We present a framework to study directed consumer search. Firms sell products with two attributes. One is readily observable, the other is observed only after visiting a firm. Search is directed as the order of search is influenced by the observ-R We are very grateful to Régis Renault, the Editor, and two anonymous referees for their constructive comments and suggestions. Previous versions of this paper circulated as "Consumer Search with Observable

No abstract

also analyzes obfuscation and the price choices of a multiproduct monopolist with horizontally differentiated products. Unlike the present article, he places fewer restrictions on the distributions of match values and focuses on markets with more than two products and observable prices. For these reasons, some of his results and welfare findings differ from the present article, including showing that obfuscation might improve welfare.By obfuscating some of its products, the monopolist controls the order in which its goods are sampled. In this sense, the present article is related to the literature on ordered searching. Arbatskaya (2007) shows that when consumers shop for homogeneous products, equilibrium prices must fall as a consumer proceeds to search for lower prices. Zhou (2011) demonstrates that when consumers shop for differentiated products, equilibrium prices increase instead. Armstrong, Vickers, and Zhou (2009) explore the role of prominence in search markets and show that a prominent product is sold at a lower price relative to nonprominent products. The main difference between those articles and the present one is that this study focuses on the role of search frictions among the products provided by a single firm. Zhou (2009) also observes that a multiproduct monopolist earns higher profits if one of its products is prominent. My monopoly analysis extends his result by allowing the monopolist to choose the degree of obfuscation, and by treating differences in product quality and design. 3 The decision of a firm to obfuscate its products is somewhat related to its decision to advertise its products. Therefore, in this sense, the present article is related to the work of Caminal (1996), Anderson and Renault (2006), and Wang (2017). In all three models, the decision to advertise affects consumers' decisions on whether to pay a costly visit to a firm, and the firm chooses the advertising option that leads to the highest possible consumer traffic. The major difference between those articles and the present one is that in those articles, a single-product advertising monopolist often earns less if fewer consumers search, whereas in the present article, the multiproduct monopolist gains from consumers who decide not to search much, in that they buy products that are more expensive, though less obfuscated. Janssen and Non (2009) show that there may be an equilibrium in a duopoly model with homogeneous products, in which advertising firms charge higher prices. Their result stems from the fact that consumers differ in their willingness to pay that is positively correlated with their search costs. In contrast in the present article, there is no search cost differentiation across consumers.By the fact that the mean valuation of a product by consumers who acquire the good decreases in the search order, my results resemble the findings of models with an intertemporal pricediscriminating monopolist (see, e.g., Coase, 1972;Stokey, 1979Stokey, , 1981Bulow, 1982). However, unlike with that monopolist, its multiproduct ...