An Economic Framework for Generative Engines: Advertising or Subscription?

A team of researchers from Carnegie Mellon University has published a significant new economic analysis titled 'An Economic Framework for Generative Engines: Advertising or Subscription?' The paper tackles the central monetization dilemma facing providers of Generative Engines (GEs) like OpenAI's ChatGPT and Google's AI Overviews. These AI systems deliver synthesized answers that often bypass traditional websites, crippling those sites' ad revenue. However, this creates a new problem for the AI providers themselves: should they insert ads into their AI-generated responses, or keep them clean to drive premium subscription sales?

Using a dynamic framework that models how design choices affect long-term user engagement and retention, the researchers derived a clear, optimal policy. They found that providers should follow a 'cutoff rule,' showing ads to a user only when the immediate payoff from that ad exceeds the long-term value of providing an ad-free response to retain that user. This cutoff shifts toward showing more ads when ad rates are high or users are less sensitive to ads. Conversely, it shifts toward ad-free responses when the potential for converting users to valuable subscriptions is greater.

Crucially, the model introduces a competitive reality often missing from theoretical analyses. The presence of rival GEs—meaning users can easily switch from ChatGPT to Claude or another service—forces the optimal policy further toward ad-free responses. In a competitive market, ad-heavy monetization becomes unsustainable as it drives users to alternatives. This creates a powerful economic incentive for companies to prioritize user experience and long-term sustainability over short-term ad revenue, potentially leading to cleaner AI products for consumers.