named
Get a free AI audit

Blog

If ChatGPT sends almost no traffic, is AI visibility worth tracking?

Saad Khan, founder of Named (named.khanex.ai)

Drafted with AI help from the sources below and Named's live site. Saad checks every claim before it goes live. Named sells AI visibility work, so read this with that in mind.

Direct answer

Yes, if your buyers ask AI engines what to buy, but don't judge it by traffic. Marie Haynes found ChatGPT sent 0.62% of the sessions Google organic did across 100 sites. That counts clicks, and a buyer can read your name in an answer without clicking. Named counts the names. It keeps one fixed prompt set, repeats it on each engine, and logs names and cited links apart. If your buyers don't ask AI, skip it.

What did the October 2026 traffic studies find?

All three found that AI assistants send a small share of the traffic and conversions Google organic sends.

Study Sample Main finding What it can't see
Marie Haynes, Oct 21, 2026 12 months of GA4 data, 100 sites, over 145 million sessions ChatGPT sent 0.62% of Google organic's sessions, and 0.73% with all other LLMs added. Google organic drove 27x more conversions than all LLMs combined Buyers who read an answer and never click a link
The SEO Works GA4, 95 sites, Sep 2025 to Aug 2026 Organic sent 39x more traffic than AI assistants. For every 100 organic conversions, AI referrals delivered 3.2 Google AI Overviews clicks, which arrive as organic, and people who read an answer and then search the brand by name
Kevin Indig, Oct 21, 2026 12 GA4 properties, plus 632 tests across 12 sites; four brands had full visibility data Couldn't establish a reliable link between AI visibility and later branded search or direct traffic He says an effect could still exist. He couldn't isolate it

The conversion numbers disagree. In Haynes' data, ChatGPT visitors converted at 13.8% against 3.02% for Google organic, across the 55 sites that tracked conversions. The SEO Works found 7.18% against 6.08%, AI traffic converting worse on 23 of 63 sites, and B2B product sites converting at 0.62 times the organic rate.

Does low AI traffic mean AI answers don't matter?

No, because referral traffic only counts people who clicked a link in the answer. A buyer can ask ChatGPT for a recommendation, read your name, and type it into Google later. The SEO Works says its own method misses "the common pattern of reading an AI answer and then searching for the brand by name."

Mike King at iPullRank made the same point in his reply to the three studies. He says they grade "a brand channel with a traffic ruler" (iPullRank, Oct 22, 2026).

To be fair to the other side, Indig went looking for that hidden effect and couldn't find it reliably in four brands. Nobody has shown that being named leads to revenue weeks later.

King calls the tactics a distraction, meaning the schema hacks and the llms.txt files, and I agree with him. Haynes goes further and calls chasing ChatGPT mentions the wrong target. That's where I disagree. If someone asks ChatGPT for the best CRM for a small law firm and it names three companies, you want to know if you're one of them. You can check that today without waiting for GA4 to catch up.

What should you measure instead of AI referral traffic?

Count whether the engine names you, ask buyers how they found you, and track what those buyers spend. Keep each number separate. That's close to what Indig recommends: track mentions on a consistent set of prompts, ask new leads where they first heard of you, and connect those answers to revenue.

What to count How to count it What it tells you Where it breaks
Mentions per engine A fixed list of buyer questions, run more than once on each engine Whether you're in the answer Results vary run to run, so one run proves little
Self-reported source Ask "How did you hear about us?" on forms and calls, in the buyer's own words Whether AI is where the buyer met you People misremember
Referral sessions and conversions GA4, compared with organic The clicks you do get Misses AI Overviews clicks and people who search your name later
Revenue from AI-sourced buyers Tie the self-reported source to closed deals Whether the work pays for itself A few deals a month is a small sample

Start with the boring part. Haynes found 45 of her 100 sites weren't tracking any conversions. If you can't count a lead, you can't tell where it came from.

How Named counts this

Named keeps one fixed prompt set and repeats it to check changes. The Namies draft fixes, and a person checks every change. Results vary run to run, so Named repeats each question before calling a change, and says when the sample is too small.

What Named logs Why
A separate total for each engine Engines pick sources their own way, so one blended number hides the engine that skips you
Whether each answer names you That's share of answer, the share of your buyers' prompts where an engine names you
The links each reply cites Shows which pages the engine read
The companies it names instead Tells you who's getting the recommendation

We don't use referral traffic as the result, because an answer can name you without anyone clicking. The rules are on our facts page, and the split between names and links is in Mentioned vs cited.

When is AI visibility not worth the money?

When your buyers don't use AI to choose, or when your Google basics are still broken. The SEO Works found that growing organic search by 5% would add more conversions than doubling AI referral traffic. It also found the sites gaining AI traffic were the same sites whose organic search was improving.

So fix Google first if it's broken. Then ask your last ten customers how they found you. If none of them mention an AI engine, wait and ask again next quarter.

FAQ

Does AI traffic convert better than Google traffic?

It depends on the study and what you compare it with. Haynes found 13.8% for ChatGPT against 3.02% for Google organic. The SEO Works found 7.18% against 6.08%, and lower than organic for B2B product sites.

Do Google AI Overviews show up as AI traffic in GA4?

No. The SEO Works says those clicks arrive as ordinary organic sessions and can't be separated in GA4.

Is llms.txt worth adding?

Not for Google. Google's AI optimization guide says you can ignore AI text files like llms.txt for Google Search (Google, updated Jul 10, 2026). We ship one anyway, but we don't count on it.

Can anyone guarantee that AI will name my business?

No. Nobody controls what an engine says, and answers change from run to run. Named doesn't promise a ranking.

Which engines should I check?

Every AI engine you care about, starting with ChatGPT, Claude, Gemini, Perplexity, Grok and Muse. Count each one on its own. Or start with the free audit at named.khanex.ai.

Sources