ChatGPT Referrals Convert at 3x the Rate of Google Search

There's a number circulating in eCommerce circles that's making a lot of marketers rethink their entire approach to digital spending. AI tools like ChatGPT are sending customers to online stores who convert at 3.6%. That's almost three times the 1.23% conversion rate of traditional Google search traffic.

The data comes from Lebesgue, a marketing analytics platform used by over 10,000 businesses globally. The company analyzed 35,000 brands on Shopify and WooCommerce. The AI referral numbers matched, almost exactly, the conversion rate of customers who searched for a brand by name. Those are the warmest possible leads.

The Funnel Has Flipped

Here's what makes this different from any other traffic source. Traditionally, a customer searches Google, lands on your site, and you spend the next few seconds convincing them to buy. With AI tools, that convincing already happened inside the chatbot conversation.

Josip Begic, Co-Founder and CEO of Lebesgue, puts it plainly: "With AI, the persuasion happens before the click." By the time a customer taps through from ChatGPT's recommendation, they've already been told your product is a good fit. They arrive with a decision mostly made.

This also explains why AI-referred visits generated 30% higher revenue per session than traditional search traffic. It's not just more conversions. Each converted customer is spending more.

The Research Is Not Unanimous

It would be easy to overstate this. The picture is genuinely mixed.

A separate academic study published on SSRN examined 973 eCommerce sites generating a combined US$20 billion in annual revenue. That research found ChatGPT referrals actually converted worse than both affiliate links (by 86%) and regular organic search (by 13%). A third study from Visibility Labs, which looked at 94 eCommerce sites, landed in the middle. ChatGPT traffic converted 31% better than non-branded Google search, but not nearly at the 3x level Lebesgue reported.

The gap between these findings comes down to methodology. Lebesgue's sample skews toward smaller Shopify merchants. The SSRN study covers a much broader range of business types and sizes. Neither is wrong. They're measuring different populations. What they agree on is that AI referral volumes are growing fast, even if the conversion story depends on your category and business model.

What This Means for APAC Brands

Asia Pacific drives more than half of global eCommerce. The region's consumers are heavy AI users, particularly across mobile-first markets in Southeast Asia, China, and South Korea. But most APAC brands are still not thinking about AI referral optimization as a priority.

Regional players like Zalora and JD.com have invested heavily in on-site AI personalization. The next step, how to show up well inside ChatGPT, Perplexity, and Gemini responses, is still largely unaddressed territory.

Generative AI referrals to eCommerce sites surged 4,700% in 2025. Volume is still small relative to Google, but the trajectory is steep. Organic search is simultaneously being squeezed. Google's AI-generated answer boxes have reduced the percentage of people who click through to websites by 61% across major categories.

Begic frames the implication for brands clearly: "The brands that attract the highest-value, pre-qualified leads will be those most visible across all the places AI looks, such as reviews, third-party content and forums."

That's a different kind of marketing investment than bidding on keywords. It's reputation management. It's getting reviewed, quoted, and referenced in the places AI systems read. For Asian marketing leaders still allocating most of their budgets to paid search and SEO, it may be time to run a quick audit of how well your brand shows up when someone asks ChatGPT for a recommendation in your category.