Because of its importance to inter-firm relationship quality and thus to inter-firm relationship continuity, this research investigates previously unstudied antecedents of satisfaction in a marketing channel. It tests, for example, whether satisfaction reduces the attractiveness of alternative supplier relationships, as it is widely believed, or whether it works the other way around. The results were surprising. They suggested that satisfaction reduced alternative attractiveness, as it is widely believed, and that alternative attractiveness reduced satisfaction in the study context. The standardized effects were of equal magnitude (β = -.29) and they were comparatively large for this type of research.

Another proposed antecedent of satisfaction, voice (constructive attempts to change objectionable relationship conditions), increased satisfaction. However, satisfaction had no effect on voice. Thus, satisfaction may not increase voice as previously thought, and voice was an antecedent of satisfaction. The standardized effect of voice on satisfaction was the largest in the study (β = .45).

With several exceptions most of the other study variables directly or indirectly affected satisfaction, and together the study variables explained nearly half of the variance in satisfaction. Specifically, retailer investment in the relationship, and the number of their employees also positively affected satisfaction. The number of their rivals negatively affected their satisfaction. However, of these relationships voice, alternative attractiveness, and investment were several times larger than the others, and the voice relationship with satisfaction was the largest in the study.

Overall, the study suggests that important antecedents of relationship satisfaction include voice, alternative attractiveness, and investment in the study context. Although it is very risky to generalize from a single study, channel relationship managers interested in fostering satisfaction in their customers may wish to sincerely solicit retailer complaints, and then work to resolve them in a mutually satisfacto­ry manner. In addition to 800 numbers and quick and competent complaint processing, they may want to provide e-mail addresses with rapid responses. They may also want to consider providing on-line customer communities, such as user community bulletin boards, where problems and solutions can be shared. Relationship managers may also wish to publicize to their retailing partners successful outcomes proceeding from other retailers' use of voice. They may want to have wholesaler sales persons actively solicit retailer voice by asking for their "number one complaint." They may also want to consider using wholesaler-sponsored retailer satisfaction surveys to facilitate voice in less vocal retailing customers, such as those firms with fewer employees, or those with higher revenue (who were less likely to be vocal in the study).

Channel relationship managers might also consider taking steps to actively reduce the attractiveness of their exchange partners' alternative relationships, and increase their relationship partners' investment in the exchange relationship. These steps might include activities such as tailoring some of their promotional activities specifically for their established customers in order to reduce their customers' alternative attractiveness. Although there is little guidance in the marketing channels literature for comparative advertising, this type of advertising may be an effective approach to reducing alternative attractiveness. Featuring the supplier's superior logistics service levels for example may be an effective tactic. Providing retailer "success stories" in a newsletter directed to retailer customers, or on the company web site, that tells of other retailers who switched from competitive wholesalers may also be useful.

Channel managers may also wish to consider increased emphasis on designing and promoting additional relationship investments to their retailing partners. Prospect theory suggests that individuals and presumably firms should view the additional costs of relationship investment noncumulatively. This suggests that with all other things being equal, retailers with both high or low levels of relationship investment should be equally likely to make additional investments of time, effert or money in the relationship. In addition, prospect theory suggests that investment opportunities should be offered as either a single large gain or a series of small costs, depending on how the prospective investor might view the proposed investment. Opportunities for selling additional relationship investment could be identified using a series of customer satisfaction surveys and systematic prospecting and follow-up selling activities aimed for example at long-time customers (who were more likely to have invested in the relationship).