TL;DR
We asked 171 marketers, founders and SEO leads how they’re tracking AI traffic. We expected some frustration. We didn’t expect the numbers to be this stark.
54% of marketers can’t find their own AI traffic
More than half of the people we surveyed described the same problem, without any prompting: they know AI is sending them visitors. They just can’t find them in the data.
“A huge chunk of AI-referred traffic just shows up as ‘direct’ or gets lumped into regular organic. I’d estimate we’re probably missing 30-40% of actual AI-driven visits because of how broken the attribution is right now.”
That’s not a rounding error. That’s a third or more of an entire channel’s activity, sitting in the wrong bucket, on autopilot, indefinitely – unless someone goes looking for it.
44% trust a signal that isn’t even in their AI report
When we asked marketers which signal they trust most for gauging AI’s impact, branded search volume beat everything else – 44% named it as their most reliable indicator. Not AI referral traffic. Not GA4. Branded search in Search Console.
Think about what that means. The metric marketers trust most to understand AI’s impact on their business doesn’t mention AI anywhere in its name.
Here’s why: people rarely click the citation inside an AI answer. They open a new tab and Google the brand name instead, out of habit or caution. GA4 logs that as organic traffic. The AI recommendation that started the whole journey gets credited to nothing.
“When AI tools recommend a business by name, users often follow up with a direct Google search. Watching that branded query trend tells me more about AI-driven awareness than any referral report does.”
One agency saw a retail client’s branded queries jump 31% in two months, with no extra ad spend behind it – purely from showing up more consistently in AI answers. That’s real commercial movement, and it would be completely invisible to anyone only watching a referrals report.
1 in 3 say AI traffic converts better than everything else
This is the stat that should change how you think about AI traffic entirely.
A third of the marketers we surveyed said AI-referred visitors convert at a higher rate than their other channels. One SEO consultant put a number on it directly:
“AI accounts for just 1.46% of my traffic, yet it’s my highest-quality converting channel.”
Low volume and low value would be safe to ignore. Low volume and disproportionately high value is a different problem entirely – it means the small slice of visitors most likely to become customers is exactly the slice your reporting can’t see.
There is no “normal” AI traffic number – and that’s the point
We asked people to estimate what share of their traffic came from AI. The answers were all over the place, and not randomly – they tracked cleanly by industry.
Estimated AI traffic by industry
Travel and hospitality sites sat at 3-5%. B2B and SaaS around 2.8%. Retail and e-commerce closer to 1%. Publishing and media as low as 0.5%.
“Anyone quoting you a precise percentage is guessing with confidence.”
That line from one of our respondents is worth sitting with. Every industry benchmark report you’ve seen – Ahrefs at 0.17%, SE Ranking at 0.32%, Search Engine Land at 1.08% – is measuring a different sample, with different methodology, and all of them can only see the visible fraction of AI traffic that survives with a referrer intact. None of them are lying. None of them are the whole picture either.
If a vendor or a competitor quotes you a single confident percentage for “average AI traffic,” that’s your cue to be sceptical, not impressed.
The real number might be hiding in your Direct traffic
Research shows 70.6% of AI traffic arrives with no trackable referrer at all, landing straight in your Direct bucket alongside genuine direct visits. Run your own Direct traffic and average deal value through that ratio, alongside a typical AI-visitor conversion rate, and most mid-sized businesses land somewhere in six or seven figures of unattributed monthly revenue.
It’s a rough estimate, not gospel. But even at half that rate, it’s not a number worth ignoring.
Why this keeps getting worse, not better
Two things are compounding the problem.
AI usage is growing fast
First, AI usage itself is growing fast, and users increasingly juggle several platforms – most people who use AI tools regularly are using an average of 3 different ones. Every one of those platforms strips referrer data slightly differently, which means the attribution gap doesn’t stay still long enough for a single fix to hold.
Attribution can be a mess
Second, if those branded verification searches happen through paid search, some businesses are now paying to recapture a customer that AI already delivered to them for free.
Worse, if a competitor is bidding on your brand name – and they very likely are – AI sends someone your way, and a rival intercepts them at the exact final moment you should have owned.
What the marketers who’ve solved this actually do
The respondents who weren’t frustrated had one thing in common: they’d stopped relying on any single metric.
“We stopped relying only on referral source attribution and started correlating spikes in branded queries, assisted conversions, and sudden engagement increases on pages frequently cited in AI summaries.”
That’s the pattern worth copying. AI referrals, branded search trend and homepage-excluded direct traffic, read together rather than in isolation, is what turns “we think AI is working” into “we can see AI is working.”
Crunch the numbers and see for yourself
Looking to learn more about AI traffic and how to track it properly? Check out our latest guides:
➡️ What is dark traffic?
➡️ How to track AI traffic easily
Download our AI traffic dashboard
Get more reliable insights into your AI data.
Download here
