The marketing and general business literature's have successfully argued that relationship marketing
                         (RM) is critically important in business-to-business selling. The objective of this thesis is to quantify the
                         effect of RM on the producing firm. The most notable economic effect is reflected in the sales volume,
                         which is determined by two factors: the number of buyers, and the buying rate. Accordingly, we have
                         formulated an aggregate-level (to determine the number of buyers) and an individual-level (to study the
                         buying rate) model. To analyze effectiveness of RM, we developed an aggregate-level model of
                         repeat-buying (that is, of the joint presence of diffusion and repeat-purchasing effects) that incorporates
                         marketing control variables such as advertising, promotions, and pricing. The model departs from the
                         extant literature in yet another significant way: as a result of the personal selling process, the innovation
                         and imitation roles of the potential buyers are observable (sales representatives deduce and report on
                         the past action of buyers, their standing in the community, etc.). This factor is in contrast with the
                         traditional Bass model of diffusion, where these roles are observationally inseparable. Accordingly, we
                         develop an extension of the Bass diffusion model and repeat-purchase that explicitly incorporates the
                         observability of the aforementioned roles, and the marketing control variables. The model is validated
                         empirically on a data set from the pharmaceutical industry.
 
 

Social Systems Simulation Group P.O. Box 6904 San Diego, CA  92166-0904 Roland Werner, Principal Phone/FAX  (619) 660-1603