Determinants of Word-Of-Mouth Communications During Product Consumption

Paula Fitzgerald Bone, West Virginia University
ABSTRACT - Word-of-mouth (WOM) communications which occur during product consumption have received little attention from consumer researchers. In this paper, factors which may "spark" WOM during consumption are examined. Findings from a field study show that WOM can be partially explained by: (1) social tie strength, (2) the presence/absence of an individual taking a committed decision maker role, (3) consumer satisfaction and, (4) perceived novelty.
[ to cite ]:
Paula Fitzgerald Bone (1992) ,"Determinants of Word-Of-Mouth Communications During Product Consumption", in NA - Advances in Consumer Research Volume 19, eds. John F. Sherry, Jr. and Brian Sternthal, Provo, UT : Association for Consumer Research, Pages: 579-583.

Advances in Consumer Research Volume 19, 1992      Pages 579-583


Paula Fitzgerald Bone, West Virginia University


Word-of-mouth (WOM) communications which occur during product consumption have received little attention from consumer researchers. In this paper, factors which may "spark" WOM during consumption are examined. Findings from a field study show that WOM can be partially explained by: (1) social tie strength, (2) the presence/absence of an individual taking a committed decision maker role, (3) consumer satisfaction and, (4) perceived novelty.

Foods, sporting goods, musical concerts, and videotaped movies all have one thing in common--they are often consumed in groups. And, when products are consumed in groups, there is the possibility that word-of-mouth communications (WOM) may occur. Indeed, Belk (1971) found that WOM is most likely to occur when individuals are in close proximity to a product. In this study, several factors which may influence the amount of WOM during consumption are examined.

The traditional notions of information seeker and information receiver seem to be inappropriate when studying WOM which occurs during product consumption since individuals in this situation are simply sharing thoughts--they are not necessarily providing and requesting information. For this reason, WOM communication is conceptualized herein as a group phenomenon--an exchange of comments, thoughts, and ideas among two or more individuals in which none of the individuals represent a marketing source.

Several factors may "spark" WOM during the consumption experience. This paper examines aspects of the social environment, objective characteristics of the situation, and individuals' perceptions of the consumption experience as determinants of how much WOM about the product/service occurs (cf. Yale 1987a). In the next section of this paper, these potential determinants of WOM are discussed. Then, an empirical investigation of these factors is presented.


Two aspects of the social environment of a consumption group are of interest here: social ties and the social role of committed decision maker. Social ties represent the strength of a consumer's relationship to the people accompanying him/her. Since relationship strength has been shown to influence information flow from consumer to consumer (Brown and Reingen 1987), it is also likely to affect discussions during consumption. Rumor researchers suggest that persons in a new social relationship spread rumors in order to learn more about each other (Fine and Rosnow 1978). Taking this concept further, individuals who are simply acquaintances or casual friends (i.e., have weak social ties) may attempt to strengthen friendships. They may look for safe, uncontroversial topics of discussion that can be used to get to know each other better (Yale 1987a). Discussion about products/services consumed may meet consumers' needs to learn more about their companions. Thus, it is suggested that the weaker the social ties that exist among group members, the more WOM will occur.

A second component of the social environment is whether one or more of the group members takes on the role of a committed decision maker. Conceptually, a committed decision maker is an individual who selects the product/service to be consumed and is enthusiastic about sharing this consumption experience with his/her companions. For instance, assume John lives in New York City and has friends coming to visit for the weekend. John wants to show off New York, so he carefully chooses a restaurant that he particularly likes. He has taken the role of committed decision maker.

The committed decision maker makes a choice that is visible and affects other people. This situation is likely to create some psychological dissonance for the decision maker, so, s/he will seek confirmation of the choice from her/his companions (Stuteville 1968). In fact, Dichter (1966) proposes that group members are motivated to report their experiences and opinions to the decision maker and that the decision maker expects such reports. This suggests that whenever there is a committed decision maker in a group, WOM is likely to occur.


Belk (1975a, 1975b) suggested five categories of objective situational variables, two of which are examined in this paper: temporal perspective and task definition. Temporal perspective deals specifically with the concept of time. Belk (1975a) proposes that the time elapsed since last purchase may be an important part of the consumption situation. This situational factor suggests that the greater the time between consumption experiences, the more attention the consumer pays to it. Attention will be maximized whenever it is a consumer's first experience with a particular product or service. Thus, greater WOM is expected to occur whenever the consumer has his/her first experience with a product or service.

