A paid ChatGPT visit earned $0.89 in revenue per session in U.S. traffic, against $2.45 for a paid Google Search visit, according to an October 2026 working paper by Maximilian Kaiser of the Vienna University of Economics and Business and Christian Schulze of the Frankfurt School of Finance & Management. Organic ChatGPT referrals to 1,898 online shops fell about 6 to 8% after ads launched, while conversion held.
Published November 3, 2026. Updated November 3, 2026. Sourced from the Kaiser and Schulze working paper, Search Engine Journal and PPC Land.
Key facts from the study:
- The paper is a working paper under review, dated October 2026, and covers 1,898 online shops that sell about $16 billion a year.
- In U.S. traffic, a paid ChatGPT visit earned $0.89 per session and a paid Google Search visit earned $2.45.
- Paid ChatGPT visitors converted at 0.37%, against 1.17% for paid search visitors.
- Organic ChatGPT referrals fell about 6 to 8% after ads arrived, and organic conversion did not change.
- At the rate of return of paid search, a ChatGPT ad click is worth about $0.26. The largest publicly known account paid $1.72 per click.
Table of contents
- How does a paid ChatGPT visit compare with paid search?
- Did ads change organic ChatGPT referrals to websites?
- What is a ChatGPT ad click worth to an advertiser?
- Can advertising pay for free ChatGPT access?
- How reliable are the study's findings?
- What should PR and marketing teams do with the findings?
- Frequently asked questions
How does a paid ChatGPT visit compare with paid search?
A paid ChatGPT visit generated $0.89 in revenue per session in U.S. traffic, a 64% shortfall against the $2.45 that a paid Google Search visit generated. Kaiser and Schulze measured the gap across advertiser websites and found that paid ChatGPT visitors also converted far less often.
| Measure (U.S. traffic) | Paid ChatGPT visit | Paid Google Search visit |
|---|
| Revenue per session | $0.89 | $2.45 |
| Conversion rate | 0.37% | 1.17% |
The authors conclude that paid ChatGPT visits perform worst among the major channels they compared, offering less than a sixth of the value for money of paid Google Search. They measured each channel against direct traffic, using conversion rate, average order value and revenue per session for 102 advertisers, and bounce rate, pageviews and visit duration for 387 websites.
The result gives advertisers a number to test against, and it does not say ChatGPT ads cannot work. OpenAI began testing ads in the United States on February 9, 2026 and opened a self-serve beta on May 5, 2026, as EPR's ChatGPT ads guide details, so formats and delivery tools are still maturing.
Did ads change organic ChatGPT referrals to websites?
Organic ChatGPT referrals to websites fell by about 6 to 8% after ads arrived, and the visitors who still arrived converted at the same rate as before. The authors treat organic referral volume and conversion as a proxy for whether ads lowered the value of ChatGPT to its users.
Referral volume declined 5.9% over the first 13 weeks after launch, with a 95% interval running from a 12.6% decline to a 1.2% increase. After allowing for ChatGPT's faster usage growth in North America than in Europe, the estimate moves to a decline of 7.2% to 8.4%.
Conversion barely moved. Organic ChatGPT conversion dipped by 0.02 percentage points from a 2.21% baseline before launch, a difference too small to separate from chance (p = .648). For brands that earn ChatGPT referrals, the practical reading is that people who click through from an answer remained as likely to buy as they were before ads appeared.
What is a ChatGPT ad click worth to an advertiser?
At the rate of return of paid search, a paid ChatGPT click is worth about $0.26 to an advertiser, according to the paper, while the largest publicly known account paid $1.72 per click. That puts actual prices more than six times above the parity price, which is the cost per click at which a ChatGPT click would return the same revenue per ad dollar as a paid search click.
Search Engine Journal's report on the study explains how any advertiser can estimate its own parity price. Divide the revenue per visit from ChatGPT ads by the revenue per visit from paid search, then multiply the result by the cost per click paid on search. With the paper's U.S. averages, that ratio is about 0.36 ($0.89 divided by $2.45), and the table applies it to four paid search costs per click.
| Paid search cost per click | Parity price for a ChatGPT click |
|---|
| $0.50 | $0.18 |
| $1.00 | $0.36 |
| $2.00 | $0.73 |
| $4.73 | $1.72 |
The table is EPR arithmetic using the paper's U.S. averages. The last row shows that the $1.72 click reported in the study returns search-level value only if a comparable paid search click costs about $4.73. An individual account's ratio will differ from the national average, so the formula matters more than the table.
