Do GEO leads convert better than Google leads? What the data says

Ahrefs published a number in June 2025 that has turned up in every GEO pitch deck since. AI search made up 0.5% of its traffic but brought in 12.1% of its signups, roughly 23 times the conversion rate of its Google organic visitors. Ten months later, a peer-reviewed study of 973 online shops found ChatGPT visitors converted worse than almost every other channel. Both findings are true. They measure different things, and the gap between them tells you more about whether GEO leads convert better than Google leads than either headline does on its own.

This guide explains how AI search picks the businesses it recommends, why the visitors it sends tend to arrive ready to buy, where that advantage disappears, and how to measure it in your own accounts so you don’t have to take anyone’s word for it. Ours included.

The short answer: for considered purchases (services, B2B, finance, legal, anything people research before they enquire), AI search visitors usually convert at several times the rate of Google organic visitors. For quick, cheap ecommerce purchases the advantage is small or reversed. Volume is tiny either way, so GEO adds to Google rather than replacing it.

On this page

• What is GEO, in plain English?
• How does AI search work?
• Do AI search visitors really convert better?
• Why do GEO leads convert better?
• Where AI search leads don’t convert better
• Why your analytics undercounts GEO leads
• How to measure GEO leads properly
• How to earn AI citations that turn into leads
• FAQ

What is GEO, in plain English?

Generative engine optimisation (GEO) is the work of making your business easy for AI assistants such as ChatGPT, Gemini, Perplexity, Claude and Google’s AI Mode to find and recommend. SEO aims for a place in a list of ten blue links. GEO aims for a mention inside the single answer the assistant writes.

The term comes from a 2023 paper by researchers at Princeton and IIT Delhi, which found that adding citations, quotations and statistics to a page could lift its visibility in AI-generated answers by up to 40%.

Not everyone likes the label. Google’s guide to optimising for generative AI search says that, from its point of view, optimising for AI Overviews and AI Mode is still SEO. That’s fair for Google, because its AI features sit on top of its normal ranking systems. ChatGPT is a different animal. Semrush found that when ChatGPT search cites a page, that page ranks outside Google’s top 20 for related queries almost 90% of the time. Ranking on Google doesn’t guarantee you’re the business being recommended in a chat window. We’ve covered the overlap in SEO vs GEO: which one does your business actually need?

How does AI search work?

AI search turns a long, specific question into several smaller searches, reads the pages those searches return, and writes one answer that cites a handful of sources. The searcher sees a shortlist rather than a results page, and that changes who clicks and why.

Google hands you the Yellow Pages. ChatGPT is the mate down the pub who has already rung round four plumbers and tells you which two to call.

Everything above the last box happens inside the assistant, before your analytics sees anything.

The prompt is a brief, not a keyword

People don’t type keywords into ChatGPT. Semrush’s analysis of 17 months of clickstream data found that between 65% and 85% of ChatGPT prompts couldn’t be matched to any keyword in its database of over 27 billion terms. Its own example sums it up: the Google search is “best project management software”, while the ChatGPT version describes a 12-person remote engineering team that keeps missing sprint deadlines.

That second query carries the buyer’s situation and their problem. It’s already half a sales qualification call.

The engine fans the question out

Google’s documentation calls the next step “query fan-out”, where the model generates a set of related searches to gather more information. Its example turns “how to fix a lawn that’s full of weeds” into separate searches about herbicides, chemical-free removal and prevention. Both AI Overviews and AI Mode may use it. ChatGPT browses in a similar way when it needs fresh information, and Semrush found it switched on web search for 34.5% of queries in February 2026.

It reads, then writes, then cites

Google calls the reading step retrieval-augmented generation, or grounding. Its systems pull relevant pages from the search index, check the information on them, and generate a response with links to the pages that back it up. By the time the user reads the answer, the comparison work is done, and only a few sources get a link.

Most people never click

Fewer links means fewer clicks. Pew Research found that Google users who saw an AI summary clicked a traditional result in 8% of visits, against 15% for those who didn’t see one, and almost never clicked a link inside the summary itself. AI Overviews now appear on more than 80% of Google searches, according to recent tracking by Semrush, SE Ranking and seoClarity, up from about half in mid-2025. The people who do click are a filtered group. That filter is the whole story of this article.

Do AI search visitors really convert better?

In most published studies, yes, and by a wide margin when the conversion is a sign-up or an enquiry. How wide depends on what’s being sold and how it’s measured.

