AI Search · The Spark
AI referral traffic converts 23 times better than Google, and what it means for revenue teams
Ahrefs found visitors arriving from AI assistants convert far above Google search. Here is Sparked's operator read on what that means for B2B pipeline.
Yes, and by a wide margin. Ahrefs recently found that visitors arriving from an AI assistant such as ChatGPT or Perplexity convert into leads at close to 23 times the rate of a typical Google search visitor. For a B2B revenue team the takeaway is simple: answers now outrank rankings, and the channel deserves a line item in the forecast.
What Ahrefs actually found
In a short update this week, the Ahrefs team walked through their own site analytics and compared conversion rates by referral source. Visitors who clicked through from an AI assistant converted at roughly 23 times the rate of visitors who arrived from a standard Google search result.
Ahrefs, "AI traffic converts 23x better than Google" (YouTube, 9 July 2026): watch the clip.
That is our summary of the finding and our analysis of what it means below, not a repeat of their commentary.
Why an AI referral converts so much better
The gap is not magic. It is sequencing. Someone who asks an assistant a detailed question has already done the thinking a search engine used to leave for the landing page: what the options are, how they compare, whether the category fits their problem. By the time they click through to a named supplier, most of the qualifying work is done.
- The assistant already filtered out people who were only browsing.
- The visitor arrives with a specific question rather than a generic keyword.
- A named recommendation carries more trust than a ranked list of ten blue links.
What this changes for a revenue team
Most B2B sites still write for rankings: broad keywords, long introductions, a hedge before every claim. That structure reads badly to an assistant looking for a direct answer to quote. The fix is not a rewrite of the whole site. It is prioritising the pages most likely to be asked about, and making sure each one answers its core question in the first few lines.
Where teams get this wrong
The usual mistake is treating this as an SEO detail for the content team to handle later. Josh Harris, founder of The Sparked Group, put it plainly: "The businesses winning citations in ChatGPT and Gemini right now are the ones that decided this was a revenue question, not a content question, six months before their competitors did." Analytics matter too: if a site is not segmenting AI referral traffic separately from organic search, the team cannot see the effect Ahrefs is describing, let alone act on it.
The traffic is smaller today and worth more. Build for it accordingly.
Frequently asked
Questions buyers ask about this
What did Ahrefs find about AI referral traffic?
Ahrefs compared conversion rates by referral source on its own site and found that visitors arriving from an AI assistant such as ChatGPT or Perplexity converted into leads at roughly 23 times the rate of visitors from a standard Google search.
Why does traffic from an AI assistant convert better than search traffic?
An assistant does much of the qualifying work before the click. The visitor has already asked a specific question and compared options, so by the time they reach a named supplier's site, most of the filtering that used to happen on the landing page has already happened in the conversation.
Does this mean B2B teams should stop investing in Google search?
No. Google still delivers most of the visits for a typical B2B site in 2026. What changes is where a revenue team should look for its highest quality traffic, and how pages should be structured so an assistant can find and quote them.
What should a revenue team do differently this quarter?
Segment AI referral traffic separately in analytics, identify the handful of pages most likely to be asked about, and rewrite the opening lines of each so they answer the core question directly rather than building up to it.
Working on a real engine? Start with a conversation.
Tell us where you are. We will tell you what we see and where we would start.