Friends, Key Players and the Adoption and Use of Experience Goods

TL;DR

This study empirically examines how social networks influence the purchase and use of experience goods, specifically video games, revealing significant peer effects and differences based on friendship types.

Contribution

It provides novel empirical evidence on peer effects in experience good adoption, using a large-scale dataset and an instrumental variable approach to identify causal influence.

Findings

Strong peer effects in game adoption within the same week

Older friends exert a stronger influence than key players

Players influenced by non-key, non-old friends spend more time playing

Abstract

This paper empirically investigates how an individual's network influences their purchase and subsequent use of experience goods. Utilising data on the network and game-ownership of over 108 million users from the world's largest video game platform, we analyse whether a user's friendship network influences their decision to purchase single-player video games. Our identification strategy uses an instrumental variable (IV) approach that employs the temporal lag of purchasing decisions from second degree friends. We find strong peer effects in the individual game adoption in the contemporary week. The effect is stronger if the friend who purchased the game is an old friend compared to a key player in the friendship network. Comparing the results to adoption decisions for a major label game, we find peer effects of a similar size and duration. However, the time subsequently spent playing the games is higher for players who were neither influenced by a peer who is a key player nor an old friend. Considering the increasing importance of online networks on consumption decisions, our findings offer some first insights on the heterogeneity of peer effects between old and key player friends and also provide evidence in consumers' biases in social learning.