Task definition explores the reason behind a particular consumption experience (Belk 1975a). The gift-giving situation is of interest here, since it may spark conversation during consumption. The gift-giver(s) is motivated to please the recipient(s); therefore, situational involvement is heightened. Additionally, the gift-giver is likely to want feedback from the recipient concerning his/her opinion regarding the product. Finally, the receiver may have questions about the product/service which would be addressed to the giver.


The final set of variables considered is individual's perceptions of the consumption experience. Three such factors, satisfaction, dissatisfaction and perceived novelty, are discussed below.

Satisfaction and dissatisfaction affect an individual's mood and increase the amount of WOM. In fact, much research shows that satisfaction and dissatisfaction lead to WOM after product experience (Dichter 1966; Richins 1983; Yale 1987b). Satisfaction and dissatisfaction may begin as a response to a situation (Belk 1975a) but then become part of the situation. This is possible because product consumption occurs over time, so satisfaction and dissatisfaction can occur early in the consumption experience and then influence WOM during the latter part of the consumption experience. Extreme levels of satisfaction and dissatisfaction heighten the consumer's awareness of the consumption experience and increase his/her likelihood of discussing the experience. Indeed, such discussions may reduce emotional tension created by a very pleasing or displeasing product/service (Dichter 1966).

A novel consumption experience is one that is perceived as out of the ordinary by an individual. This may be a function of the consumer's lifestyle and experiences, characteristics of the product/service, and/or the manner in which the product/service is presented. A situation which is perceived as novel will receive the consumer's attention, making product conversation more likely.


A restaurant setting was used in this investigation since individuals often consume restaurant food in groups and the consumption experience has a clear beginning and ending. A wide variety of restaurants participated in the study ranging from fast-food chains to expensive "atmosphere" restaurants. A total of 321 people participated, representing 153 groups of restaurant patrons. Forty-three percent of the respondents were male, 72% were under the age of 45, and 46% had incomes over $30,000.

Interviewers were instructed to approach every group of two or more individuals after they had completed their meals and ask all group members to participate in the study. Respondents were asked to complete the questionnaire individually without consulting the other group members.


The following hypotheses are of interest. The amount of WOM in a group is expected to increase:

H1: As the number of weak social ties among group members increases.

H2: When a group member takes on the role of committed decision-maker.

H3: As the number of individuals in the group who are experiencing the restaurant for the first time increases.

H4: As the number of individuals in the group who are experiencing a particular entree in the restaurant for the first time increases.

H5: When the consumption experience represents a gift-giving situation.

H6: As the number of people in the group who experience extreme satisfaction increases.

H7: As the number of people in the group who experience extreme dissatisfaction increases.

H8: As the number of individuals in the group who perceive the consumption experience as novel increases.


Word-of-mouth. An index of the amount of WOM was developed for this study since the usual dichotomous measures could not adequately test the hypotheses. Three items measured the amount of WOM. The first was an eleven-point graphic rating scale anchored with "we did not talk about the food at all" and "we talked about the food a lot." A seven-point Likert-type item asked whether the food eaten was a large part of the mealtime conversation. Finally, a seven-point item asked how much of the table conversation dealt with the food being eaten. Responses ranged from "nothing was said about the food" to "the biggest topic of conversation was our food." These measures were standardized and summed to yield a reliable index (alpha=0.79).

Social Ties. Participants indicated their social relationship with each person sitting at their table using an seven-point item. Responses ranged from "not very close--casual acquaintances" to "so close it's hard to imagine life without him/her" (Johnson and Reingen 1987).

Committed Decision Maker. This social role was measured by three items. First, participants were asked who chose the restaurant. Only respondents who indicated they had made the decision were considered for the committed decision maker role. The psychological dimension of this role was measured using two seven-point Likert-type items. The first stated, "I was very excited to bring my companions to this particular restaurant." The second stated, "I encouraged my companions to come here today because this is one of my favorite places to eat and I wanted to share this restaurant with them" (r=0.69).

Gift. Respondents were asked whether they were purchasing the meal as a gift for their companions.

Satisfaction/Dissatisfaction. These states were measured using two scales. The first was a seven-point delighted to terrible scale (Westbrook 1980b). The second asked respondents how satisfied they were with the food. Responses could range from 0% (not at all satisfied) to 100% (completely satisfied; Westbrook 1980a). These two items were standardized and summed (r=0.70).



