Personalized Collaborative Recommendations in the Mass-retailing Sector: the Impact of the Recommended Products and the Accompanying Message on Consumer Behavior

Vandenbulcke, Virginie [UCL] Ducarroz, Caroline [UCL] Fouss, François [UCL]

This study examines the impact of personalized collaborative recommendations and the accompanying message on consumer-behavior in the mass-retailing sector. We realize a between-subject experiment where we sent a newsletter on 229 mass-retail customers. This exploratory study aimed at testing if personalized recommendations based on collaborative filtering techniques are more efficient than non-personalized ones. These recommendations are defined on what other shoppers, who act similarly in terms of shopping basket, bought. Our main results show, first, the message effect is positive and significant on the quantity of recommended products bought. There is also an interaction effect showing that the kind of recommendation (personalized vs. non-personalized) received by the customer increases the message effect on purchases. Secondly, customers who receive collaborative personalized recommendations buy a larger quantity of products in the product category of each recommended product.