A Study of Consumer Dissatisfaction Before Purchase

Robert A. Westbrook, Duke University
ABSTRACT - Little is known currently about the level of satisfaction or dissatisfaction the American public experiences in their affairs as consumers. This paper reports recent findings on the extent of consumers dissatisfaction during the pre-purchase period based on a survey of prospective buyers of major durables. Findings suggest consumers may not be as discontent as often assumed.
[ to cite ]:
Robert A. Westbrook (1977) ,"A Study of Consumer Dissatisfaction Before Purchase", in NA - Advances in Consumer Research Volume 04, eds. William D. Perreault, Jr., Atlanta, GA : Association for Consumer Research, Pages: 142-148.

Advances in Consumer Research Volume 4, 1977   Pages 142-148


Robert A. Westbrook, Duke University


Little is known currently about the level of satisfaction or dissatisfaction the American public experiences in their affairs as consumers. This paper reports recent findings on the extent of consumers dissatisfaction during the pre-purchase period based on a survey of prospective buyers of major durables. Findings suggest consumers may not be as discontent as often assumed.


Interest is rapidly growing in the assessment of consumer satisfaction, both as a component of the larger overall quality of life (Andrews and Withey, 1975), as well as a topic in its own right (Leavitt, 1975; Czepiel et. al., 1974). Currently available research evidence is limited, in part because the overall topic of consumer satisfaction encompasses a wide variety of different issues, and in part because conceptual and measurement problems have deterred intensive study. Notwithstanding the shortage of factual data, opinions abound that the public is broadly dissatisfied with the modern market place and its offerings of goods and services (Barber, 1966; Holton, 1967; Aaker and Day, 1974; Kotler, 1972).

In spite of the emerging interest in consumer satisfaction, there is no generally accepted definition of the term; nevertheless, the words do seem to have roughly uniform semantic usage. Broadly, consumer satisfaction refers to the relative "goodness" of the subjective experiences accompanying an individual's consumer behavior. Similarly, consumer dissatisfaction is taken to indicate the degree of unfavorability of an individual's experiences associated with his or her behavior.

Thus understood, consumer satisfaction and dissatisfaction are widely inclusive concepts. Not only do they subsume the full extent of an individual's Product and service purchases, but they also include markedly different aspects of consumer behavior--namely, product acquisition, consumption and disposition. In other words, for any given purchase, consumers may be satisfied or dissatisfied to varying degrees in connection with each of the separate activities of acquiring the product, using or consuming its benefits, and disposing of it.

Much of the published research bearing on consumer satisfaction and dissatisfaction either fails to make these potentially significant distinctions or addresses the topic solely in connection with post-purchase product usage (Cardozo, 1965; Anderson, 1973; Pfaff, 1972). Few studies have explicitly considered satisfaction or dissatisfaction arising during the pre-purchase period of consumer behavior, i.e. when consumers are actively engaged in decision processes leading up to a final purchase choice. As a result, very little is known about this important facet of consumer affairs.

Studies of consumer shopping behavior have noted that this activity may be enjoyable in its own right (Rich and Portis, 1963; Tauber, 1972). Katona and Mueller (1954) reported that roughly twice as many appliance buyers liked to shop with care as did not. These results are suggestive of pre-purchase satisfaction, at least during period of shopping.

Friedman (1965), however, reported that a sizable proportion of consumers in a field experiment were unable to identify the most economical package size across a variety of grocery product categories. While the results are suggestive of consumer confusion, the research did not conclusively establish whether consumers actually experienced confusion or other felt dissatisfactions during the process of making purchase choices.

Studies of public opinion on consumerist issues and contemporary marketing practices (Barksdale and Darden, 1972; Hustad and Pessemier, 1973) have suggested widespread consumer dissatisfaction. Such studies are rarely made specific to particular products or purchases actually made by the respondent; rather, respondents typically indicate their agreement or disagreement with various attitude statements. Findings may therefore not reflect satisfactions or dissatisfactions actually experienced by consumers during their own pre-purchase activity.

