Analysis of Umbrella Branding with Crowdsourced Data

Abstract

We test for umbrella effects among private label products using a new type of data: crowdsourced data, or data contributed by users of an in-store shopping app for their own, and other users' benefit. We model umbrella effects in our crowdsourced data using a random-parameter logit model with private-label preference parameters correlated among product categories. Our results support the existence of umbrella effects in our limited sample data, where umbrella effects are defined as correlated preferences for private labels across categories. [JEL Classifications: D12; D43; L13; L83; M31]. C 2016 Wiley Periodicals, Inc.