Some Limits on the Potency of Word-Of-Mouth Information

ABSTRACT - Past research documents the substantial impact of word-of-mouth (WOM) information on the product evaluations and purchase intentions of potential buyers, especially if the products are novel and unbranded. This study examines the potency of WOM information if the potential buyer expresses an evaluative position regarding the product prior to exposure to WOM information. Our results indicate that individuals' receptivity to both positive and negative WOM information is determined largely by its "fit" with their (prior) evaluative position. These results suggest that understanding the influence of WOM information on the acceptance or rejection of new products may be greatly improved by understanding the evaluative predispositions of potential buyers in conjunction with the valence of the WOM information.

Citation:

William R. Wilson and Robert A. Peterson (1989) ,"Some Limits on the Potency of Word-Of-Mouth Information", in NA - Advances in Consumer Research Volume 16, eds. Thomas K. Srull, Provo, UT : Association for Consumer Research, Pages: 23-29.

Advances in Consumer Research Volume 16, 1989      Pages 23-29

SOME LIMITS ON THE POTENCY OF WORD-OF-MOUTH INFORMATION

William R. Wilson, Rice University

Robert A. Peterson, University of Texas - Austin

ABSTRACT -

Past research documents the substantial impact of word-of-mouth (WOM) information on the product evaluations and purchase intentions of potential buyers, especially if the products are novel and unbranded. This study examines the potency of WOM information if the potential buyer expresses an evaluative position regarding the product prior to exposure to WOM information. Our results indicate that individuals' receptivity to both positive and negative WOM information is determined largely by its "fit" with their (prior) evaluative position. These results suggest that understanding the influence of WOM information on the acceptance or rejection of new products may be greatly improved by understanding the evaluative predispositions of potential buyers in conjunction with the valence of the WOM information.

SOME LIMITS ON THE POTENCY OF WORD-OF-MOUTH INFORMATION

There is a sizable body of evidence which suggests that word-of-mouth (WOM) information has a substantial impact on the product evaluations and purchase intentions of potential buyers (e.g., Brown and Reingen 1987; Katona and Mueller 1955; Weinberger and Dillon 1980). Furthermore, the influence of negative WOM information, compared to positive WOM information, has been found to be especially potent (Arndt 1967; Lutz 1975; Mahajan, Muller, and Kerin 1984; Mizerski 1982; Weinberger, Allen and Dillon 1981). What is not clear from past research, however, are the conditions under which potential buyers will be receptive or unreceptive to WOM information.

Even though there is a substantial amount of research documenting the pervasive influence of WOM information, some researchers have begun to question the narrow band of conditions under which those findings have been obtained (e.g., Scott and Tybout 1981; Weinberger, Allen and Dillon 1981). Typically, past research has attempted to model or examine the effect of WOM information on potential buyers for products that have no market history, are unbranded, or when the product category is novel. All of the above situations would minimize the role of the potential buyer's prior expectations or predisposition in determining his/her receptivity to the incoming WOM information. This narrow focus stands in contrast to the reality of a marketplace where, in the majority, products are line extensions or new brands within an existing product line and/or the company is known to the potential buyer (Heany 1983). In the latter case, many -- perhaps a majority -- of the potential buyers will have non-neutral feelings regarding the product and these predispositions may mediate their receptivity to information transmitted by previous buyers.

In recent research, Wilson and Peterson (1989) examined the impact of positive or negative WOM information for new products where the potential buyers had well-established positive or negative affective predispositions toward the products. They found that the WOM information was only accepted to the extent the valence of that information matched the affective predisposition of the receiver.

A substantial literature documents the mediating influence of the receiver's predisposition or prior expectations on receptivity to and interpretation of new information (e.g., Allport 1961; Bettman 1979; Crocker 1981; Einhorn and Hogarth 1986; John, Scott, and Bettman 1986; Kassin 1979; McGuire 1968; Metalsky and Abramson 1981; Neisser 1966; Nisbett and Ross 1980; Petty and Cacioppo 1979; Sherif, Sherif and Nebergall 1965). In general, incoming information that does not "fit" existing beliefs, expectancies, and/or affective predispositions tends to be distorted, transformed, discounted, or ignored by receivers. It is not always the case, however, that an individual's prior feelings or beliefs will dominate the treatment of new information (e.g., Bettman. John and Scott 1986). Thus, the challenge is to determine the conditions that enhance or attenuate the influence of new, incoming information, especially when it apparently conflicts with existing information, beliefs, or feelings.

