Does Traffic From ChatGPT and Perplexity Convert Better Than Google Traffic

Traffic from ChatGPT and Perplexity converts better than Google traffic in the two single-site case studies published so far, and worse in the largest e-commerce study. If you lead marketing at an AI startup and need to justify budget for AI visibility, the honest position is that the early evidence favours AI referrals for considered software purchases, the samples are small, and your own data is the only benchmark that settles it. The studies disagree mainly because of what is being bought: software is researched inside the answer, while shoppers check prices elsewhere before paying.

Does traffic from ChatGPT and Perplexity convert better than Google traffic?

In the two single-site case studies, Ahrefs' own SaaS site and one Seer Interactive client, the published data says yes, often by a wide margin. For e-commerce, the one large multi-site study says no. Every study agrees on one thing: AI referrals are a small share of visits, so a high rate can rest on a few dozen conversions.

Swipe sideways to see every column

Study	Sample	What it found	Watch out for
Ahrefs, Patrick Stox (June 2025)	Ahrefs' own website, last 30 days	0.5% of traffic came from AI search and produced 12.1% of signups, a 23x higher conversion rate than traditional organic search; most of it came from ChatGPT	One well-known SaaS brand; the author expected the rate would not hold if volume scaled
Seer Interactive (June 2025)	One Seer client (industry not disclosed), October 2024 to April 2025	Conversion rates of 15.9% from ChatGPT, 10.5% from Perplexity, 5% from Claude and 3% from Gemini, against 1.76% from Google organic	Just under 11,000 AI sessions against almost 14 million organic sessions
Microsoft Clarity (November 2025)	1,277 domains from publisher and news sites	Sign-up rate of 1.66% from LLM referrals against 0.15% from search, and subscription rate of 1.34% against 0.55%	Publisher sign-ups are newsletters, not software trials; AI was under 1% of traffic
Semrush (June 2025)	Not disclosed	The average AI search visitor is 4.4 times as valuable as a traditional organic visitor, based on conversion rate	No method or underlying rates published
Kaiser and Schulze, University of Hamburg and Frankfurt School (November 2025)	973 e-commerce sites with $20 billion combined revenue, 12 months, 50,000+ ChatGPT transactions	ChatGPT's conversion rate and revenue per session were significantly lower than Google's paid and organic search; it beat only paid social	E-commerce purchases, where buyers often verify elsewhere before paying

Sources: Ahrefs, Seer Interactive, Microsoft Clarity, Semrush and the University of Hamburg's summary of the Kaiser and Schulze study. For an AI startup selling software, the first two are the closest match, and neither is a controlled experiment.

Why do AI referrals often convert at higher rates?

AI referrals tend to arrive later in the buying process. The comparison work, which a Google searcher does across several tabs, happens inside the answer, so the person who clicks through has often already shortlisted you. Semrush gives the same explanation for its 4.4x figure: the answer equips users with the information they need before they visit.

Clicks are also filtered harder. Patrick Stox's rough model at Ahrefs estimated that users of AI search click links about 75% less than users of traditional organic search, so the clicks that do happen tend to be deliberate. Google makes a similar point about its own AI features: Google Search Central says clicks from results pages with Google AI Overviews are higher quality, meaning users are more likely to spend more time on the site.

The landing page has shifted too. Similarweb reported that after ChatGPT made brand names clickable in May 2026, homepage visits rose from roughly 26% to 29% of ChatGPT referrals to about 62% to 63%. A visitor who clicks your name inside an answer behaves less like a searcher and more like someone who has already decided to look at you.

Why do some studies find AI traffic converts worse?

The strongest counter-evidence comes from e-commerce, and the reasons are partly about the purchase and partly about measurement. In Kaiser and Schulze's 973-site study, ChatGPT visitors showed favourable bounce rates, so the traffic was relevant, yet they bought less and spent less per session than visitors from Google's paid and organic search. Kaiser's own forecast was slow convergence, with no full parity with organic Google search expected within a year.

• Verification behaviour. A shopper who finds a product in an answer often checks the price or reviews elsewhere and buys on a later visit, which hands last-click credit to another channel.
• Small denominators. A few hundred sessions can swing a conversion rate by whole percentage points from one month to the next.
• Misclassified traffic. Loose GA4 rules can file ordinary .ai referral sites as AI assistants, which we explain in tracking AI search referral traffic in GA4.
• Dark traffic. AI-influenced visits that arrive as Direct or branded search raise those channels' rates and lower the AI channel's.

Which AI engines send enough traffic to matter?

ChatGPT is the largest source of AI referral visits in the data that breaks it down by engine. At Ahrefs it drove the majority of AI visits and signups, and for Seer's client it sent 61% of AI visits, with Perplexity at nearly 24% and Gemini at 15%. The total is small everywhere: under 1% in Microsoft Clarity's publisher data, and 0.07% of organic traffic for Seer's client.

The smaller engines can still convert well. In Microsoft Clarity's data, subscription conversion from Perplexity ran at 7 times the rate of search and direct traffic, Gemini at 4 times direct and 3 times search, and Copilot at 17 times direct and 15 times search. Usage is also spreading: Similarweb's figures show ChatGPT's share of generative AI web visits falling from about 76% to roughly 53% between June 2025 and May 2026. We have not seen reliable per-engine conversion data for SaaS beyond single-site case studies.

Google AI Overviews and Google AI Mode are missing from every comparison above for a structural reason. Their clicks are counted as Google organic traffic in GA4 and Search Console, so the Google organic baseline in these studies already contains some AI-assisted clicks. Search Console's Generative AI performance report now shows impressions in those features, but not clicks or conversions, so it cannot fill the gap.

How do you compare signup and demo rates by source in your own data?

Separate AI referrals in GA4, pick two outcomes, compare like with like for a full quarter, then follow the leads into your CRM. The aim is not a single headline multiple but a defensible answer to whether a visitor from an AI engine is worth more to you than one from Google or Bing.

Benchmarking AI referral conversion

Swipe sideways to see every column

Source	Sessions	Signups	Signup rate	Demo requests	Demo rate
ChatGPT	180	9	5.0%	3	1.7%
Perplexity	40	1	2.5%	0	0.0%
Google and Bing organic	12,000	240	2.0%	36	0.3%

What should you change if AI visitors convert better or worse than organic?

If AI visitors convert better on your site but send little traffic, invest in visibility on the prompts that already send visits, and fix the page they land on, which is now usually the homepage. If they convert worse, check whether engines describe your product accurately and whether the landing page answers what the prompt asked. In both cases, rates alone are not a reason to move paid budget; the trade-off is covered in when to move budget from paid ads to organic search at an AI SaaS.

Conversion quality is the reason AI visibility gets a budget line, and reach is the reason it stays small or grows. Track both: share of voice on a fixed prompt set, as explained in what share of voice in AI search is and how to calculate it, and referral outcomes in GA4. How the pieces fit into one plan is in building an AI search visibility strategy for a SaaS startup, and the measurement layer is a core part of AI visibility work.