Personalizing Ad Load to Optimize Subscription and Ad Revenues: Product Strategies Constructed from Experiments on Pandora
The role of advertising as an implicit price has long been recognized in economics and marketing, yet the effects of personalizing these implicit prices on firm profits and consumer welfare have not been explored. Using a large-scale field experiment on Pandora, we exogenously changed the ad load–the number of ads played per hour–for over seven million users over 18 months. We find that the impact of ad load on consumption takes over a year to stabilize, while subscription responses stabilize in less than six months. Using a machine learning model, we analyze the heterogeneous effects of changing ad load on ad and subscription revenues. We then show that reallocating ads across users can enhance subscription profits by 7% without decreasing overall advertising profits. To achieve the same subscription rate with a uniform ad allocation policy, the firm would need to increase ad load by more than 22%. Our results show that, on average, consumer welfare drops by 2% with the proposed personalization strategy, and the effect seems to be more pronounced for users who have a higher willingness to pay.This version: 8 July 2024