In the area of covariation assessment, Alloy and Tabachnik (1984) have argued persuasively for the need to look at the interaction between prior expectations and currently available information to accurately predict outcomes. Specifically, they suggest that the stronger an individual's prior feelings, the more the feelings will dominate the interpretation and use of WOM information. Conversely, even strongly held prior expectations or beliefs can be overwhelmed by contradictory new information if it is sufficiently strong, salient, and/or if a substantial amount has been accumulated.

in the current research, we attempt to predict the impact of WOM information on product evaluations and purchase intentions by examining the joint influence of an individual's prior affective predisposition and WOM information. Similar to Wilson and Peterson (1989), we assumed the potency of positive or negative WOM information would generally be determined by its fit with the receiver's prior feelings. Thus, for example, individuals who have a positive predisposition toward a product will be unreceptive to negative information but very receptive to positive information; the mirror image should be true for individuals with a negative predisposition. Individuals with relatively neutral feelings should be receptive to both types of information. Thus, for example, contrary to past findings on the potency of negative information, we predict negative information will produce weaker effects than positive information if directed toward receivers who are favorably predisposed toward a product.

Unlike Wilson and Peterson (1989), however, we wanted to examine this relationship in a context where the affective predisposition was newly established and not well-founded. In addition, we wanted to parallel past WOM research where the product had no market history, was unbranded, or the product category was novel but where the potential buyer was asked to render an evaluative judgment about the product prior to the presentation of WOM information.

Specifically, subjects were given two brief product descriptions and asked if Brand A appeared to be a better value than Brand B or no different in value. We predicted that subjects who favored one brand over another would be very receptive to positive WOM information about that brand and unreceptive to negative WOM information (the opposite effect would be obtained for the non-favored brand). No difference was expected in receptivity to WOM information about the two brands if the subject had not indicated an evaluative position favoring one brand over the other. The primary purpose of this research was to determine if the impact on product evaluations and purchase intent in this context could be better understood by examining the joint influence of an individual's evaluative predisposition and WOM information compared to past research that has only examined the strength and valence of the WOM information.

METHOD

Sample. Data for this study were collected during August 1987 from a national consumer panel through a mail survey. The sample was selected from members of the panel to represent a cross-section of the adult population in the United States. The original sample size was 1,600. Responses were received from 998 persons, resulting in a response rate of 62.3 percent. Such a response rate is typical for a study using a national consumer panel and where data are collected by a mail survey not employing any follow-up contact.

To better reflect national population characteristics, sample data were weighted by sex and education of respondent. Slightly more than 52 percent of the weighted sample consisted of females, and 72 percent of the sample were married. Forty-nine percent of the weighted sample were age 50 and over; 64 percent had no college education; and 37 percent reported a (1985) household income of at least $30,000. The average household contained 2.7 people.

Survey Questionnaire. All subjects were first provided a brief, basic description of a new product category: digital tape recorders (DTR). They were told that DTRs were just being introduced into the marketplace and were expected to do well because of their superior sound reproduction compared to traditional tape recorders. Moreover, they were informed that a consumer testing service had rated 8 DTR brands and had recommended two brands (A and 3) as best buys.

Brand A was described as quite expensive 5500) but it had received the highest ratings for quality of parts and construction. It also had the longest warranty (2 years). Although it was expensive, it received a "best buy" rating because of its high quality. Brand B had the lowest price (5250). Its sound quality was rated as good as Brand A but parts and construction were judged to be of only moderate quality. Its warranty was only for six months. Although the judged quality was only moderate, it received a "best buy" rating because of its ow price.

Based on the above information subjects were asked to indicate which recorder appeared to be a better value. Subjects could either indicate Brand A, Brand B, or about the same. The purpose of this question was to get subjects to make an evaluative judgment about the recorders. We assumed this judgment would influence how subjects would interpret subsequent WOM information. That is, the rendering of an evaluative judgment would elicit an affective predisposition toward brands in the category which would determine the subject's receptivity to and interpretation of further information about the brands.

In the next part of the survey, subjects were presented with a WOM negative performance information manipulation. They were presented the following scenario:

Suppose you found out your neighbor had purchased Brand LA or B] and that he had to take it back to the dealer after three months because of a hissing sound. He has had it back home for two weeks without further problems.

