Ali Kakhbod

Strategic Information Diffusion: Spread vs. Exploit

Frequent drops of prices to zero is a common phenomenon in price trends of many smartphone applications. The only means by which many of these apps spread is the word of mouth of their users. Motivated by these observations, we study the problem of optimal dynamic pricing in a social network where agents can only get informed about the product via word of mouth from a friend who has already bought the product. We show that for a durable product such as many apps, the optimal policy should drop the price to zero infinitely often, giving away the immediate profit in full to expand the informed network in order to exploit it in future. We further show that, beside the word of mouth nature of the information diffusion, this behavior crucially depends on the type of the product being offered. For a nondurable product, although the firm may initially make some free offers to expand its network, after a finite period, it will fix the price at a level that extracts the maximum profit from the already informed population.

Advising the Management

We study the optimal size and composition of an advisory committee when shareholders differ in preferences and beliefs and strategically acquire and communicate information. If shareholders and management have similar objectives but disagree due to different beliefs, and information is cheap, the optimal advisory body includes all shareholders. Conversely, if agents have conflicting preferences or information is sufficiently costly, the optimal advisory body is a strict subset of shareholders. Thus, advisory voting (board) is optimal in the former (latter) case. Similar implications hold if the committee also has authority, but unlike purely advisory committees, committees with authority are more diverse.

Optimal Contracting in Networks

We study optimal contracting between a firm selling a divisible good that exhibits positive externality and a group of agents in a social network. The extent of externality that each agent receives from the consumption of neighboring agents is privately held and is unknown to the firm. By explicitly characterizing the optimal multilateral contract, we demonstrate how inefficiency in an agent’s trade propagates through the network and creates unequal and network-dependent downward distortion in other agents’ trades. Furthermore, we describe bilateral contracts (non-linear pricing schemes) and characterize their explicit dependence on the network structure. We show that the firm will benefit from uncertainty in an agent’s valuation of other agents’ externality. We describe the profit gap between multilateral and bilateral con- tracts and analyze the consequences of the explicit dependence of the contracts on network structure. When the network is balanced in terms of homogeneity of agents’ influence, network structure has no impact on the firm’s profit for bilateral contracts. On the other hand, when the influences are heterogeneous with high dispersion (as in core-periphery networks) the restriction to bilateral contracts can result in profit losses that grow unbounded with the size of networks.

Types of Conflict in Strategic Communication

We study communication in coordination games with either conflicting preferences or conflicting beliefs. We show these two ways of modeling conflict lead to drastically different conclusions. While for any type of conflict truthful communication is always beneficial from the ex-ante perspective, there is however a credibility issue at the interim stage, i.e., information transmission pattern crucially depends on the nature of disagreement. Under conflicting preferences information transmission exhibits a negative externality effect: greater information obtained by some agent discourages further information accumulation by harming the credibility of other agents. In contrast, under conflicting beliefs information transmission exhibits a positive externality effect: greater information obtained by some agent encourages further information accumulation by improving the credibility of other agents. By applying our findings to several frameworks, we show how strategic motives, depending on the types of conflict, crucially affect information aggregation in polls and policy outcomes, direction of information flow in communities, communication network formations, and segregations and aggregations in communications due to adding new information. Our result highlights the importance of taking into account the types of conflict between individuals in strategic communication models.