Peer Effects from Friends and Strangers: Evidence from Random Matchmaking in an Online Game

50 Pages Posted: 31 May 2022

See all articles by Daniel Goetz

Daniel Goetz

Rotman School of Management

Wei Lu

Rotman School of Management, University of Toronto

Date Written: May 22, 2022

Abstract

The product adoption decisions of friends and strangers can both strongly influence consumers' purchase behaviour, but their importance relative to each other is unknown. In this paper, we use novel data from an online gaming app to quantify the relative magnitudes of these social interaction (peer) effects from friends and strangers in the same setting. Our empirical strategy leverages the game's quasi-random matchmaking system to identify the causal effect of observed adoptions on in-game purchasing. We find that the marginal effect of friends' product adoptions is almost twice as large as the marginal effect of strangers' product adoptions, and that the two types of peer effects are substitutes. Our results have managerial implications for optimal product seeding, and we write and estimate a network diffusion model to evaluate alternative seeding strategies that leverage peer effects from friends versus strangers. Finally, we provide evidence that although both types of peer effect increase awareness of the product, observing friends' adoptions also generates a mere exposure effect.

Keywords: Peer effects, product seeding, social learning, online video games

JEL Classification: D40, L2, L82, M3

Suggested Citation

Goetz, Daniel and Lu, Wei, Peer Effects from Friends and Strangers: Evidence from Random Matchmaking in an Online Game (May 22, 2022). Available at SSRN: https://ssrn.com/abstract=4116806 or http://dx.doi.org/10.2139/ssrn.4116806

Daniel Goetz (Contact Author)

Rotman School of Management ( email )

United States

Wei Lu

Rotman School of Management, University of Toronto ( email )

Toronto, Ontario
Canada

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
44
Abstract Views
170
PlumX Metrics