Cross sectional studies of consumer complaints indicate varying levels of dissatisfaction. One study, for example reported that some 35% of a national probability sample felt upset about their treatment as consumers (Wetland, 1974). Another study based on a convenience sample in a large Southeastern city found that 24% of these surveyed had attempted to obtain redress for some dissatisfaction in connection with durable goods (Thomas and Shuptrine, 1975). While adding valuable insight, such studies of complaints do not focus on the quality of consumers' various experiences prior to purchase. Rather, they may indicate the more salient negative points as recalled for that period.

In addition to limited present knowledge about the extent of consumer satisfaction/dissatisfaction, little is known about the factors or personal characteristics with which pre-purchase satisfaction/dissatisfaction is associated. Studies of public attitudes toward business practices and consumerist issues have observed some relation between such attitudes and various demographic characteristics (Hustad and Pessemier, 1973; Barksdale and Darden, 1972; Coulson, 1971). These findings raise the possibility that pre-purchase satisfaction may vary along demographic lines.

Other potential correlates of pre-purchase satisfaction/dissatisfaction are suggested by the research of Jacoby, Speller, and Kohn (1974), who reported that measures of the subjective states of consumers in a brand choice task were significantly affected by the experimental manipulation of information load. In particular, greater information loads were associated with more favorable subjective reports about the ease of choosing, reduction of uncertainty, etc. These findings, while borrowed from the context of a laboratory experiment, suggest that satisfaction with pre-purchase activity may be related to the amount of information acquired through search activity.

In sum, relatively little is known about the actual incidence of satisfaction or dissatisfaction during the period prior to purchase, as well as the factors which explain its occurrence. Most of what is known is based on recall after purchase rather than on reports of persons at the time of pre-purchase activity; such may or may not cloud the true picture. Accordingly, this study was conceived to investigate the nature of consumers' pre-purchase experiences as they were occurring at the time. Using this data, the study sought to 1) identify the proportion of consumers experiencing different levels of satisfaction or dissatisfaction prior to purchase, and 2) search for potential explanatory factors to account for differences in reported satisfaction/dissatisfaction.

The scope of the research in this study was limited to a single product category in order to afford the opportunity for more detailed investigation than would otherwise be possible. While this obviously reduces the generality of the results, it also enables substantially improved insights into the nature of consumers' actual buying experiences. Major household appliances were chosen as the product category since they typify purchase decisions in which substantial dissatisfaction might be expected (Holton, 1967).


Data for the study were obtained from interviews with a sample of 158 prospective buyers of major household appliances located in the suburban areas of Detroit, Michigan during the fall of 1974. For the purposes of the study, prospective buyers were defined as persons stating plans to purchase one or more appliances within two months from the date of initial contact. Appliances included were automatic washing machines, clothes dryers, range/ovens, refrigerators, freezers, and dishwashers.

Sample Design

The low incidence of prospective appliance purchasers in the general population cross-section precluded a simple probability sample of adults from the study area. [This was determined during pilot testing in which a random sample of area telephone subscribers was screened for intent to purchase. Twenty interviews were completed using finalized questionnaires and were included in the total of 158.] Respondents were instead sampled from the clientele of the major retail appliance outlets in the study area. Selection was accomplished in two ways: (1) an interviewer personally intercepted shoppers departing from the appliance display section of each of four leading retail outlets, and 2) shoppers at the remaining major retail outlets were identified by tracing the registered ownership of vehicles parked outside these outlets. The first selection method contributed 69% of the total respondents, reflecting the heavier customer traffic at the respective retail outlets. For obvious reasons of efficiency, respondents were sampled at times of peak shopper traffic, i.e. evenings and weekends.

Demographic characteristics of the sample respondents are shown in Table 1. For purposes of comparison, this sample is on the average somewhat earlier in the family life cycle, better educated, higher in occupational status, and more affluent than a national probability sample of appliance buyers interviewed in 1968 (Staelin, 1969). These differences are attributed mainly to the limitation of the present sample to the suburban areas of a major city; variations such as these are generally in accord city-suburban contrasts as reported by U.S. Census data.

Sample respondents were screened to ensure that they were the individuals in the household principally responsible for selecting the brand, model, and features of the appliance. Eighty percent were married female heads of household and another six percent were single female heads.

