ChatGPT ads proof now stops where user trust starts. Adweek reported that monetization chief Benji Shomair told partners OpenAI will not share ChatGPT conversations or sell user data to advertisers. Buyers can still bid and pass conversions back; they cannot audit the prompt or treat outside studies as platform truth.
Updated November 9, 2026
Key Takeaways
- OpenAI will not share ChatGPT conversations or sell user data to ad partners.
- U.S. paid ChatGPT visits returned $0.89 a session, versus $2.45 on Google Search.
- OpenAI lacks IAS, DoubleVerify, and Zefr, plus reliable cross-advertiser benchmarks.
- Adthena tracked 6,333 ChatGPT ad brands in August, up from 1,082 in June.
What can ChatGPT ads proof still cover?
INMA reported that cost-per-click carries the majority of spend, per global ads chief Dave Dugan, with conversion-optimized CPC and a target CPA rolling out in August. On August 24, OpenAI added a “maximise results” automated bid strategy, platform targeting, and one-day view-through conversions. Digiday reported that OpenAI recently said CPC and outcome-optimized bidding account for the majority of its ad campaigns.
Buyers can also upload first-party lists, build custom audiences, set geo include and exclude, and use negative “context hints.” LiveRamp was the first conversions-API partner in June; AppsFlyer followed on August 4, with Grubhub and 40 other brands testing in-app attribution, plus Kochava and WorkMagic. In this analysis, closed-loop proof covers outcomes the advertiser already owns: a click, a passed-back conversion, or a one-day view-through, not the chat that triggered the ad.
What can’t buyers prove without conversation data?
As Adweek reported from a Luma Partners event, Shomair said: “There are places where our partners would like us to do certain things that we just can’t because of how we think about making decisions.” Conversation exports and any sale of user data sit on the far side of that line.
- ChatGPT conversations are not shared with outside ad partners.
- User data is not sold to advertisers.
- No IAS, DoubleVerify, or Zefr integration, and no reliable cross-advertiser benchmarks.
- Exclusion targeting was a small-group test after buyers complained of little control or visibility into where ads run.
Relevance sits outside the log, too. Three independent studies reported by Adweek found ads on about a quarter of commercial queries, roughly one in seven with no topical connection to the prompt, and only 38% directly relevant. Accuracast group CEO Farhad Divecha told Digiday that the latest Magnite and PubMatic integrations have not necessarily addressed the underlying tracking issues. “We continue to see discrepancies there,” he said. “Some are teething problems because it’s a new system that has to be set up on advertiser systems, but some problems are down to attribution being messed up.”
How does ChatGPT ads proof compare with paid search?
The clearest outside comparison is a working paper under review by Maximilian Kaiser and Christian Schulze, based on ecommerce sites and reported by Digiday. Read the table as an outside estimate, not as an OpenAI benchmark.
| U.S. paid-visit metric | ChatGPT | Google Search |
|---|---|---|
| Revenue per session | $0.89 | $2.45 |
| Conversion rate | 0.37% | 1.17% |
Agency reads split. PMG has said ChatGPT returns about twice the average of other platforms at today’s low spend levels. Lifesight’s Rajeev Nair told AdExchanger the platform is “very rudimentary” and dollars “will remain in ‘test’ budgets.” INMA reported that MediaPost collected small-advertiser tests with 0.6% clickthrough rates, US$7 clicks, zero conversions, and OpenAI click counts two to three times what Google Analytics recorded. In this analysis, none of that replaces a query log.
What should buyers keep on search and social?
Keep on ChatGPT the briefs that survive a black box: first-party list matches, geo include and exclude, negative context hints, and conversion bids where the advertiser’s conversions API is the source of truth.
Keep on search and social any brief that needs a query audit, third-party verification, or a conversion rate a quarterly review will defend. Sensor Tower data in Digiday shows why the split is already live: U.S. shopping’s share of ChatGPT ad spend fell from 38% in April to 20% in September, while financial services rose from 2% to 13%. In this analysis, those later categories usually need harder proof than a test line can carry.
Frequently asked questions
Can advertisers see the ChatGPT conversations behind their ads?
No. Adweek reported that Benji Shomair said OpenAI will not share ChatGPT conversations with outside ad partners or sell user data to advertisers.
How do paid ChatGPT visits compare with Google Search?
A working paper under review, reported by Digiday, found paid ChatGPT visits generated $0.89 per U.S. session and converted at 0.37%, versus $2.45 and 1.17% for paid Google Search.
What measurement can buyers actually connect?
INMA reported LiveRamp, then AppsFlyer, Kochava, and WorkMagic as conversion partners, plus target-CPA bidding and one-day view-through. IAS, DoubleVerify, and Zefr are not integrated.
Are early ChatGPT returns enough to leave test budgets?
Lifesight’s Rajeev Nair told AdExchanger the platform is “very rudimentary” and dollars “will remain in ‘test’ budgets.” PMG has cited about twice the return of other platforms, but only at today’s low spend levels.
