[Séminaire CREM] In-house vs. Outsourced Data Analytics : Implications on Competition and Consumer Targeting

Abstract:
In a digital context, firms can gather abundant data on consumer behavior and use these data to implement data analytics to target consumers through price discrimination. To do so, they can either implement data analytics internally (in-house) or externally (outsourcing) via a more efficient data intermediary. This paper investigates the impact of the firms’ trade-off between in-house versus outsourced data analytics on consumer targeting and competition and contrasts two types of pricing strategies for the data analytics service (lump-sum payment vs. wholesale pricing). In the case of a lump-sum payment, we find that when firms’ costs related to data analytics are symmetric, only one firm outsources if product differentiation is low, while both firms outsource otherwise. When firms’ data processing costs are asymmetric, only the high-cost firm outsources if product differentiation is low; only the low-cost firm outsources if product differentiation is moderate; and both firms outsource if product differentiation is high. When the data intermediary uses a wholesale pricing strategy, and when data processing costs are symmetric, we find that both firms prefer in-house data analytics if competition in the downstream market is low ; and both firms outsource if competition is relatively high. Finally, when costs are asymmetric, the optimal contracting strategy still depends on the intensity of market competition, yet in a more subtle way i.e., gradually increasing the intensity of market competition makes the firm ”oscillates” between full internalization and outsourcing. We also study how the data intermediary is able to strategically exploit its cost-efficiency to drive the firms’ trade-off to an outcome that will favor its position.