Competitive information provision in sequential search markets

We study competitive information provision in search markets. Consider the used car market: as a consumer searches, she receives noisy signals of the values of cars. She can consult an expert (say Carfax or a mechanic) to find out more about the true value before deciding whether to purchase a particular car or keep searching. Prior research has studied the pricing problem faced by a monopolistic expert who provides searchers with perfect information. Here, we study a richer model that augments prior work in two important respects. First, we analyze expert duopolies; thus each expert must now reason about the influence of her strategy on the other. Second, we consider experts who provide uncertain information, with higher quality experts providing more certainty; experts can compete on both price and quality. We show that, in equilibrium, prices will be set such that the searcher consults the worse quality expert for low or high signals, and the higher quality expert for intermediate signals. Surprisingly, we find cases where an improvement in the quality of the higher-quality expert (holding everything else constant) can be pareto-improving: not only does that expert's profit increase, so does the other expert's profit and the searcher's utility.