How crowdsourcing improves prediction of market-oriented outcomes

Abstract

Firms often struggle to be proficient in predicting uncertain market conditions and forecasting the outcomes of their business initiatives. This research introduces crowdsourcing as an innovative tool that can enhance market information processing, and in turn, improve prediction of market-oriented outcomes (e.g., sales). We field test a forecasting tournament with employees at a Fortune 100 consumer packaged goods firm, and examine the extent to which predictions based on the “wisdom of the crowd” outperform those generated by traditional forecasting approaches. We find that crowdsourcing produces results superior to the firm's incumbent approaches almost three-fourths of the time across a broad range of business decisions. Additionally, we conduct a survey with participants to open up the “black box” of crowdsourcing. We find that differences in information acquisition and interpretation are the underlying mechanisms that can explain the improved prediction accuracy found through crowdsourcing.