Network Traces on Penetration: Uncovering Degree Distribution from Adoption Data
Presentation
Based on Marketing Science article Dover, Yaniv, Jacob Goldenberg, and Daniel Shapira (2012), “Network Traces on Penetration:
Uncovering Degree Distribution from Adoption Data,” Marketing Science, forthcoming.
Teaching Objective
The objective is to show how the consumers' network structure (distribution of the individual-level number of links) impacts sales. We also discuss a method for uncovering the connectivity distribution of the consumer network underlying the product dissemination process, based exclusively on limited early-stage penetration data. We then explain how the method for estimating the network can be used to also improve sales forecasting. Essentially, the method estimates the network connectivity distribution and the diffusion parameters, at an early stage (15%) of the penetration process and then uses the estimation to forecast future sales.
Keywords: social network, word of mouth, diffusion, forecasting