly, those who were taken to an unfamiliar store were given the option of replacing all or part of their purchases coupled with an offer to take them to their "favorite" store. Only one subject took this option. Subjects were then escorted home.

RESULTS

Manipulation Checks

The actual time taken by shoppers in both the time-pressure-present (TPP) and time-pressure-absent (TPA) conditions was noted to the nearest minute. The mean times of 13.58 minutes and 31.11 minutes for the two groups are significantly different (t-7.50, p<0.001). After completing the shopping trip, subjects were asked to rate the amount of time pressure they felt during the shopping trip on a 7-point scale (1=No time pressure at all, 4=Moderate time pressure, 7=Very high time pressure). The difference between the two group means of 5.91 and 1.37 was significant (t=16. 11, p<0.001).

On a 12-point scale [The 12-point scale used was as follows: 0=Never shopped, 1=1-5 times, 2=6-10 times, 3-11-20 times, 4=21-30 times, 5-31-50 times, 6=51-75 times, 7=76-100 times, 8=101105 times. 9-151-200 times, 10-201-300, 11=301 or more.] subjects were asked to estimate the number of times they had shopped in the store chosen for the experimental purposes. Those in the low-prior-knowledge (LPK) treatment had never shopped in that particular store, and hence the mean response for that group was zero. On the same scale, the high-prior-knowledge (HPK) group had a mean score of 8.84 (101-150 times). These two means are significantly different (t-20.11, p<0.001). After the shopping trip was completed, subjects were asked to rate their perceived familiarity with the store layout on a 7-point scale (1=No knowledge at all, 4-Moderate knowledge, 7-Very high knowledge). The two group means were 1.67 and 6.06; these are significantly different (t-13.78, p<0.001).

Hypotheses Testing

Shortfall was defined as a count of the items that a shopper intended to purchase but did not. This was operationalized as those items on a shopper's original list that were not in the shopping cart at the check-out. Implied in this operationalization is the assumption that the verbal report of the items provided before going to the grocery store represents the shopper's intentions.

Time pressure and prior knowledge of the store layout were the two independent factors, each of which had two levels. Since the actual time taken for shopping was expected to have a direct bearing on the total number of items purchased, it was used as a covariate. Shortfall-defined and measured in the above manner--was used as a dependent variable in an ANCOVA. There was a slight possibility of order of entry problem because of unequal cell sizes. The analysis was rerun by varying order of entry of the independent variables. There were no differences in the results obtained.

Actual shopping time, which was used as a covariate, was highly significant (p<0.002) and its coefficient (-0.055) was in the expected direction. Other important aspects of the ANCOVA results are reported in Table 1. There was no significant interaction effects, but the main effects of both the independent variables are significant (p<0.06 or better), therebY supporting hypotheses 1 and 2 respectively.

TABLE 1

RESULTS OF ANCOVA

A planned pairwise comparison [All planned pairwise comparisons were performed using a one-tailed t-test with the mean sum of square errors from the overall analysis of variance.] of the mean shortfall in the LPK-TPP condition (3.63) with the mean shortfall in all other conditions was conducted. The difference between the mean shortfall in this condition and those in the LPK-TPA condition (1.50), EPK-TPA condition (0.73), and HPK-TPP condition (2.65) were all significant at p<O.05 level or better. This clearly supports hypothesis 3a. In accordance with the last hypothesis, the mean shortfall in the HPK-TPA condition (0.73) was compared with those in the HPK-TPP condition (2.65) and LPK-TPP condition (3.63). The differences were significant at p<O.005 level or better. However, when compared to the shortfall in the LPK TPA condition (1.50), the difference was only modestly significant (p<O.10). This modestly supports hypothesis 3b. Overall, one could state that all the hypotheses were reasonably well supported.

DISCUSSION

Shortfall--when a shopper does not buy items on her/his plan--is an important phenomenon, since it represents a lost opportunity to conclude a sale, which from the buyer's viewpoint was all but concluded. In spite of its importance, it is an overlooked phenomenon in the consumer shopping literature, and most of the previous research has focused only on unplanned purchasing behavior.

As was ewidenced from this study, shortfall need not occur only when there are stock-outs. In fact, non-availability of an item was not the most frequent reason for shortfall; its frequency was a little over 11 percent of the total pool of reasons (see Table 2). However, in a grocery shop ping situation where repeat purchases are an important objective of overall marketing strategy, the total cost of lost sales as a result of stock-out can be extremely high. Situational parameters such as time pressure and knowledge of the store layout seemed to have some influence over shortfall.

In order to understand the nature of the influence of time pressure and knowledge of the store layout on shortfall, the reasons for shortfall were further analyzed. There were a total of fourteen reasons advanced by the shoppers; these were cited 162 times when all the responses were pooled. Seven of these reasons constituted approximately 85 percent of the total responses. The detailed breakdown of these reasons under the different experimental conditions is provided in Table 2. Space constraints prevent a more detailed analysis of its presentation.