The longer the bar, the bigger the AI advantage. Only one study comes in under 1×. Here’s what sits behind each one.

Ahrefs (one software site). AI search sent 0.5% of visits and 12.1% of signups over 30 days, a 23x higher conversion rate than traditional organic search. Patrick Stox, who ran the analysis, said he doubted the advantage would hold at that size if AI search became the default.

Microsoft Clarity (1,200+ publisher sites). Visitors from AI assistants signed up at 1.66%, compared with 0.15% from search, about 11 times higher, even though AI referrals were still under 1% of traffic.

Marie Haynes (100 sites, 55 tracking conversions). In an analysis reported by MarTech in October 2026, ChatGPT visitors converted at 13.8% against 3.02% for Google organic, roughly 4.6 times higher.

Semrush (cross-industry). The average AI search visitor was worth 4.4 times the average traditional organic visitor, based on conversion rate.

Visibility Labs (94 ecommerce brands, 2025). As reported by Search Engine Land, ChatGPT traffic to product and category pages converted at 1.81% against 1.39% for non-branded organic, 31% higher, and beat organic in 10 of 12 months.

Kaiser and Schulze (973 ecommerce sites). The one study that points the other way. Across all sites, Google organic converted 13% better than ChatGPT. More on that below, because the detail flips the conclusion for lead-gen businesses.

Two more data points are worth knowing. Adobe, which tracks over a trillion visits to US retail sites, reported that AI traffic converted 42% better than non-AI traffic in March 2026, having converted 38% worse a year earlier. And a single-client agency case study cited in the Marketing Science paper put ChatGPT at 15.9% against 1.8% for organic search.

The pattern is hard to miss. The biggest multiples come from sign-ups and enquiries. The smallest come from retail checkouts.

Why do GEO leads convert better?

AI search visitors convert better because most of the buying journey happens inside the chat, before they reach your site. By the time they click, they’ve described their problem and been handed your name as a recommendation.

The research happens before the click

Semrush puts it plainly: by the time an AI search user visits your site, they have likely already compared their options and perhaps even learned about your value proposition. Visibility Labs calls the same effect intent compression. The narrowing down that used to take five Google searches and a dozen open tabs gets squeezed into one conversation.

A Google visitor who lands on your services page might be on tab three of eight. An AI visitor has usually closed the other seven.

The prompt does the qualifying for you

Picture two people looking for an accountant. One types “accountant Leeds” into Google. The other asks ChatGPT: “We’re a 25-person construction firm in Leeds turning over about £4m. Our accountant is slow on VAT returns and we want someone who knows CIS. Who should we talk to?”

If you’re recommended in answer to the second prompt, the person who clicks already knows you suit their size and sector. The Google searcher could just as easily be a sole trader or a student. The prompt has filtered out most of the people who were never going to buy, which is why the conversion rate looks so different even when the click count is small.

A recommendation borrows trust

Semrush notes that AI responses are presented like personal, word-of-mouth recommendations, which may give them more persuasive power than a list of links. Trust in the tools is rising too. In Adobe’s survey of more than 5,000 US consumers, 66% said they believe AI tools provide accurate results.

Being named in an answer feels closer to a referral than an advert, and people behave accordingly.

Fewer clicks, better clicks

Because AI answers settle most questions on the spot, the casual browsers drop out. Ahrefs estimated that users from AI search click links 75% less than they do in traditional organic search. Google says something similar about its own AI features: clicks from results pages with AI Overviews are higher quality, with visitors more likely to spend more time on the site.

They land on the page that answers their question

Microsoft’s Clarity team found that AI-referred visitors bypass the homepage, land deep inside the site and convert at a much higher rate. At Ahrefs, AI visitors also viewed 50% more pages than visitors from search.

Where AI search leads don’t convert better

AI search leads don’t reliably convert better for simple, low-cost products, and they never make up for the volume Google sends. Anyone who only quotes the 23x figure is telling you half the story.

The 973-shop study

The most rigorous study so far is Maximilian Kaiser and Christian Schulze’s paper in Marketing Science. Using 12 months of Google Analytics data from 973 ecommerce sites with $20bn in combined revenue, they found ChatGPT referrals converted below every traditional channel except paid social.

Read on, though, and the picture changes. The authors split sites by how complex their products are. In high-complexity categories, which include business and consumer services, finance, law, vehicles, heavy industry and jobs, ChatGPT’s conversion rate beat paid social, referral, email, organic search and direct. Those sites also got 4.6 times more of their traffic from ChatGPT.