Measurement of Consumer Dissatisfaction

The measure of consumer dissatisfaction in this study was based on the extent to which respondents reported unfavorable experiences or sentiments in connection with their pre-purchase decision activity.



Since pre-purchase behavior for major household durables is characterized by considerable diversity and complexity (Katona and Mueller, 1954; Newman and Staelin, 1972), respondents were questioned about a variety of different aspects of their purchase decisions representing potential sources of dissatisfaction. The area questioning included:

1) How respondents felt about the selection of appliances from which to choose--too many, too few, or satisfactory.

2) Whether or not respondents expected to be able to find an appliance meeting their desires.

3) How respondents felt about visiting retail outlets to shop for appliances-dislike, neutral, enjoyment.

4) How respondents felt about deliberations to select appliances features and specifications--dislike, neutral, or enjoyment.

5) Whether respondents encountered difficulty in trying to evaluate the performance qualities of alternative brands.

6) Whether respondents encountered difficulty in trying to evaluate the cost of optional product features and specifications.

7) Whether respondents felt more product information would help them to make a better purchase decision.

8) How much respondents vacillated in their decision making.

9) Whether respondents worried about the outcome of their purchase decisions.

10) Whether respondents felt rushed and overly hurried in the course of their purchase decisions.

All respondents were interviewed via telephone using trained professional interviewers. Since most respondents had been obtained by in-store intercept, these persons also had the benefit of personal contact, thus encouraging full and frank response.

An index measure of dissatisfaction was then constructed by assigning one point to each response to the preceding questions indicating an unfavorable sentiment or experience. Prior to combining items in this manner, however, inter-item relationships were examined to verify that only items with positive associations were entered into the index (Lansing and Morgan, 1971).

Tests of the validity of the constructed measure of dissatisfaction are not possible due to the absence of other accepted measures of consumer dissatisfaction. However, the measure appears to have at least some content validity. Moreover, evidence presented below reveals the majority of the component measures to be positively intercorrelated.

Explaining Pre-Purchase Dissatisfaction

Various factors were examined for their potential to explain dissatisfaction during pre-purchase activity. Candidate variables were drawn from the limited prior research evidence as well as common sense. They fall into five groups: 1) demographic characteristics 2) personality attributes 3) previous ownership and purchasing experience 4) pre-purchase information search activity, and 5) timing of the purchase decision process. A detailed list of the specific variables in each group appears in Figure 1.

The search procedure to identify the strongest predictors of dissatisfaction employed the AID3 algorithm (Sonquist, Baker, and Morgan, 1971), which produces a sequence of binary splits on the total sample of respondents based on the combinations of levels of the independent variables that best reduce unexplained variation in the dependent measure. The results of this procedure indicate which of the independent variables examined are most useful for segmenting the sample on the basis of homogeneity with respect to pre-purchase dissatisfaction. These methods are appropriate when the goal of research is to find variables related to the phenomenon in question short of exact testing of specific hypotheses, as when relevant theory is lacking.




The incidence of unfavorable sentiments or experiences for each of the ten items dealing with potential areas of pre-purchase dissatisfaction is shown in Table 2. While the level of negative response varies by item, on no single item did a majority of respondents give indication of dissatisfaction. However, nearly half of the prospective buyers (49.4%) reported some difficulty in judging the value of optional product features (e.g. self-defrosting refrigerators, self-cleaning range/ovens, etc.). Furthermore, over one-third of the sample (33.5%) expressed a need for more information than currently available to buyers.

Table 3 shows the extent of association between items in terms of Kendall's tau-b. While the magnitudes of the taus are not great, all but one of the items are positively related to at least one other item at a level of significance. Few of the relationships are negative, and none at a level of significance. These findings indicate some tendency for unfavorable feelings on different items to be associated with each other, which allows the separate items to be combined meaningfully into an overall index (Lansing and Morgan, 1971).



The distribution of the index measure of consumer dissatisfaction is shown in Table 4. This may also be viewed as a simple count of the number of individual question items to which respondents gave some indication of negative feeling. The average index value for all respondents is 2.28, with a standard deviation of 1.70. Of particular note is the finding that over 75% of the respondents give two or fewer individual mentions of dissatisfaction.