Half the subjects were told their neighbor had purchased Brand A and half were told Brand B. Subjects were then asked to indicate if the buyer's experience had raised, lowered, or had no effect on their perception of the value of Brand [ A or B].

Finally, subjects were provided with the affective (positive or negative) evaluation manipulation of Brand A or B. Half of the subjects were told that the buyer "loved" his DTR and the other half were told the buyer "hated" his DTR. Subjects were then asked to indicate if that information made them more likely, less likely, or had no effect on their consideration of purchasing that particular brand.

RESULTS

To evaluate the impact of affective predisposition on receptivity to W-O-M information. three groups of respondents were formed on the basis of their value comparison judgments of Brand A versus Brand B. Of the 966 subjects who gave a response, 71.3% (689 subjects) indicated that Brand A was a better value than Brand B (labeled the A>B group). Only 7.9% (76 subjects) indicated Brand B was a better value than Brand A (B>A group). The remaining 20.8% (201 subjects) indicated that the brands were of about the same value (A=B group).

TABLE 1

OVERALL RESPONSES FOR THE THREE VALUE COMPARISON GROUPS

Negative Performance Information. The fact that the buyer reported a performance problem with the brand purchased had a negative effect on perceived value of that brand for a fair number of respondents. As can be seen in Table 1, about 41% of all respondents lowered their value ratings of that brand. As expected, there were no differences among the three value comparison groups when brand type is not taken into account. But as can be seen in Figure A, however, the impact of such information for each group is quite different if brand type is examined. As predicted, subjects in the A>B group are much more likely to lower the value rating of Brand B than Brand A ([ =9.37; p < .001) while the opposite effect is observed for subjects in the B>A group (t= 2.92; p<.005). Although subjects in the A=B group were not expected to respond differentially as a function of brand, more subjects lowered their value rating for Brand A than for Brand B (t=2.17; p< .05) in response to the negative performance information. This difference may be due to the fact that respondents had higher expectations about the performance reliability of Brand A than Brand B because much of the value of the former was derived from quality of its parts and construction. Therefore, a performance problem with NSX Brand A elicited more uncertainty about its true quality than did a performance problem with Brand B.

For comparison purposes, differences between the A>B and B>A group responses for the same brand were tested. Significant differences were obtained between the proportion of respondents in each group who lowered their value ratings for both Brand A (t=3.92; p<.001) and Brand B (t=4.38; p<.001). That is, relatively more A>B subjects than B>A subjects lowered their value rating of Brand B, while relatively more B>A subjects than A>B subjects lowered their value ratings for Brand A when faced with negative performance information.

Positive versus Negative WOM Evaluation Information. The overall impact of positive and negative WOM evaluations on purchase intention is presented in Table l. In general, positive WOM information enhances the purchase intent of subjects while negative WOM information depresses purchase intent. As expected, negative WOM information affected the purchase intent of subjects more than did positive information.

When subjects' reactions to positive WOM information by brand are examined, a much different pattern of responses is observed. This is apparent from a comparison of the responses of the three value groups (see Figure B). As predicted, relatively more subjects in the A>B group were likely to increase their purchase intent if the owner provided a positive evaluation of Brand A than Brand B (t=7.77; p<.001) In contrast, subjects in the B>A group demonstrated a greater increase in purchase intent if the brand given a positive evaluation was B rather than Brand A, although the effect does not reach statistical significance (t= 1.55). Brand differences were not expected for the A=B group and there appears to be no effect.

FIGURE A

PERCENTAGES OF SUBJECTS LOWERING BRAND EVALUATION RATINGS AFTER A REPORT OF NEGATIVE PERFORMANCE INFORMATION BY BRAND TYPE AND VALUE COMPARISON GROUP

The increase in purchase intent for Brand A following the presentation of positive WOM was higher for subjects in the A>B group than subjects in the B>A group (t=2.63; p<.01). In contrast, B>A group members indicated a greater increase in purchase intent than A>B group members when Brand B was the brand of interest. These effects are consistent with our predictions

In general, negative WOM had a very powerful impact on lowering purchase intent of the brand toward which the comments were directed. But, as predicted, negative WOM had more impact on purchase intentions for Brand B than Brand A within the A>B group (t=2.55; p<.05), whereas no differential impact for brand was observed for members of the A=B group. Contrary to our prediction, however, members of the B>A group were equally affected by negative WOM for both brands A and B.