The authors’ explanation is that AI assistants are most useful when a purchase needs research and deliberation, and add little to routine buys. For lead-gen businesses, that’s the most important finding in this article: the more thought a purchase needs, the better AI-referred visitors convert.

Smaller baskets

Where AI visitors do buy online, they tend to spend less. Visibility Labs found average order value was $204 for ChatGPT visitors against $238 for organic, a 14.3% gap, and the Marketing Science paper found AI order values falling over time. Revenue per visit still came out 10% higher in the Visibility Labs data, a much thinner margin than the conversion rate suggests.

The volume problem

This is the bit the pitch decks skip. In Marie Haynes’ 100-site analysis, ChatGPT sent just 0.62% as many sessions as Google organic, and Google organic still drove 27 times as many conversions as every AI assistant combined. Haynes called GEO a “distraction”. The 973-shop study put ChatGPT at under 0.2% of all traffic.

Run the numbers on a typical site. Say you get 10,000 Google organic visits a month and convert 3% of them, which gives you 300 enquiries. At the 0.73% ratio Haynes found for all AI assistants combined, you’d get about 73 AI visits. Even at a 13.8% conversion rate, that’s around 10 enquiries. Useful, probably high quality, but not a replacement for anything.

So why bother? Because AI answers increasingly shape which brands people search for next, and most of that influence never gets credited to an AI assistant.

Why your analytics undercounts GEO leads

Standard analytics only credits AI search when someone clicks a link in the chat and lands on your site in one go. Plenty of AI-influenced leads take a different route and get filed under branded search or direct.

People check on Google first. Semrush found that over 21% of ChatGPT’s referral traffic goes to Google, and suggests users may be confirming what ChatGPT told them. If someone hears your name in ChatGPT, googles you and fills in a form, GA4 records a branded organic lead.

The ecommerce study saw the same leak. Visibility Labs flagged that shoppers often get a recommendation from ChatGPT, then search for the brand on Google before buying, so the sale is credited to branded search. Its advice was to add post-purchase surveys.

Last-click attribution can’t see it. Kaiser and Schulze point out that last-click data understates any channel that mainly helps with discovery, which is exactly what AI search does.

Tracking breaks. For about a week in mid-October 2026, ChatGPT links lost the tracking tag OpenAI adds to them (utm_source=chatgpt.com). Adobe’s Cedric Huesler looked at around 500 pages where ChatGPT drove at least 5% of visits and found ChatGPT visits down about 31,000 a day, while direct visits to the same pages rose by about 26,000. He later confirmed it was a bug and the tags came back. Anyone who reported a ChatGPT traffic collapse in October was looking at an attribution fault, not lost visitors.

How to measure GEO leads properly

You measure GEO leads by tracking AI referrals as their own channel, asking every lead where they heard about you, and judging the channel on closed revenue rather than sessions. Here’s the setup we’d recommend.

1. Use GA4’s AI Assistant channel. Since May 2026, Google Analytics has grouped visits from recognised assistants such as ChatGPT, Gemini and Claude into an AI Assistant channel in the default channel group. Google’s help centre also shows how to build a custom channel group for AI assistants with a referrer regex, which is worth running alongside it to catch tools Google doesn’t recognise yet.
2. Ask every lead. Add a “How did you hear about us?” field to enquiry forms with “ChatGPT or another AI assistant” as an option, and ask the same question on discovery calls. It’s the only reliable way to catch people who researched in ChatGPT and came back through Google.
3. Follow leads into the CRM. Compare AI-sourced and Google-sourced leads on qualification rate, close rate, deal value and time to close. A website conversion is only the first step.
4. Watch branded search and direct. If your AI visibility improves, branded searches and direct visits to the pages AI tools cite should rise as well, even if the AI channel itself stays small.
5. Check Search Console. Google’s Generative AI performance report shows how your content is performing in AI Overviews and AI Mode.
6. Give it 90 days. AI samples are small. Visibility Labs had months with only 15 to 37 ChatGPT conversions across 94 brands. One good week proves nothing.

How to earn AI citations that turn into leads

You earn citations that turn into leads by publishing specific, evidenced answers to the questions your buyers ask, making sure AI crawlers can read them, and getting talked about on the sites AI tools trust. Then you make the landing page easy for a pre-sold visitor to act on.

Write for the brief, not the keyword. Most prompts don’t match keywords, so build pages around real buying situations: “accountant for construction firms using CIS”, “Shopify agency for brands moving off Magento”, “what a Google Ads audit should cost”. Semrush’s research suggests AI systems are more likely to cite content tailored to specific use cases and audiences.

