Short answer
Researchers at Frankfurt School and the Vienna University of Economics and Business find that ChatGPT ads deliver clearly lower returns than Google search ads on the e-commerce sites they studied. The study is still under peer review, so read the findings in that light and back any budget decision with your own account data.
Highlights
- According to the study, every 1 dollar spent on ChatGPT ads returned 0.52 dollars in revenue, against 3.40 dollars on Google Ads.
- Revenue per visitor was 0.89 dollars for ChatGPT ads and 2.45 dollars for Google search ads; conversion rates were 0.37 per cent and 1.17 per cent.
- The study is still under peer review; free referral traffic from ChatGPT fell by 6 to 8 per cent after ads launched.

3 min readAuthor: UNIT Journal EditorEditor-in-chief: Uğur Deniz İlhan
The first comparative data has arrived for brands that want to advertise in ChatGPT. According to a study reported by Marketing Türkiye, ChatGPT ads trail Google search ads by a wide margin on e-commerce sites.
What did the study measure?
The study was written by Christian Schulze of Frankfurt School of Finance & Management and Maximilian Kaiser of the Vienna University of Economics and Business. It examined about 10 billion site visits to 1,898 international online stores between June 2025 and August 2026. Ad performance was assessed across 387 websites, and return on investment was calculated for the 102 advertisers with complete revenue data.
How big is the gap between ChatGPT and Google ads?
Revenue per visitor was 0.89 dollars for ChatGPT ads and 2.45 dollars for Google search ads. The purchase conversion rate was 0.37 per cent for ChatGPT and 1.17 per cent for Google. Measured against spend, every 1 dollar spent on ChatGPT returned 0.52 dollars in revenue, while Google Ads returned 3.40 dollars; the researchers find Google 6.6 times more efficient on this measure. Of the 102 advertisers with complete revenue data, 82 per cent made no sales at all from ChatGPT ads.
Did the ads affect organic traffic too?
According to the study, free referral traffic from ChatGPT to e-commerce sites fell by 6 to 8 per cent after ads launched. The conversion rate of visitors arriving through those referrals did not change significantly.
The study has not yet passed peer review. Its findings should be read in that light when applied to other markets and budget decisions; the data comes from international e-commerce sites and includes no Turkey-specific measurement.
What should brands do?
These results do not show that ChatGPT ads are worthless, only that their current performance trails Google search. In Google Ads management and performance marketing work that decides budgets channel by channel, these steps stand out:
- Before trying a new channel, set up separate conversion and revenue tracking for it; separate ChatGPT traffic with a source tag.
- Run the test with a small, time-limited extra budget without cutting your existing Google Ads spend, and compare revenue per visit and revenue per unit of spend.
- Protect the proven conversion data in your search ads; plan AI channels as a separate discovery and testing budget.
- On the organic side, remember that appearing as a source in ChatGPT answers has value independent of ads; monitor your free referral traffic separately.
- Until the study clears peer review or your own tests produce results, do not shift channel budgets by a large amount.
What happens next?
The peer-review outcome and the development of the ChatGPT ad product will show whether this picture changes. Your own account's measurement is a stronger basis for a decision than a single study.
Sources
- Marketing Türkiye ·
Follow UNIT Journal
What's new in search, AI and technology, in your feed every day.


