No Evidence of Attraction Effect Among Recommended Options: A large-scale field experiment on an online flight aggregator

This article is the first to test whether the attraction effect can be replicated among recommended options in a real digital marketplace, namely – an online flight aggregator. For this purpose, we conducted a large-scale field experiment on an existing flight aggregator, in which we varied the number of the options recommended at the top of the result list. More precisely, we investigated whether recommending an additional “decoy” option, which is asymmetrically dominated by one of the two former recommended options (the cheapest or the fastest itinerary), increases users' conversion rate and impacts the market share of the recommended options. Our results, based on the analysis of more than 140,000 search sessions, suggest that this is not the case in general. We then conduct analyses on subsamples to investigate the boundary conditions of the attraction effect. This study questions the relevance of the attraction effect in online marketplaces and recommender systems and proposes new research avenues.