New working paper · Can advertising fund free AI? Evidence from ChatGPT's international advertising rollout → Paid ChatGPT visits returned about a third of a paid-search visit's value Organic ChatGPT referrals fell by about 6 to 8% after ads arrived; conversion held
ecommerceresearch.org Available for advisory & talks

§ Research / chatgpt-ads

/chatgpt-ads

Can Advertising Fund Free AI? Evidence from ChatGPT's International Advertising Rollout

Maximilian Kaiser · Christian Schulze

StatusWorking paper · under review
Versionv1 · October 2026
Data1,898 websites · 60 weeks
DesignStaggered rollout · EU untreated
Download the paper (PDF) ↓
On this page 01Summary02Findings03Data04Materials05In the press06Cite07Versions08Updates

01 · Summary

Ad performance does not yet cover the cost of free access

Advertising could keep AI assistants free and make ChatGPT a Google Search competitor if advertisements do not devalue the product and justify their price. This study tests both conditions during ChatGPT's staggered international rollout, with the European Union untreated, using data from 1,898 online shops selling $16 billion annually.

The introduction of advertisements moderately reduces the volume but not the conversion of organic ChatGPT referrals, a proxy for product value. However, paid ChatGPT visits perform worst among major channels, offering less than a sixth of the value for money of paid Google Search. Current performance is not sufficient to finance free access costs.

02 · Findings

Findings

2.1

After the introduction of ads, organic ChatGPT referrals fell by about 6 to 8%.

Volume changed by −5.9% over the first 13 post-launch weeks (95% interval −12.6% to +1.2%). Allowing for ChatGPT's faster usage growth in North America than in Europe moves the estimate to −7.2% to −8.4%.

2.2

Visitors who still arrived converted as before.

Organic ChatGPT conversion changed by −0.02 percentage points from a 2.21% pre-launch baseline (p = .648).

Event studies of organic ChatGPT referral volume and conversion around the ad launches
Figure 3Organic performance at the ad launches and its eligibility sensitivity: referral volume (A) and conversion rate (C) in Sun–Abraham event studies with website-clustered 95% intervals; (B, D) the same estimates as the assumed share of sessions from ad-eligible users falls from 100% to 66%.
2.3

Paid ChatGPT visits generated less value for advertisers than paid search.

In US traffic, a paid ChatGPT visit generated $0.89 in revenue per session against $2.45 for a paid Google Search visit, a 64% shortfall. Paid ChatGPT visitors converted at 0.37% versus 1.17%.

2.4

Actual prices are more than six times the parity price compared to Google Search.

At paid search's rate of return, a paid ChatGPT click is worth about $0.26. The largest publicly known account paid $1.72 per click.

Paid ChatGPT versus established channels on six metrics, each relative to direct traffic
Figure 4Performance of paid ChatGPT versus established channels, each measured against direct traffic: conversion rate, average order value and revenue per session for 102 advertisers (A–C); bounce rate, pageviews and duration for 387 websites (D–F). Whiskers show 95% intervals.

03 · Data and design

A staggered rollout, measured from outside the platform

1,898Websites · organic panel
387Advertiser websites · paid panel
60Weeks · Jun 2025 – Aug 2026
~10 bnVisitor sessions
22.4 mOrganic ChatGPT referrals
9 + EUTreated countries · untreated EU
Map of rollout cohorts and weekly organic ChatGPT sessions by region
Figure 2The rollout and the organic ChatGPT channel: treated cohorts and the untreated EU (A), weekly organic ChatGPT referral sessions by region (B), and pre-rollout conversion of organic ChatGPT and organic search (C).

The data comes from Grips Intelligence, which provided it free of charge and had no role in the research questions, design, analysis or writing. Maximilian Kaiser serves as its scientific advisor.

04 · Materials

Paper · PDFWorking paper (PDF)The complete working paper with the Web Appendix.PDF ↓ Full text · MDFull text in MarkdownThe complete working paper with the Web Appendix as plain text, for reading tools and AI assistants..md ↓ Press · PDFFact sheet for journalistsThe story and the key numbers on one page.PDF ↓ Slides · PPTXSlides with all figuresAll figures of the paper with the key numbers, 17 slides, for teaching and talks.PPTX ↓

05 · In the press

Search Engine Land2 Nov 2026 ChatGPT ads generate 64% less revenue per visit than Google Search: Study PPC Land2 Nov 2026 ChatGPT ad visits earn shops $0.89 against $2.45 for Google search ads Search Engine Journal2 Nov 2026 Google Search Ad Visits Out-Earned ChatGPT Ads In U.S.

More coverage →

06 · How to cite

Kaiser, M., & Schulze, C. (2026). Can Advertising Fund Free AI? Evidence from ChatGPT's International Advertising Rollout. Working paper. https://ecommerceresearch.org/chatgpt-ads/

@unpublished{kaiser_schulze_2026_ads,
  author = {Kaiser, Maximilian and Schulze, Christian},
  title  = {Can Advertising Fund Free AI? Evidence from
            ChatGPT's International Advertising Rollout},
  year   = {2026},
  note   = {Working paper},
  url    = {https://ecommerceresearch.org/chatgpt-ads/}
}

07 · Versions

v1Oct 2026First public version