§ Research / organic-llm
/organic-llm
ChatGPT Referrals to E-Commerce Websites: How Do LLMs Compare Against Traditional Channels?
01 · Summary
Organic LLM traffic underperforms most traditional channels
Organic LLM traffic (oLLM) is visitors who reach websites through organic, unpaid links in ChatGPT's answers rather than through search, ads, email or social media. This study is the first to measure it at scale, with 12 months of first-party Google Analytics data (August 2024 to July 2025) from 973 e-commerce websites with $20 billion in combined revenue.
One year after launch, oLLM converts and earns per session below all traditional channels except paid social. At under 0.2% of all visits it is promising, but not yet a mass channel. ChatGPT accounts for over 90% of all LLM-referred sessions.
02 · Findings
Findings
A new sales channel, measured for the first time.
When someone asks ChatGPT and clicks through to a website, that visit is organic LLM traffic. The study traces over 50,000 ChatGPT-referred purchases on 973 websites and benchmarks them against 164 million purchases from the eight established channels: organic search, paid search, direct, email, affiliate, referral, paid social and other.
Today it is a niche, not yet a mass channel.
Adjusted for website, device, time and data sparsity, an oLLM visit converts less often than almost every established channel. Affiliate links convert about 86% more often than oLLM and organic search about 13% more often; paid social converts about 53% less often. On revenue per visit, too, oLLM ranks above paid social and below every other channel. oLLM accounted for less than 0.2% of all visits.
It is improving, most likely because customers are getting better at using AI.
Over the year, oLLM's conversion rate rose while traditional channels stayed flat, and its average order value fell. Websites whose visitors are more LLM-savvy show the same mix of higher conversion and lower order value. Because the pattern appears both over time and across audiences, the most consistent explanation is growing consumer skill at buying through AI. The authors call this suggestive, not proven cause and effect.
For complex products, it already out-converts five of eight channels.
Websites selling complex products receive about 4.6 times more oLLM traffic share than websites selling simple products. On these websites, oLLM's conversion rate beats five traditional channels: paid social, referral, email, organic search and direct.
03 · For managers
Match the response to the product
Simple, routine products. The study measures the lower funnel, whether a visit converts and what it earns, and there oLLM is still weak for routine purchases. Do not shift lower-funnel budget out of proven performers such as affiliate and search. LLMs may still matter in discovery, which last-click data cannot capture.
Complex products. Here oLLM already deserves attention: it pairs comparatively strong conversion and revenue per visit with low bounce rates. Treat it as a high-intent complement to organic and paid search, and make product content easy for AI assistants to read and cite.
04 · Data and design
What the study is based on
The data comes from Grips Intelligence, read directly from each website's own Google Analytics. Maximilian Kaiser is affiliated with Grips Intelligence.
05 · Materials
06 · In the press
07 · Questions
Frequently asked questions
Q1What is organic LLM traffic (oLLM) / AI referral traffic?
Organic LLM traffic (oLLM) is human visitors who reach an online store by clicking an organic link that ChatGPT — or another LLM platform — shows when someone expresses purchase intent. It's real people clicking through, not the AI crawlers or bots that scrape sites for training. The channel was effectively born in August 2024, when ChatGPT began adding outbound product links to its answers, creating a genuinely new way for customers to arrive at an e-commerce website.
Q2Does “organic” mean it's free?
Yes. Throughout the study period these links were generated by the model itself, based on what it judged most relevant — they were not paid placements or ads, and no retailer could buy its way into them. We call it organic LLM traffic (oLLM) precisely to separate it from any paid LLM traffic (pLLM), such as the sponsored placements and ad formats platforms have since begun testing.
Q3Is AI traffic its own channel, or part of search?
We treat oLLM as a distinct new channel and benchmark it head-to-head against the eight channels retailers already budget for: organic and paid search, direct, email, referral, affiliate, paid social, and “other”. It behaves differently enough from each of them — in who it brings, how they engage, and whether they buy — that it earns its own line in your analytics rather than being folded into referral or direct.
Q4Does this study tell me how to optimize for ChatGPT (GEO)?
No, and that's deliberate. This is a descriptive study of how oLLM actually performs across hundreds of websites, not a how-to guide for getting recommended by ChatGPT. We flag generative engine optimization (GEO) as an important but still-open question: it isn't yet clear how a retailer can reliably raise its exposure in AI answers, its click-through, and its conversions, or how those steps relate. That's a frontier the paper points to, not one it solves.
Q5Does the study cover ChatGPT's ads, instant checkout, or shopping agents?
No — we measure organic ChatGPT referrals only. Features such as in-chat Instant Checkout, agentic shopping assistants, and paid ad placements all arrived after our observation window closed, and each could materially change the picture. We note them as developments to watch, but they are outside what this study measures, so we don't make claims about them.
Our follow-up working paper studies ChatGPT's ads directly: Can Advertising Fund Free AI?
Q6How much traffic does ChatGPT actually send to e-commerce sites?
Very little, so far. One year after launch, oLLM accounted for under 0.2% of all sessions in our data — roughly 200× smaller than Google organic search.
08 · How to cite
Kaiser, M., & Schulze, C. (2026). Frontiers: ChatGPT Referrals to E-Commerce Websites: How Do LLMs Compare Against Traditional Channels? Marketing Science, 45(4), 699–715. https://doi.org/10.1287/mksc.2025.0489
@article{kaiser_schulze_2026_ollm,
author = {Kaiser, Maximilian and Schulze, Christian},
title = {Frontiers: ChatGPT Referrals to E-Commerce Websites:
How Do LLMs Compare Against Traditional Channels?},
journal = {Marketing Science},
volume = {45}, number = {4}, pages = {699--715},
year = {2026},
doi = {10.1287/mksc.2025.0489}
}
09 · Updates
Get an email when we publish research updates
We email you when we publish new research or update our studies. No other mail.
Double opt-in. Unsubscribe in every email. Sent by Christian Schulze.


