Discussion

Early data puts ChatGPT ad visits behind Google Search, with an important pricing caveat

In The AI Economy

OpenAI Watch
OpenAI WatchParticipantOpening post
#13755

Early data from 102 websites found that visitors arriving via ChatGPT ads generated less revenue per session, and converted less often, than visitors from Google Search. For advertisers weighing where to spend, the measured results are more useful than a headline return-per-dollar figure, which depends on assumed click prices.

OpenAI Watch analysis

What happened

Search Engine Journal’s Matt G. Southern says the study found US ChatGPT ad sessions generated $0.89 in revenue per session, compared with $2.45 for paid search. Transactions per session were 0.37% for ChatGPT ads and 1.17% for search. The analysis covered 60 weeks of data through 22 August, with up to 28 weeks of ChatGPT ad sessions available for US traffic. Of 87 advertisers using both channels, 84 generated less revenue per session from ChatGPT ads.

The study describes those results, rather than proving that switching a particular advertiser’s budget would cause performance to fall. The more eye-catching $0.52 in revenue per ChatGPT ad dollar, compared with $3.40 for search, combines measured session revenue with click prices drawn from different sources and accounts. It is an estimate, not a like-for-like measurement of what every advertiser earns.

Why it matters

An advertiser could be asked to move budget on the strength of a per-dollar comparison that depends heavily on what they pay per click. The reported session figures offer a clearer starting point, but they still do not settle whether ChatGPT ads make sense for a particular campaign, audience or objective.

Southern also notes that the study’s authors call the work descriptive, not causal. OpenAI told him the study captures an earlier period and a small sample, and pointed to products launched since then, including Target CPA and Target CPC. That is relevant context, but it does not erase the results or establish how newer bidding tools perform.

Our read

The sensible takeaway is not “move every pound to search” or “ChatGPT ads do not work”. It is that the $0.52 figure is a poor shortcut for a budget decision unless its click-price assumptions fit your account. Compare actual spend and attributed revenue over the same period, with the same attribution settings, and put session counts beside the results. Marketing dashboards have enough opportunities to look conclusive without borrowing certainty from mismatched prices.

What to watch

  • Whether further research measures newer ChatGPT ad formats using larger or later datasets.
  • Whether advertisers publish comparable results using matched campaigns and clearly stated attribution settings.
  • Whether the weaker per-session results hold across different industries, campaign sizes and objectives.

Discussion spark: Would you test ChatGPT ads despite the lower reported revenue per session, or should advertisers wait for larger, matched comparisons before shifting budget?

Sources and evidence

OpenAI Watch is independently operated by WittyWires. It is not affiliated with, endorsed by, or operated by OpenAI.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.