OpenAI's own bidding guidance sits well above the parity range. EPR's guide to buying ChatGPT ads reports that OpenAI's Ads in ChatGPT: The Basics guide suggests a starting maximum bid of $3 to $5 per click on Clicks campaigns. A maximum bid is a ceiling and not the price paid, so actual costs can land lower.
Can advertising pay for free ChatGPT access?
On the paper's numbers, advertising does not yet cover the cost of free ChatGPT access. The authors conclude that current ad performance is not enough to finance what free access costs.
PPC Land's summary of the paper reports that advertising currently brings in about $1 per free user per year, while serving a free user costs about $5 in inference alone. The same summary adds that OpenAI's projections for 2030 imply about $40 per free user.
The study frames the question in two parts. Ads could keep AI assistants free and help ChatGPT compete with Google Search, but only if they do not make the product worse for users and if advertisers get enough value to justify the price. The organic results speak to the first condition, because referral conversion held. The paid results speak to the second condition, where the gap is large.
The market is moving quickly. The paper's data ends August 22, 2026, and Search Engine Journal reports that OpenAI has since expanded ChatGPT Ads to 89 countries and says more than 100,000 advertisers have run campaigns. Search Engine Journal also notes that OpenAI's ads FAQ says ChatGPT Ads does not yet have performance benchmarks across advertisers, industries or campaign types.
How reliable are the study's findings?
The findings come from a version 1 working paper that is still under review, so the figures are estimates that later versions may revise. The paper, titled "Can Advertising Fund Free AI? Evidence from ChatGPT's International Advertising Rollout," is dated October 2026 and is published on the authors' research site.
Kaiser and Schulze used Google Analytics e-commerce tracking provided free by Grips Intelligence, an e-commerce data company. Search Engine Journal reports that Kaiser is a scientific advisor to Grips and previously worked there. The paper states that Grips had no role in the research questions, design, analysis or writing.
The design compares countries where ads launched at different times with the European Union, where ads had not launched by the end of the data window. The window ran for 60 weeks from June 2025 to August 2026, and EU ads went live on August 24, two days after the data ended.
What data did the study use?
The organic panel covers 1,898 websites that sell about $16 billion a year, with about 10 billion visitor sessions and 22.4 million organic ChatGPT referrals. The paid panel covers 387 advertiser websites. The rollout covers nine treated countries plus the untreated EU.
One caution applies to the organic result. The 95% interval around the first estimate includes zero, so the 5.9% decline is a best estimate and not a certainty.
What should PR and marketing teams do with the findings?
PR and marketing teams can use the study to set a cost ceiling for ChatGPT ad tests while continuing to build the earned mentions that ads cannot buy. Five steps follow from the evidence.
- Calculate the parity price from your own analytics. Divide ChatGPT ad revenue per visit by paid search revenue per visit, multiply by your paid search cost per click, and compare the result with the bids in Ads Manager.
- Start with the 25 USD daily minimum described in EPR's ChatGPT ads advertiser guide, and judge the test on revenue per session and conversion, the two measures where the study found the widest gaps.
- Tag paid and organic ChatGPT traffic separately with UTM parameters, so that organic referrals, which held their conversion rate in the study, are not blended with paid visits, which converted at 0.37%. Paid media in the answer-engine era covers how teams track clicks alongside citation share.
- Keep earning mentions inside answers. OpenAI states that ads run on separate systems from the chat model, so a campaign cannot buy a recommendation. EPR's guides on how to get recommended by ChatGPT and why ChatGPT recommends a competitor explain what moves an answer, and EPR's report on ChatGPT shopping retailers found four retailers holding 85.2% of cosmetics results across 34,637 prompt runs.
- Recheck the numbers when the next version of the paper appears, since the authors say they update their studies as new data comes in.
Sources: Kaiser and Schulze, Can Advertising Fund Free AI? working paper, Search Engine Journal report on the study, and PPC Land report on the study.
Related links: AdTech and MarTech, EPR research hub, ChatGPT ads advertiser guide, and ChatGPT shopping retailers.
