Platform-Controlled Search and Distortions in Attention Allocation

Digital platforms allocate buyer attention across sellers that differ in quality and breadth of appeal. We study a monopoly platform that sets meeting rates between buyers and two seller types --- niche sellers whose high-quality good is valued by a fraction of buyers and mass-market sellers whose good is valued by all. Sellers compete by posting prices à la Burdett and Judd (1983), so buyer surplus requires competition, while platform revenue requires seller rents. This difference creates a systematic distortion: as search capacity grows, the platform keeps high-quality niche attention just past the point where extra exposure stops creating rents and starts eroding them — its saturation point — and diverts the rest to mass-market sellers. The resulting buyer-surplus loss converges to a finite limit proportional to the quality gap between the two goods, split equally between an allocative loss from too little high-quality exposure and excess rent extracted by sellers. Applying the model to Amazon product search and Google passage-ranking data indicates that, for captive buyers, both platforms operate past the saturation point, with buyer-surplus losses of 63 and 44 percent of the planner's benchmark, respectively. Allowing buyer participation to respond to the platform's recommendation strategy disciplines the platform and shrinks this loss.