Web Search and Purchase Behavior in Durable Goods Markets with Network Effects

Abstract

I develop a dynamic structural model to explain consumers web search and purchase behavior in adoption of a new durable product in markets with network effects. In the model, consumers engage in pre-purchase web searches to form expectations on the network size of each brand, and optimize their purchase timing and brand choice based on those expectations. Decisions on web searches and purchases are the outcome of dynamic utility maximizing behavior. I apply the model to the online search volume data from Google Trends and the sales data from the U.S. video game console industry. The model is estimated with Nested Fixed Point algorithm, and the estimation results indicate that consumers are classified into three segments with different intrinsic preferences for each brand, sensitivity to information from web search, and search cost. In a policy simulation, I quantify the impact of purchasing a competitors brand name as a keyword for search advertising and draw managerial implications regarding keyword search advertising strategies.