Put proof on the page. Statistics and quotes from named sources are what the original GEO research found boosted visibility in AI answers. Prices, timelines, case study numbers and named experts give an assistant something concrete to repeat.

Make it readable by machines. Adobe’s AI visibility checker found that product pages on US retail sites scored just 66%, meaning a third of their content couldn’t be read by AI models. Keep key information in plain text that crawlers can reach. You don’t need an llms.txt file for Google, which says its search ignores them.

Get mentioned in the right places. AI tools lean heavily on what others say about you. Semrush found Quora and Reddit were the most-cited sites in Google AI Overviews, and half the links in ChatGPT answers pointed to business and service websites. Genuine reviews and expert commentary in trade press help. Google warns that chasing inauthentic mentions doesn’t.

Keep your SEO strong. Google’s AI features run on its core ranking systems, and a fifth of ChatGPT’s referrals end up on Google anyway. GEO without SEO is building on sand.

Fix the landing page for someone who’s already decided. AI visitors arrive deep in the site with their mind half made up. Put pricing guidance and a short enquiry form on the pages they land on, not three clicks away.

For the step-by-step version, see our guides on how to rank on ChatGPT and AI SEO and GEO.

FAQ

Do ChatGPT leads convert better than Google leads?

Usually, for services and other considered purchases. Published studies put the gap anywhere from about 1.3 times higher for ecommerce to 23 times higher for one software company’s signups. A large study of online shops found Google organic converted 13% better overall, but ChatGPT came out ahead for complex products.

How much traffic does AI search send compared with Google?

Very little so far. One 100-site analysis found ChatGPT sent 0.62% as many sessions as Google organic, and Google still drove 27 times more conversions. Treat AI search as a small, high-intent channel that sits alongside Google rather than a replacement for it.

Is GEO different from SEO?

Partly. Google says optimising for its AI Overviews and AI Mode is still SEO, because they run on its normal ranking systems. ChatGPT and other assistants often cite pages that rank poorly on Google, so being recommended there takes specific, evidenced content and mentions on the sites those tools trust.

How do I track AI search traffic in GA4?

Open the Traffic acquisition report and look for the AI Assistant channel, which Google added to the default channel group in May 2026. Add a custom channel group with a referrer regex to catch newer tools, and put a “How did you hear about us?” question on your forms, because many AI-influenced leads arrive through branded search.

Which businesses benefit most from GEO?

Businesses selling considered, higher-value products or services, such as B2B and professional services, finance, legal, trades, healthcare and complex ecommerce. Research on 973 online shops found AI referrals performed best where purchases needed comparison and deliberation, and worst for simple, routine buys.

The pub mate test

Back to the mate down the pub. He doesn’t send you many plumbers, but the ones he sends usually get the job, because he’s done the vetting. GEO leads look like that today. There aren’t many of them, but they arrive unusually ready to buy, and standard analytics misses a good share of them. The data doesn’t support abandoning Google for AI search. It does support making sure you’re the name that comes up when a buyer with a real problem asks.

If you want to know how often AI assistants recommend your business, and what those leads are worth, our GEO team can audit your visibility across ChatGPT, Gemini, Perplexity and Google’s AI features. Book a free consultation or see how we approach lead generation.

Sources

• Does AI search traffic convert better than traditional search? Ahrefs, June 2025
• ChatGPT referrals to e-commerce websites: how do LLMs compare against traditional channels? Kaiser and Schulze, Marketing Science, 2026
• AI traffic converts at 3x the rate of other channels, Microsoft Clarity, 2025
• ChatGPT sends less than 1% of Google’s traffic, MarTech, October 2026
• We studied the impact of AI search on SEO traffic, Semrush, July 2025
• ChatGPT traffic analysis: insights from 17 months of clickstream data, Semrush, April 2026
• ChatGPT ecommerce traffic converts 31% higher than non-branded organic search, Search Engine Land, February 2026
• U.S. retailers see surge in AI traffic, Adobe, April 2026
• Optimizing your website for generative AI features on Google Search, Google Search Central
• AI features and your website, Google Search Central
• Google users are less likely to click on links when an AI summary appears, Pew Research Center, July 2025
• What’s new in Google Analytics: AI Assistant traffic measurement, Google, May 2026
• GEO: Generative Engine Optimization, Aggarwal et al., arXiv