Another way to examine the differential effects of positive versus negative WOM information is to compare the net difference for each group between how many subjects were more likely to purchase the brand given positive evaluations minus how many subjects were less likely to purchase the brand given negative evaluations. These results are provided in Table 2. As can be seen in the table, positive WOM appears to be equally potent as negative WOM in those situations where the subjects encountering the information took an initially positive evaluative position toward the brand (i.e., respectively, Brand A: A>B group and Brand B: B>A group). In the absence of a positive commitment to the product (Brand A: B>A and A=B group; Brand B: A>B and A=B group), however, the unfavorable impact of negative WOM information on purchase intent was quite dramatic.

DISCUSSION AND CONCLUSION

In general, the research findings suggest the influence of word-of-mouth information on product evaluations and purchase intentions of potential buyers is much more complex than has heretofore been reported. Past research has documented the powerful influence of WOM information, especially if that information is unfavorable. There appear to be forces, however, that influence both the reception and retention of WOM information beyond the simple content, strength, or valence of such information. A better understanding of such forces should lead to more accurate predictions of the net influence of WOM information on potential buyers' purchase behavior.

FIGURE B

PERCENTAGES OF SUBJECTS REPORTING HIGHER PURCHASE INTENT AFTER A POSITIVE OWNER EVALUATION BY BRAND TYPE AND VALUE COMPARISON GROUP

Consistent with predictions, receptivity to positive or negative WOM information was influenced by evaluative feelings toward the product, even though the product had no market history, was unbranded, and the product category was relatively novel. The results indicate clearly that once an individual assumes an evaluative position toward a product, he/she will begin to filter information about that product. To the extent the information "fits" the evaluative position, it will be accepted. These results suggest that predicting WOM influence on acceptance of new products or services may be greatly improved by understanding the evaluative predispositions of potential buyers as well as the valence of the WOM information.

Wilson and Peterson (1989) found that receptivity to positive or negative WOM information was almost completely dependent on whether the valence of that information matched the receiver's affective predisposition toward the product. The present results parallel to a great extent those findings. One major difference, however, is worth noting. In their research, subjects expressed very strong feelings (positive or negative) about the products being considered (the products were new car models by well known nameplates). In fact, subjects with a positive predisposition increased their purchase intent when faced with negative performance and affective WOM information regarding the target product. In the present research, subjects were impacted by negative WOM information even when they had taken a positive evaluative position on the brand. This result seems reasonable given that the evaluative position assumed by most subjects was based on minimal information and their position had been established only briefly before counter information was encountered.

Overall the results suggest that the amount of time may be quite brief, following the introduction of new products, that information about those new products will be processed by potential buyers in a relatively unbiased fashion. It appears that the influence of WOM information, although quite powerful, may be severely curtailed to the extent it runs counter to the individual's feelings toward the product. Obviously, the present findings call into question the validity of virtually all contemporary innovation/diffusion models in marketing (cf., Mahajan and Wind 1986) that assume a given level of receptivity of all potential buyers towards the positive and negative WOM information they encounter. Our findings suggest that innovation/diffusion rates for new products might be better understood and predicted by taking into account the receptivity of the receiver to WOM information in addition to measuring the valence, strength, and/or amour t of WOM information transmitted.

Almost two decades ago, Robertson (1971) argued that, in spite of its recognized importance, personal influence was one of the most "elusive" concepts in the consumer behavior literature, primarily due to a lack of research activity in the area. While considerable research findings have accumulated in the interim, there are still significant gaps in our knowledge. The present results suggest the influence of word-of-mouth informal on cannot be fully understood without considering how it interacts with the receiver's evaluative predisposition toward the product.

FIGURE C

PERCENTAGES OF SUBJECTS REPORTING LOWER PURCHASE INTENT AFTER A NEGATIVE OWNER EVALUATION BY BRAND TYPE AND VALUE COMPARISON GROUP

TABLE 2

NET DIFFERENCE BETWEEN POSITIVE VERSUS NEGATIVE W-O-M EVALUATION EFFECTS ON PURCHASE INTENT FOR VALUE COMPARISON GROUPS X BRAND TYPE

REFERENCES

Alloy, Lauren B. and Naomi Tabachnik (1984), "Assessment of Covariation by Humans and Animals: The Joint Influence of Prior Expectations and Current Situational Information," Psychological Review, 91 (January), 112-149.

