SIDE PAYMENTS IN MARKETING

Side payments, known politely as gainsharing and pejoratively as bribe ry, are prevalent in marketing. Indeed, many management schools have a dded ethics modules to their basic marketing courses to discuss these issues and there is much discussion of side payments in the literature (e.g., Adams 1995, Borrus 1995, Mauro 1997, Mohl 1996, Murphy 1995, P eterson 1996, and Rose-Ackerman 1996). We seek to provide insight with respect to one class of marketing side payments. We hope that our ana lyses clarify some of the issues and suggest how these side payments a ffect marketing activities. We begin by focusing on one common example of potential side payments-salesforce ratings of internal sales suppo rt. We derive two formal results and speculate on how these results ge neralize. The two results are (1) that having one group of employees r ate another implies that there are almost always incentives for side p ayments, but (2) the side payments need not reduce the firm's profit. At least in theory, the firm is always able to revise the reward syste m to factor out these side payments. The first result, based on a stra ightforward proof, has important practical implications for managers w ho may wish to preclude side payments. They may be unable to design a ratings-based reward system that does not have inherent incentives for side payments. The second result, in our opinion, is quite surprising . It suggests that marketing managers might be advised to invest more time into understanding how side payments affect employee reactions to reward systems. They might want to reconsider costly efforts to monit or, police,or preclude such side payments. While our results do not su bstitute for a moral discussion of side payments, we hope that the for mal structure for one common marketing situation provides valuable ins ight. The system we analyze is based on a practical managerial problem we have observed. The salesforce evaluates a sales support group with a real-valued rating. The sales support group is rewarded based on th at rating, whereas the salesforce is rewarded based on outcomes, such as sales or customer satisfaction, that indicate incremental profits t o the firm. The reward to the salesforce might also depend upon how it rates sales support. For example, the salesforce might be held to a h igher standard whenever it rates sales support as ''excellent.'' (We a rgue in the paper that the firm will want this to happen.) In addition , the salesforce might ask for a side payment from the sales support g roup as compensation for high ratings. We cast the practical problem a s a formal game and incorporate the following issues: (1) incremental actions taken by the salesforce and by sales support are perceived to be onerous, (2) the measure of incremental profit is a noisy measure, (3) both the salesforce and sales support are risk averse, (4) given t he reward system Imposed by the firm, both the salesforce and sales su pport will maximize their well-being, and (5) given the structure of t he reward system, the firm will seek to maximize expected profits. We first show that there are almost always incentives for side payments. Specifically, we demonstrate that sales support is better off with a s ide payment, while the salesforce is no worse off. This is not surpris ing because the reward to sales support is increasing in the rating, w hile in the absence of a side payment, the salesforce will select a ra ting such that its net marginal returns to increasing the rating are z ero. The exception occurs when the rating is constrained by the firm t o be less than this ''optimal'' racing, but even then there might be i ncentives for side payments. We next show that the firm can anticipate these side payments and design a reward system to factor them out at no loss of profit. The intuition is straightforward. The firm first ad justs the marginal returns in the reward functions for sales support a nd for the salesforce such that they will each take the ''optimal'' ac tions even though they engage in side payments. Then the firm adjusts their fixed compensation so that the firm extracts its full profit. Th e proof is difficult because we must show that adjusted reward systems exist and we must show that they allow the full profit to be extracte d. Throughout the paper we discuss the practical implications of our r esults. We close by highlighting future research opportunities.