Restaurant Patronage and Entree Purchasing Frequency. Respondents indicated how often they ate at the restaurant and how often they ordered the entree selected on six-point items. A response format ranging from "at least once or more a week" to "never before today" was used.

Perceived Novelty. Perceived novelty was measured by three seven-point Likert-type items asking whether the food was presented in a different manner than expected, whether the food was presented in an interesting or unusual manner, and whether the food was out of the ordinary for the respondent (alpha=0.64).

Data Analysis

Several issues had to be addressed during data analysis. Some groups had incomplete data. Groups were only included if the majority of the members completed the questionnaire. Respondents who did not complete more than 3 items were excluded from the analysis as were individuals who did not complete all three of the WOM items. This eliminated persons who did not thoughtfully complete the questionnaire. Missing data were handled by substituting the mean value for the missing data point. This conservative procedure may have attenuated the results since it treats an individual as average when, in fact, s/he may have possessed one of the critical characteristics. Thirty-nine groups were dropped from the analysis due to the above procedures, resulting in usable data from 114 groups.

In order to test the hypotheses, information from the individual questionnaires was combined to provide group data. Table 1 provides the rules used in determining the value of the independent group variables used in the final analysis.


A regression model was used to investigate the hypotheses. The condition index for the data set was 9.27, indicating that multicollinearity is unlikely to be a problem (Belsley, Kuh and Welsh 1980). In order to control for group size, the analysis was conducted using the residual WOM scores after the effects of group size were taken into account. Hypotheses were tested using the residual group WOM score as the dependent variable and the weak ties, committed decision maker, new restaurant patron, new entree, gift, high satisfaction, low satisfaction, and distinctiveness group scores as the independent variables. The overall model was statistically significant (F=4.54, p<0.01, adjusted R2=0.20). Table 2 contains the regression coefficients. (Note that several potential interactions were examined; however, none were statistically significant.)



The results indicate that four of the variables are significant. It appears that having (1) a committed decision maker in the group, (2) highly satisfied group members, and (3) group members who perceive the consumption experience as novel creates greater product discussion. Thus, H2, H6, and H8 are supported. The relationship between weak social ties and WOM is significant but in the opposite direction than hypothesized. Four of the independent variables are not statistically significant: (1) a new restaurant patron, (2) new entree consumed, (3) gift-giving situation, and (4) low satisfaction. Thus, H3, H4, H5, and H7 are not supported.


This field study suggests that the social environment and individuals' perceptions of the consumption experience play a role in determining the amount of product discussion during consumption, explaining 20% of the variance in WOM behavior. Additionally, it appears that WOM during consumption is a common phenomenon since more than 80% of the respondents indicated that at least one or two comments were made about the food during the meal. Given that WOM can have a strong effect on performance perceptions (Bone 1990), creating positive WOM during consumption should be an important part of a marketing strategy.

The variables found to spark WOM can be used to create positive WOM during consumption. Marketers may wish to create group consumption situations which include a committed decision maker. Many firms have "regulars" who are likely to be psychologically committed to the establishment. Special promotions can be used to entice these customers to bring their friends and associates along as well. The finding that high satisfaction leads to greater WOM emphasizes the importance of providing quality products and services. Quality which exceeds a consumer's expectations will be rewarded by positive WOM. Finally, offering novel products via unusual presentations or unique product/service offerings can also increase the amount of WOM during consumption.

A few of the findings from the field study were surprising. Most surprising was the inverse relationship between social ties and WOM. It is possible that individuals who have weak social ties spend more time talking about each other's past experiences and interests in an effort to find common interests and to learn more about each other. On the other hand, individuals with strong ties may spend more time talking about the present since they already know so much about one another. Their conversations may focus on daily events (e.g., How was the office today? What do you want to watch on TV tonight? Do you like what you ordered?).

Also surprising is the fact that high satisfaction led to more WOM during consumption, but low satisfaction did not. One problem may be that the vast majority of the respondents in the study were relatively satisfied with the food they received. Only one subject indicated that the food was terrible and only three indicated that they were mostly dissatisfied (out of 321 respondents). Thus, even though individuals identified as dissatisfied had satisfaction scores well below the mean, it is possible that they experienced only moderate dissatisfaction.

