For years, marketers have treated “direct traffic” as a sign of strength.
It meant brand awareness. Loyalty. Intent.
If someone typed your URL directly into their browser, that was a win. But that assumption is quietly breaking down — and most teams haven’t realized it yet.
What looks like direct traffic is increasingly something else entirely: a growing pool of influence that analytics platforms can’t properly trace. And as artificial intelligence becomes a primary layer in how buyers research, that blind spot is expanding fast.
Welcome to the era of dark traffic.
The Journey You Can’t See
A modern buyer journey rarely starts where your analytics say it does.
Instead, it begins in places like ChatGPT, Perplexity AI, or embedded copilots inside tools buyers already use. Questions are asked, options are evaluated, and shortlists are formed — all before a single click happens.
By the time a user lands on your site, the decision-making process is already well underway.
But here’s the problem: analytics platforms only capture the click, not the context.
So that visit? It gets labeled as “direct.”
Not because it was direct — but because everything that influenced it happened somewhere your tracking tools can’t see.
When Attribution Stops Reflecting Reality
Traditional attribution models were built for a different internet.
They assume that influence happens across trackable touchpoints: search queries, ad clicks, referral links. The journey may be complex, but it’s ultimately visible.
AI breaks that assumption.
When a buyer asks an AI assistant for recommendations, the response is generated from a mix of sources, signals, and synthesized knowledge. There’s no referral URL. No campaign parameter. No breadcrumb trail.
The influence is real — but it’s invisible.
According to Pew Research Center, users are significantly less likely to click on traditional search links when AI-generated summaries appear in results — further compressing the visible portion of the buyer journey. The implication is clear: more decisions are being shaped without producing measurable interactions.
Attribution, as most teams understand it, isn’t just incomplete. It’s increasingly disconnected from how decisions are actually made.
The B2B Blind Spot
This shift is particularly pronounced in B2B.
Buyers are no longer arriving as blank slates. They come in informed, pre-filtered, and often with a shortlist already in mind. The evaluation process — once distributed across search, content, and conversations — is now being condensed into AI-driven interactions that companies never see.
That creates a dangerous illusion.
Marketing teams believe they’re influencing demand because traffic is steady. Sales teams believe they’re driving conversion because prospects seem qualified. Meanwhile, the most critical part of the journey — the formation of preference — is happening elsewhere.
As Shane H. Tepper, cofounder of Resonate Labs, has observed in his work with B2B teams, much of what shows up as “direct” traffic today may actually be the result of AI-influenced research that never gets properly attributed.
In other words, companies may be benefiting from channels they don’t measure — or worse, losing to competitors in places they don’t even know exist.
Optimizing for the Wrong Signals
The consequences are subtle, but significant.
If “direct traffic” is partially composed of misclassified AI-influenced visits, then many of the conclusions drawn from it start to wobble:
- Brand strength may be overestimated
- Channel performance may be misread
- Budget allocation may be misdirected
Teams continue optimizing based on what they can measure — even as the most important influences move outside that measurement framework.
It’s not that the data is wrong.
It’s that the system producing it no longer reflects the full picture.
A Shift From Tracking to Understanding
There’s no simple fix for this.
No new dashboard will suddenly illuminate every AI-driven interaction. No attribution model will fully reconstruct a journey that unfolds inside a closed system.
What’s emerging instead is a different way of thinking about performance.
Less about tracking every step.
More about understanding where and how decisions are formed.
Because the uncomfortable truth is this: the most important part of the buyer journey is no longer happening on your website, or even on channels you control.
It’s happening in environments where visibility is partial, attribution is fragmented, and influence is inferred rather than observed.
And in that world, success won’t belong to the brands with the cleanest dashboards.
It will belong to the ones that learn how to operate — and compete — in the dark.