Highlights of the AID3 analysis to identify the strongest predictors of pre-purchase dissatisfaction appear schematically in Figure 2. Splits were based upon the independent variables having the highest BSS/TSS ratio, subject to a minimum constraint (BSS/TSS $ 0.8%) and a minimum resulting group size (N $ 20). Splits were also constrained to adjacent predictor categories.

It is evident that the AID procedure succeeded in splitting the sample into segments differing substantially in average values on the dissatisfaction index.

The first split was on amount of information seeking yet planned by respondents. Prospective buyers planning little or moderate amounts of additional search were grouped together (group 2) and revealed a mean dissatisfaction index of 2.08. Buyers planning a great deal of search (group 3) had a mean index of 3.48. This binary split is the best possible among all predictors, yielding a BSS/TSS ratio of .085, over twice that of the nearest competing predictor.



The remaining partitions of the sample are detailed in Figure 2. In addition to the amount of planned information seeking, other important explanatory variables are the index of personal competence of the respondent, occupation of head of household, number of retail stores already visited, and number of other types of major appliances bought previously.

The highest mean pre-purchase dissatisfaction score (3.48) was observed for respondents yet planning a great deal of search activity (group 3). The next highest average (3.28) was observed for respondents planning a low or moderate amount of search, but who are relatively low in personal competence, have already shopped 3 or more retail stores and have bought relatively few (2 or less) other major appliances in the past (group 9). The lowest mean dissatisfaction score (1.35) is observed for respondents who do not plan high levels of additional search, are high in personal competence, and have heads of household relatively lower in occupational status, i.e. blue collar and lower white collar jobs (group 10).

Results of a one-way analysis of variance on the six final groups identified in the AID analysis (3, 10, 11, 6, 8, 9) are shown in Table 5. They indicate that the explanatory variables used to split the sample were able to account for some 19.2% of the total variation in dissatisfaction index scores (BSS/TSS). These results are statistically significant.

Examination of variables on which splits almost occurred, i.e. whose BSS/TSS ratios were fairly close to that of the predictor on which the split was made, reveals three additional explanatory factors of potential consequence: 1) whether information was sought from Consumer Reports magazine, 2) satisfaction with the previously owned appliance (if any), and 3) duration of the purchase decision process up to the time of interview.






The results of the survey suggest that dissatisfaction during the pre-purchase period is limited. Two thirds of the prospective appliance buyers interviewed gave two or fewer indications of discontent. Mention of more than five was rare. When specific areas of dissatisfaction are considered, the results still do not indicate extensive consumer unhappiness. Fewer than a quarter of the respondents were displeased with the market selection of appliances from which to choose. Even less reported the task of evaluating different brands of appliances as difficult. Nor were there many individuals who experienced worry about the outcome of their purchase deliberations. The great preponderance of consumers report making up their minds about their purchase without vacillation. If buyers are truly confused because of product proliferation and groping in their choices because of incomplete information, one would clearly not expect to observe results such as those described.

This is not to say, however, that prospective buyers are totally satisfied. Many of the appliance buyers in this study noted their difficulty in judging the desirability of optional product features in light of their added cost. Whether this represents a serious pre-purchase problem is debatable. Most respondents also point out that the process of deliberating about such features is not unpleasant.

Some one third of the respondents acknowledged the need for more information than was currently available to buyers. It is unclear from the results whether this report reflects a fairly realistic appraisal of market information sources, or simply temporary feelings because the individual might not have yet completed his purchase decision process. Nevertheless, there are few persons who corroborate this information need by also indicating difficulty judging brands, extensive vacillation, worry, etc.

How well these findings mirror the level of pre-purchase dissatisfaction experienced by the general buying public is not known. There are several reasons why the limited levels of dissatisfaction noted here may not accurately reflect the national cross-section. First, the respondents interviewed were drawn from a limited geographic area, the suburbs of a major metropolitan area in the Midwest. These persons are notably different than rural or exclusively urban populations. The direction of any bias resulting from the sampling limitation is unclear, however. There is some evidence to suggest that over-representation of the more educated and affluent may overstate the incidence of dissatisfaction (Hustad and Pessemier, 1973).

