Google’s Zero Moment of Truth
Consumer have the first experience of an object that they want to buy when they go online and read related reviews of previous customers.


Understanding the structure of the social networks and the information channels in a region of the developing world is fundamental for the success of any economic or technological project that involves the citizens’ participation. Word-of-mouth is often the only way information is efficiently transmitted, and some individuals have a role in society that makes them more central and therefore more effective in the information transmission. The method of information distribution may be different (word-of-mouth vis-à-vis social network) or the speed at which such information is spread might vary but in essence they have a very similar network structure and pricing schemes can be more aptly designed to cater to such behavior.

Revenue management problem with social learning

Launching a new product involves uncertainty. Specifically, consumers may not initially know the true quality of the new product, but learn about it through some form of a social learning process, adjusting their estimates of its quality along the way, and making possible purchase decisions accordingly. The dynamics of this social learning process affect the market potential and realized sales trajectory over time. The seller’s pricing policy can tactically accelerate or decelerate learning, which, in turn, affects sales at different points in time and the product’s lifetime profitability. We study a monopolist’s pricing decision in a market where quality estimates are evolving according to such a learning process. We contribute in three ways. First, in terms of modeling, by specifying a social learning environment that tries to capture aspects of online reviews as well as the possible bounded rationality of consumers. Second, by proposing a tractable methodological framework, based on mean-field approximations, to study the learning dynamics and related price optimization questions in the presence of social learning. Third, in addressing some of the pricing questions faced by revenue maximizing sellers in such settings.

This is the first step to analyze a larger class of problems of incentive design in the presence of learning dynamics in the system. This broader analysis would not only provide a unification of the small existing literature, it would also simplify the analysis of future problems of this type that have not been examined, yet.