In Graphic Detail: Why OpenAI’s ad business is still a work in progress

The key tradeable signal is not “ChatGPT ads work,” but that OpenAI is likely to monetize first through control of discovery, not through high-volume ad load. That favors infrastructure providers that help brands become machine-readable and measurable, while punishing legacy search and SEO intermediaries whose value proposition depends on ranking opacity. The second-order effect is that budget shifts will initially come from experimentation dollars, not core performance budgets, which means ad tech and martech vendors with clear LLM optimization, feed management, and attribution hooks should see earlier traction than pure media sellers.

The engagement slowdown matters because it caps near-term inventory growth exactly when investor narratives are demanding a hypergrowth ramp. If time spent keeps drifting lower over the next 1-2 quarters, OpenAI will be forced into a tighter monetization mix: higher ad load, more sponsorship-like placements, or heavier commercialization of “commercial intent” queries. Any of those can improve ARPU, but they also raise the risk of user pushback and retention deterioration, creating a classic platform tradeoff that usually shows up with a 3-6 month lag rather than immediately.

Consensus appears too eager to extrapolate long-dated revenue targets into near-term equity upside for the broader AI stack. The more likely near-term winners are firms that sell the plumbing behind AI visibility and ad measurement, while the losers are businesses dependent on search traffic arbitrage and generic content scaling. If OpenAI’s rollout remains cautious, the market may overprice the benefit to large ad platforms and underprice the benefit to niche data/optimization vendors that can win share before the ad product fully scales.