Second, by sampling respondents from retail outlets, there is some tendency to over-represent "heavy" shoppers, since they are accorded more opportunities to get chosen than persons who shop only sparingly. Furthermore, respondents were chosen only at times of peak retail floor traffic (weekday evenings and all day during weekends); persons who shop predominantly at other times are accordingly under-represented in the sample. These effects may have biased the observed incidence of dissatisfaction downward, so that it is understated compared to the total population. Whether such actually occurred is, of course, unknown.

Finally, it is obvious that results for a single class of products may not reflect the entire spectrum of consumers' buying experiences. Major household appliances were chosen since there were considerable a priori reasons for suspecting extensive dissatisfaction. While relatively few indications of discontent were found, the same may not be true of other durables, as well as non-durables and intangible services.

To what factors is pre-purchase dissatisfaction related? Several were identified in this study and should serve to guide future research. Evidence of a relationship was found between information search plans and dissatisfaction; the more extensive the consumer's plans for search, the greater the reported dissatisfaction. It is unlikely that search plans "cause" dissatisfaction, however, even if they are able to account for substantial variation in the dependent variable. More likely, search plans are a proxy for some basic psychological characteristic which mediates the experience of dissatisfaction in the buying situation. One candidate is risk aversion, which might cause the person to feel uncomfortable with various aspects of the purchase decision, and at the same time lead him/her to attempt to relieve the unpleasantries through further information seeking.

Pre-purchase dissatisfaction was also seen to be related to actual search behavior. Higher average dissatisfaction scores were noted for persons reporting 3 or more store visits and for those reporting usage of a specialized product information source--Consumer Reports. As noted above, the direction of causal effects is unclear. Nevertheless, the positive relationship between search and dissatisfaction is surprising in light of findings by Jacob y et. al. (1974) that more favorable subjective states were produced by heavier information loads. Perhaps extended search activity is not required to yield high information load when consumers deliberate and choose in natural purchase settings.

Pre-purchase dissatisfaction was seen to be related to the. nature of the consumer's personality. Individuals characterized by high levels of personal competence, or ego-strength, tended to report fewer dissatisfactions. Andrews and Withey (1975) found similar results on satisfaction with the overall quality of life. These results are intuitively appealing. Persons characterized by greater control over their destinies and effectiveness in life are not likely to permit themselves to experience (or even acknowledge) great upset during buying activities, nor in other personal affairs, for that matter.

Interestingly, none of the demographic variables excepting occupation was especially useful for categorizing respondents in terms of dissatisfaction. The absence of education as a substantial explanatory variable is puzzling and there is no ready rationale why occupation should be related positively to dissatisfaction.

None of the explanatory variables alone is able to account for much of the variation in reported pre-purchase dissatisfaction. Moreover, the best five variables when used to group respondents into maximally different segments enable only a fraction of the total variation to be accounted for. It is evident that further research is needed, both to reduce measurement error which attenuates explanation and to identify new relationships of consequence.

This study has attempted to take a first step in developing a workable measure of pre-purchase dissatisfaction and exploring a number of potential explanatory factors. Several interesting findings have emerged, but the interpretation of these results must remain cautious in view of the limitation in sample size and location. The findings of the AID3 analysis, in particular, are susceptible to small sample variability.


David A. Aaker and George S. Day, eds., Consumerism: Search for the Consumer Interest, 2nd ed. (New York: Free Press, 1974), pp. xxi-xxvi.

Ralph E. Anderson, "Consumer Dissatisfaction: The Effect of Disconfirmed Expectancy on Perceived Product Performance," Journal of Marketing Research, 10 (February 1973): pp. 38-44.

Frank M. Andrews and Stephen B. Withey, "Developing Measures of Perceived Life Quality," Social Indicators Research, 1 (1974).

Richard J. Barber, "Government and the Consumer," Michigan Law Review, 64 (May 1966): 1226ff.

Hiram C. Barksdale and William R. Darden, "Consumer Attitudes Toward Marketing and Consumerism," Journal of Marketing, 36 (October 1972), pp. 28-35.