Allport, Floyd H . (1955), Theories of Perception and the Concept of Structure, New York: John Wiley & Sons, Inc.

Arndt, J. (1967), Ford of Mouth Advertising A Review of the Literature, New York: Advertising Research Foundation. Wilson, William R. and Robert A. Peterson (1988), 'The Influence of Category-Based Processing in Determining the Diagnosticity of Product-Related Information, (unpublished manuscript; under editorial review).

Bettman, James R. (1979), An Information Processing Theory of Consumer Choice, Reading, MA: Addison-Wesley.

Bettman, James R., Deborah Roedder John, and Carol A. Scott (1986), "Covariation Assessment by Consumers,' Journal of Consumer Research, 13 (December), 316-326.

Brown, Jacqueline Johnson and Peter H. Reingen (1987), "Social Ties and Word-of-Mouth Referral Behavior," Journal of Consumer Research, 14 (December), 350-362.

Crocker, J. (1981), "Judgment of Covariation by Social Perceivers," Psychological Bulletin, 90 (September), 272-292.

Einhorn, Hillel J. and Robin M. Hogarth (1986), "Judging Probable Cause," Psychological Bulletin, 99 (January), 3-19.

Heany, Donald F. (1983), "Degrees of Product Innovation," The Journal of Business Strategy, (Spring), 3- 14.

John, Deborah Roedder, Carol A. Scott, and James R. Bettman (1986), "Sampling Data for Covariation Assessment: The Effect of Prior Beliefs on Search Patterns," Journal of Consumer Research, 13 (June), 3847.

Kassin, Saul M. (1979), 'Consensus Information, Prediction, and Causal Attribution: A Review of the Literature and Issues," Journal of Personality and Social Psychology, 37 (November), 1966-1981.

Katona, George and Eva Mueller (1955), "A Study of Purchasing Decisions," in Consumer Behavior: The Dynamics of Consumer Reaction, ed., L. H. Clark, New York: New York University Press.

Mahajan, Vijay and Yoram Wind (1986), Innovation Diffusion Models of New Product Acceptance, Cambridge, MA: Ballinger Publishing Co.

Mahajan, Vijay, Etian Muller, and Roger Kerin (1-984), "Introduction Strategy for New Products with Positive or Negative Word of-Mouth,' Management Science, 30 (December), 1389-1404

McGuire, W. J (1968), "Personality and Susceptibility to Social Influence," in Handbook of Personality Theory and Research, eds., E. F. Borgatta and W. W. Lambert, Chicago: Rand-McNally.

Metalsky, G.I. and L. Y. Abramson (1981), "Attributional Styles; Towards a Framework for Conceptualization and Assessment," in Assessment Strategies for Cognitive-Behavioral Interventions, eds., P. C. Kendall and S. D. Hollon, New York: Academic Press, 13-58

Mizerski, Richard W. (1982), "An Attribution Explanation of the Disproportionate Influence of Unfavorable Information," Journal of Consumer Research, 9 (December), 301-310.

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

Robertson, Thomas i. (1971), Innovative Behavior and Communication, New York: Holt Rinehart and Winston, Inc.

Scott, Carol A. and Alice M. Tybout (1981), "Theoretical Perspectives on the Impact of Negative Information: Does Valence Matter?," in Advances in Consumer Research, Vol. 8, ed., Kent Monroe, Ann Arbor, MI: Association for Consumer Research, 408-409.

Sherif, C. W., M. Sherif and R. E. Nebergall (1965), Attitude and Attitude Change, Philadelphia: W.'B. Saunders.

Weinberger, Marc G. and William R. Dillon (1980), 'The Effects of Unfavorable Product Rating Information," in Advances in Consumer Research, Vol 7, ed. Jerry C. Olson, Ann Arbor, MI: Association for Consumer Research, 528-532.

Weinberger, Marc G., Chris T. Allen, and William R. Dillon (1981), "Negative Information: Perspectives and Research Directions," in Advances in Consumer Research, Vol 8, ed., Kent B. Monroe, Ann Arbor, MI- Association for Consumer Research, 398-404.

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