It is obvious, nonetheless interesting, that no shopper in the TPA condition suggested that lack of time was a reason for shortfall. Equally interesting was the fact that memory failure, as a reason for shortfall, was overwhelmingly mentioned by shoppers in the TPP condition. This seems to strongly indicate that time pressure is not merely a constriction in the time available, but also seems to affect information retrieval. In other words, time pressure does not allow the retrieval of the detailed information base stored in the long-term memory; rather, it seems to switch the individual into a scriptal mode whereby holistic action sequences could be retrieved. In the process, the efficiency of the processing is enhanced but at the cost of accuracy. There is further reason to believe that the TPP condition calls for enhanced efficiency because those under the condition seem to drop their tentative plans. In other words, shoppers in the TPP condition seem to switch to a scriptal mode. One of the steps that a script adopts in the drive for increased efficiency is to overlook the goals that are tentative.

One potential reason for shortfall could be that the item was not available. However, it is possible that the specific item was really available, except that it was in a location that the shopper did not search for. In other words, these two reasons could be merely manifestations of an underlying attribution process, which itself could be a function of the shopper's perceived self-confidence in their knowledge of the store layout. Thus, shoppers in the HPK condition more often cited non-availability as a reason for shortfall than those in the LPR condition. The results were exactly the opposite with shoppers in the LPK condition citing their inability to locate as a reason for shortfall more often than those in the HPK condition. This distinction is important because in one case the shopper attributes the cause for shortfall to oneself whereas in the other case the store is viewed as the cause. In the long run, the latter attribution could hurt the image of the store.

TABLE 2

REASONS FOR SHORTFALL

FIGURE 1

TIME PRESSURE

FIGURE 2

PRIOR KNOWLEDGE

REFERENCES

Abelson, Robert P. (1981),"Psychological Status of the Script Concept," American Psychologist, 36, pp 715-729.

Abelson, Robert P. (1976),"Script Processing in Attitude Formation and Decision Making," in J.S. Carroll and J.W. Payne (eds.), Cognition and Social Behavior, Hillsdale, NJ: Erlbaum.

Anderson, John R., Paul J. Kline, and Charles M. Beasley (1980),"A General Learning Theory and Its Application to Schema Abstractions," in Gordon Bower (ed.) The Psychology of Learning and Motivation: Advances in Research and Theory, New York, NY: Academic Press, pp 277-316.

Bronner Rolf (1982) Decision Making Under Time Pressure: An Experimental Study of Stress Behavior in Business Management, Lexington, MA: Lexington.

Down, Roger M. and David Stea (1977), Maps in Minds: Reflections on Cognitive Mapping, New York, NY: Harper.

Iyer, Easwar S. and C. Whan Park (1986), "Effects of Time Pressure and Prior Knowledge of Store Environment on Grocery Purchasing Behavior," Working Paper, University of Massachusetts, Amherst, MA.

Kollat, David T. and Ronald P. Willet (1967), "Customer Impulse Purchasing Behavior," Journal of Marketing Research,

Kozminsky, Edward (1977),"Altering Comprehension: The Effect of Biasing Titles on Text Comprehension," Memory and Cognition, 5, p 482-490.

Krugman, Herbert E. (1965), "The Impact of Television Advertising: Learning without Involvement," Public Opinion Quarterly, 29, p. 349-356.

Langer, Ellen (1978), "Rethinking the Role of Thought in Social Interaction," in J. Harvey, W. Ickes, and R. Kidd (eds.) New Directions in Attribution TheorY. Vol. 2, Hillsdale, NJ: Erlbaum.

Langer, Ellen, Arthur and Benzion Chanowitz (1978), "The Mindlessness of Ostensibly Thoughtful Action: The Role of 'placebic' information in interpersonal interaction," Journal of Personality and Social Psychology, 36, pp 635

Lynch, John G. Jr. and Thomas R. Srull (1982), "Memory and Attentional Factors in Consumer Choice," Journal of Consumer Research, 9, p 18-37.

Miller, G.A., E. Galanter, and K.H. Pribram (1960), Plans and the Structure of Behavior, New York, NY: Holt Rinehart.

Minsky, Marvin (1975), "A Framework for Representing Knowledge," in Peter Winston (ed.) The Psychology of Computer Vision, New York, NY: McGraw-Hill, pp 211-280.

Mitchell, Andrew (1981), "The Dimensions of Advertising Involvement," K. Monroe (ed.), Advances in Consumer Research, Chicago, IL: Association for Consumer Research, pp 25-30.

Petty, Richard and John Cacioppo (1979), "Issue Involvement Can Increase or Decrease Persuasion by Enhancing Message or Relevant Cognitive Responses," Journal of Personality and Social Psychology, 37, pp 1915-1926.

Pitchert, John W. and R.C. Anderson (1977), "Taking Different Perspectives on a Story," Journal of Educational Psychology, 69, pp 309-315.

Raju, P.S. (1980), "Optimum Stimulation Level: It's Relationship to Personality, Demographics, and Exploratory Behavior," Journal of Consumer Research, 7, pp 272-282.

Rumelhart, David E. and A. Ortony (1977), "The Representation of Knowledge in Memory," in R.C. Anderson, R.J. Spiro, and W.E. Montague (eds.), Schooling and the Acquisition of Knowledge, Hillsdale, NJ: Erlbaum, pp 99-148.

Wright, Peter (1975), "Consumer Choice Strategies: Simplifying vs. Optimizing," Journal of Marketing Research, 11, pp 60-67.

----------------------------------------