Angus Campbell, Philip E. Converse, W. E. Miller, and D. E. Stokes, The American Voter (New York: John Wiley and Co., 1960), p. 517.

Richard N. Cardozo, "An Experimental Study of Consumer Effort, Expectation, and Satisfaction," Journal of Marketing Research, 2 (August 1965): 244-49.

John S. Coulson, "New Consumerists' Breed Will Fade Away," Marketing News, Mid-June 1971, p. 1ff.

John A. Czepiel, Larry J. Rosenberg, and Adelbayo Akerele, "Perspectives on Consumer Satisfaction," "Proceedings of the Fall Conference of the American Marketing Association (Chicago: American Marketing Association, 1974), pp. 119-123.

Ralph Day and E. Laird Landon, "Collecting Comprehensive Consumer Complaint Data by Survey Research," Advances in Consumer Research, Vol. 3 (College Park, Md.: Association for Consumer Research, 1976), pp. 263-68.

Monroe P. Friedman, "Consumer Confusion in the Selection of Supermarket Products, " Journal of Applied Psychology, 50 (1966): 529-534.

Richard H. Holton, "Government-Consumer Interest: Conflicts and Prospects (The University Point of View)," in Chart in Marketing Systems, ed. Reed Moyer (Chicago: American Marketing Association, Winter 1967), pp. 15-17.

Thomas P. Hustad and Edgar A. Pessemier, "Will the Real Consumer Activist Please Stand Up: An Examination of Consumers' Opinions about Marketing Practices," Journal of Marketing Research, 10 (August 1973), pp. 319-324.

Jacob Jacoby, Donald E. Speller, and Carol Kohn, "Brand Choice Behavior as a Function of Information Overload," Journal of Marketing Research, 11 (February 1974): 63-69.

George Katona and Eva Mueller, "A Study of Purchase Decisions," in Lincoln H. Clark, ed., Consumer Behavior: The Dynamics of Consumer Reaction, Vol. 1 (New York: New York University Press, 1955), pp. 30-87.

Philip Kotler, "What Consumerism Means for Marketers," Harvard Business Review, 50 (May-June 1972): 48-57.

John B. Lansing and James N. Morgan, Economic Survey Methods, (Ann Arbor, Michigan: Institute for Social Research, 1971), p. 288.

Clark Leavitt, "Consumer Satisfaction as the Ultimate Life Force," Advances in Consumer Research, Vol. 3 (College Park, Maryland: Association for Consumer Research, 1976), pp. 252-258.

Joseph W. Newman and Richard Staelin, "Prepurchase Information Seeking for New Cars and Household Appliances,'' Journal of Marketing Research, 9 (August 1972): 249-57.

Anita B. Pfaff, "An Index of Consumer Satisfaction," Proceedings of the 3rd Annual Conference of the Association for Consumer Research, (College Park, Maryland : Association for Consumer Research,1972)

Stuart U. Rich and Bernard Portis, "Clues for Action from Shopper Preferences," Harvard Business Review, 41 (March-April, 1963): 147.

John P. Robinson and Philip R. Shaver, Measures of Social Psychological Attitudes (Ann Arbor, Michigan: Institute for Social Research, 1969), pp. 102-105.

John A. Sonquist, Elizabeth L. Baker, and James N. Morgan, Searching for Structure, Rev. Ed. (Ann Arbor, Michigan: Institute for Social Research, 1973).

Richard Staelin, "Information Seeking and Duration of the Purchase Decision Process for New Cars and Major Household Appliances," (unpublished Ph.D. dissertation, University of Michigan, Ann Arbor, Michigan, 1969).

Edward M. Tauber, "Why Do People Shop?", Journal of Marketing, 36 (October 1972): 46-59.

William R. Thomas and F. Kelley Shuptrine, "The Consumer Complaint Process: Communication and Resolution," Business and Economic Review, 21 (June, 1975): 13-22.

Rex H. Warland, Robert O. Herrman, and Jane Willits, "Dissatisfied Consumers: Who Gets Upset and What They Do About It," Unpublished manuscript, Pennsylvania State University, November, 1974.