If ChatGPT barely shows up in analytics, is it fair to pay more for AI visibility? AI search traffic is real and measurable, and it is small. How small depends on whose data you read, what you divide by, and whether the site tracks conversions at all. Three analyses published in October 2026 also reveal that AI assistants send a sliver of what Google sends. It’s still too soon to judge whether you should skip AI visibility altogether. 

In this guide, we’ll show you how to read the latest studies on AI search traffic and learn what they measured, why clicks understate visibility, and how to track the AI search traffic your own site gets.

What the Studies Say About AI Search Traffic 

Barry Schwartz rounded up three analyses published in October 2026 that AI assistants send a small fraction of the traffic Google organic search sends in a Search Engine Roundtable roundup. Each compares AI assistant referrals directly with Google organic search, and each found little measurable payoff from AI visibility in analytics. Let’s explore both the headline numbers from the studies and their sample limitations:

 

Study Sample Reported finding
Marie Haynes, “GEO is a distraction” (21 October) 100 sites, 12 months of GA4 data, more than 145 million sessions ChatGPT sessions equaled 0.62% of Google organic volume, and 0.73% with all other LLMs added. Among the 55 sites tracking conversions, ChatGPT visitors converted at 13.8% against 3.02% for Google organic, yet Google drove 96.5% of conversions.
The SEO Works, AI referral traffic benchmark 95 websites Organic search delivered 39 times more traffic than AI assistants. For every 100 conversions from organic, AI referrals delivered 3.2.
Kevin Indig, Growth Memo 12 sites of different sectors and sizes AI-driven direct traffic to long-tail pages was not detectable, and he found no reliable link between AI visibility and branded search or Direct traffic.

What we learn: These three don’t measure the same thing, and that’s the first lesson. Haynes and The SEO Works count sessions and conversions. Indig looked for AI influence hiding inside direct and branded search and couldn’t find it. 

ChatGPT Referral Traffic: Why the Volume Looks So Small

ChatGPT referral traffic looks small because it only counts visits where someone clicked a link, and your analytics tool could see where they came from. That’s just a thin slice of how ChatGPT exposes your brand.

Start with the denominator. Haynes compared ChatGPT sessions with Google organic sessions. She did not report ChatGPT’s share of all site traffic, which is a different number, so don’t paste her 0.62% into a sentence that says “of total traffic.” 

As far as we can tell, the post also leaves out exact reporting dates, the industry mix, how a conversion was defined, and any change in Google organic traffic over the period.

Then the trend. Haynes reported that the exponential rise in ChatGPT sessions across her sample ended around August 2025, and that ChatGPT sessions stopped accelerating around August 2025, around the GPT-5 release, and have declined since. Other data points the other way: The SEO Works saw AI’s share of organic sessions edge up from about 2.3% to 3.0% after May 2026, though it calls that four-month window provisional. Neither settles your case, so chart your own sessions before drawing conclusions.

And then the missing surface. Per Practical Ecommerce’s walkthrough of GA4’s AI Assistant channel, notes that clicks from Google’s AI Overviews and AI Mode are reported as ordinary Organic Search.

So when a study sets AI assistants against Google organic, part of Google’s own AI exposure sits inside the organic number. 

Pew’s browsing data shows why that matters. Across 900 US adults and 68,879 Google searches in March 2025, people clicked a traditional result in 8% of visits when an AI summary appeared and 15% when it didn’t. They clicked a link inside the summary in 1% of visits. People see these answers. They rarely click out of them.

Also read: How to do a content refresh for AI 

AI Traffic vs Organic Traffic: The Volume Gap and What Moves It

Organic search sent 39 times more visits than AI assistants in The SEO Works’ study, and about 137 times more in Haynes’ study. (100 ÷ 0.73) Both looked at roughly 100 sites, yet the results differ by more than three times. That’s because every site is different. A study’s ratio describes its own sites, not yours, so check your own numbers.

Watch out for one more thing. A ratio changes if either side changes. If Google traffic falls while AI traffic stays flat, AI looks like it’s catching up when it hasn’t grown at all. Track both in absolute sessions, not just as a ratio.

