Key findings
AI platforms referred an average of 46.8 million visits a month to marketplaces, more than any other industry. Similarweb.
Monthly referrals rose from 33.2 million to 89.9 million in a year. Similarweb.
For the largest platforms it is still under 1% of bookings. Booking Holdings, via PhocusWire.
How much traffic do AI assistants send to marketplaces?
More than to any other industry. Similarweb measured referral visits from AI platforms worldwide between June 2025 and May 2026. Marketplaces led at an average of 46.8 million visits a month. The monthly figure climbed from 33.2 million in June 2025 to 89.9 million in May 2026. Across all industries, AI platforms sent an average of 770.7 million referral visits a month.
It is still a small share of the total. Booking Holdings told investors in August 2026 that referrals from large language models were less than 1% of its room nights, as reported by PhocusWire. Etsy said on its February 2026 earnings call that agentic traffic in the fourth quarter of 2025 was about 15 times the year before, from a small base, and that orders from ChatGPT tended to be higher in value than those from more mature channels.
Why does the answer name one platform and not another?
Because the question is usually where, not what. People ask where to book a cabin, where to hire a designer, where it is safe to buy a refurbished phone. The answer names one or two platforms and gives a reason for each.
That reason comes from pages about categories and policies, not the homepage. An engine that is asked about vetted freelance designers looks for a page that says, plainly, how designers are vetted, what they cost and what happens if the work is poor. A marketplace is named for the categories it describes clearly.
Which sources do the engines read?
Category and city pages. What the platform is good for, in a form that can be quoted.
Listing data. Titles, attributes, prices and availability, written by thousands of sellers and read as the platform’s own facts.
Help centre and policy pages. Fees, guarantees and dispute terms, which decide trust questions.
Review sites and forums. Where customers and sellers compare platforms directly.
Partner integrations. Yelp, for example, announced in August 2026 that its reviews, ratings and business details are surfaced directly inside ChatGPT.
Booking Holdings said in November 2026 that its organic referrals increased after new ChatGPT models improved citation rates, again as reported by PhocusWire. When engines cite more, the platforms with clear, citable pages gain first.
What to do first, and sources
Write the where-should-I questions for your top categories and cities and run them by engine. Note which platforms are named and which of your pages, if any, are cited. Fix category and policy pages first, then the listing templates and seller guidance that shape every listing. How SourceNode does this for marketplaces and platforms.
Yelp news release, 10 August 2026, as quoted by PYMNTS: Yelp content surfaced inside ChatGPT.
The researcher
Odin Lund is co-founder and COO of SourceNode and leads its data analytics and research, including the daily tracking behind this work across ChatGPT, Gemini and Google AI Overviews. Odin founded First Query, the AI search practice that became SourceNode, and co-founded The Pixel Paper.
Meet the research team
About this research
Published by SourceNode
SourceNode is a data and research-first AI search firm. We record what AI models say about a business every day, trace each claim to its source, and use that data to get businesses named in AI answers. Every figure we publish carries its method and its sample.
Share Blog
Related research
AI search
AI search statistics for 2026: the numbers we trust, with sources
Read the research
AI search
What McKinsey’s AI search research says, and what to do about it
Read the research