This study reveals that WOM during product consumption is a rather common event which can be partially explained by only a handful of variables. Of course, findings may be limited to the respondent population and environments examined herein, but the study does open a variety of issues for further investigation. For instance, only a small portion of the factors which may affect WOM were investigated. Other characteristics such as the consumption atmosphere (e.g., upscale vs. downhome environments), the structure of the consumption group (e.g., family vs. business), mood states, time of day/year, and product consumption with the purpose of determining adoption of the product/service by a group may also affect the amount of WOM during consumption.

In the future, other product consumption situations should be examined. Specifically, alternative product such as movies, compact disks, sporting equipment, video games, etc. can be examined to determine how frequently conversations occur and what variables spark such conversations. In addition, we can examine product trial situations and shopping experiences to determine the frequency and content of such discussions.

Given the high levels of satisfaction reported by the respondents, it is probable that the majority of the conversations held were positive in nature. Yet, future research should examine the content of product conversations--the valence of the conversations, common themes which run through conversations, the demographic and psychological characteristics which play a role in the content of WOM, and the effect of these conversations on repurchase decisions.

In sum, this study shows that several variables (social ties, committed decision maker role, satisfaction and perceived distinctiveness) systematically influence the amount of WOM during consumption. In addition, the paper offers a multi-item measure of the amount of WOM which appears reliable. Given that WOM conversations can explain over 20% of the variance in product judgements (Bone 1990), it is clear that this topic is deserving of further research.


Belk, Russell W., (1971) "Occurrence of Word-of-Mouth Buyer Behavior as a Function of Situation and Advertising Stimuli," in Combined Proceedings of the American Marketing Association, Fred C. Allvine (ed.), Chicago, IL: American Marketing Association, 419-22.

Belk, Russell W. (1975a), "Situational Variables and Consumer Behavior," Journal of Consumer Research, 2 (December) 157-164.

Belk, Russell W. (1975b), "The Objective Situation as a Determinant of Consumer Behavior," in Advances in Consumer Research, Mary Jane Schlinger (ed.), Chicago, IL: Association for Consumer Research.

Belsley, David A., Edwin Kuh and Rowy E. Welch (1980), Regression Diagnostics: Identifying Influential Data and Sources of Collinearity, John Wiley & Sons, Inc. New York, Chap. 3. pp. 100-105.

Bone, Paula Fitzgerald (1990), "The effect of Word-of-Mouth Communications During Consumption on Performance Perceptions," in 1990 AMA Educator's Proceedings: Enhancing Knowledge Development in Marketing, William Bearden et al. (ed.), Chicago, IL: American Marketing Association, p. 26.

Brown, Jacqueline, and Peter H. Reingen (1987), "Social Ties and Word-of-Mouth Referral Behavior," Journal of Consumer Research, 14, 3, 350-362.

Dichter, Ernest (1966), "How Word-of-Mouth Works," Harvard Business Review, 44 (November/ December) 147-166.

Fine, Gary and Ralph Rosnow (1978), "Gossip, Gossipers, and Gossiping," Personality and Social Psychology Bulletin, 4 (Winter), 161-168.

Holmes, John H. and John D. Lett, Jr. (1977), "Product Sampling and Word-of-Mouth," Journal of Advertising Research, 17 (5), 35-40.

Lutz, Richard J. and Pradeep Kakkar (1975), "The Psychological Situation As a Determinant of Consumer Behavior," in Advances in Consumer Research, Mary Jane Schlinger (ed.) Chicago, IL: Association for Consumer Research.

Richins, Marsha L. (1983), "Negative Word-of-Mouth by Dissatisfied Consumers: A Pilot Study," Journal of Marketing 47 (Winter), 68-78.

Stuteville, John R. (1968), "The Buyer as a Salesman," Journal of Marketing, 32 (July), 14-18.

Westbrook, Robert A. (1980a), "Intrapersonal Affective Influences on Consumer Satisfaction with Products," Journal of Consumer Research, 7 (June), 49-54.

Westbrook, Robert A. (1980b), "A Rating Scale For Measuring Product/Service Satisfaction," Journal of Marketing, 44 (Fall), 68-72.

Yale, Laura J. (1987a), "Word of Mouth Advertising: A Synthesis and Review," Working Paper, Graduate School of Management University of California-Irvine.

Yale, Laura J. (1987b), "An Empirical Study of Word of Mouth Behavior: Antecedents, Processes and Outcomes," Working Paper, Graduate School of Management University of California-Irvine (1987b).