AI Search Conversion Rate: What the Numbers Actually Mean 

An AI search conversion rate claim (“AI converts 4x better”) means nothing until it names its baseline, its sample, and its definition of a conversion. The published numbers swing wildly because those three things differ.

Source Sample Reported comparison
Ahrefs, June 2025 Its own site AI search was 0.5% of traffic and 12.1% of signups, 23x the conversion rate of organic search visitors
Marie Haynes, October 2026 55 sites tracking conversions ChatGPT 13.8% against Google organic 3.02%, more than 4.5x
Siege Media 78 sites Median AI to organic ratio of 1.26x, with AI matching or beating organic on 72% of sites
The SEO Works, October 2026 95 site benchmark 0.67x against paid search and 2.48x against direct traffic

Read the spread. Ahrefs tested itself, and Patrick Stox said in that post that it might be an outlier for Ahrefs. Haynes’ conversion comparison is based on 55 sites because 45 of her 100 are not tracking conversions. And The SEO Works compares AI to paid search and direct in the same dataset. Anyone who gives you a multiple without referencing the base is giving you an unverifiable number.

Now combine a volume figure with a conversion rate. Here’s an illustration using Haynes’ numbers

For every 10,000 Google organic sessions, ChatGPT sends about 62 (0.62%).

 At 13.8%, it’s about 10 conversions. The 10,000 organic sessions, at 3.02%, produce about 302. ChatGPT traffic converts about 4.5 times better and still accounts for only about 3% of conversions in this example.

Also read: The modern topic cluster strategy for ChatGPT, Gemini and AI Search 

AI Citations vs Clicks: What the Click Count Leaves Out

Clicks measure visits generated by an AI answer. Citations and mentions measure how many times an AI answer refers to you, and that difference can be significant.

Per BrightEdge’s analysis, ChatGPT references brands 3.2 times more frequently than it cites them: 2.4 mentions per prompt vs. 0.74 citations. Google’s AI Overviews flipped that in BrightEdge’s data, citing brands more often than they named them.

When ChatGPT recommends a brand without a link, there is no click to count and no referral for GA4 to record.

Will an unlinked mention ever convert down the line, in a branded search or a direct visit? The plausible mechanism is straightforward. As Indig explained in November 2025, unlinked mentions build familiarity, so teams should optimize for mentions, not citations, and stop trying to make referral traffic pay off. 

AI Traffic Tracking: How to Measure Your Own Site

The fastest way to start AI traffic tracking is the AI Assistant channel Google added to GA4’s default channel group.

  1. Open the channel. Go to Reports, then Acquisition, then Traffic acquisition, and look for AI Assistant under the session default channel group.
  2. Know its limits. It’s referrer-based. AI sessions without a recognized referrer fall into Direct or Unassigned, and AI Overviews and AI Mode clicks stay in organic search.
  3. Add a custom channel group. Use a regex such as chatgpt|openai|gemini|claude|copilot|perplexity|grok on session source, and run it beside the native channel for a quarter.
  4. Compare against organic and paid. Write the baseline next to every multiple, every time.
  5. Confirm conversions are tracked. In Haynes’ sample, 45 of 100 sites weren’t tracking any.
  6. Ask people. Add “how did you hear about us” to forms, include AI tools as an option, and tie answers to revenue.
  7. Be careful with Direct. Direct growth can come from bots and typed visits as easily as from AI. Split it by landing page and browser version, and compare total sessions with engaged sessions before you credit AI.

On the evidence above, AI search traffic is small today and worth measuring. It isn’t the whole story, though: AI answers shape which brands people name and search for, and clicks capture only part of that.

Links and earned coverage factor into how pages rank, and ranking helps pages get retrieved by AI assistants. If any provider, us included, promises a citation count, ask for the method.

You Know How It Works. Let Us Run It.

Book a call with a strategist or check out our services to find the best fit for you.

Written by Ahmed Ehtesham

Strategists and outreach leads who've shipped 4,000+ links and recovered dozens of penalized domains.

Let's Talk About What's Holding
Your Rankings Back

Most brands we audit find they're missing 20-30% of potential traffic, usually due to
indexation gaps or thin internal linking. Get your audit and see what you're missing.

Scroll to Top