A striking number surfaced in Digiday’s marketing coverage this week: 73% of CMOs have invested in AI visibility monitoring. Nearly three-quarters of the buy side is now paying to know when and where their brands appear in AI Overviews and ChatGPT answers. The problem is what happens after the dashboard lights up — which, at the moment, is nothing measurable.
Tools like Scrunch, Profound and Semrush can track the citations. What they cannot do is connect a citation to revenue. So marketers are left triangulating — stitching together conversion data, traffic patterns and custom marketing-mix models to guess at whether being mentioned by a chatbot moved a single unit of product. Skyword CEO Andrew Wheeler compressed the frustration into five words: “Citation alone isn’t good enough.” Roast’s John Barham was blunter about the tooling landscape: “There is no one tool out there that can paint you a picture of the universe.”
Most publishers will read this as a buy-side curiosity — marketers wrestling with attribution, as they eternally do. That is the wrong reading. What this story actually describes is a funded, unfilled brief: a measurement hole the buy side has already budgeted against, bought tools for, and failed to close. In a market where publishers spend most of their time defending CPMs against shrinking open-web budgets, a documented gap that 73% of CMOs are actively spending to fill is not a curiosity. It is a sales opportunity with the budget line already open.
01The gap sits under your pitch too
Here is why this belongs in a publisher briefing rather than a marketing one: the same measurement hole sits underneath every AI-era pitch a publisher makes. Sponsored content that gets cited in AI answers. Licensing deals priced on the value of your content to a model. First-party audience products sold on quality of attention. In every case, the publisher’s argument eventually arrives at the same cliff the CMO’s dashboard does — a mention happened, and then what? Whoever can evidence a downstream commercial outcome, rather than a citation, wins the budget conversation. Right now, on this specific question, almost nobody can.
Which is what makes the publisher position unusual. The visibility vendors are structurally stuck: they observe the AI surfaces from outside and have no view into what a user does afterwards. Platforms could close the loop but have little incentive to. Publishers, alone in this chain, own the ground where the after happens — on their own properties, with their own logged-in users, measured with their own first-party data. The newsletter signup, the return visit, the commerce click, the registered account, the subscription start: all of it is observable without a browser vendor’s or a platform’s permission, because it happens on infrastructure the publisher controls.
That does not mean publishers can attribute a chatbot citation to a sale — nobody can, and pretending otherwise would be exactly the over-claiming this desk flags in vendor studies. What publishers can do is evidence post-exposure behaviour on their own properties with a rigour the visibility tools cannot match, and sell that evidence as part of the package.
02Why this matters for publishers
| The budget already exists | 73% of CMOs have spent money on AI visibility tooling that stops at the mention. A publisher offering outcome evidence is not asking for a new budget line — it is selling into one that is open and underserved. |
|---|---|
| First-party outcome data is the asset only you hold | Vendors see the citation; platforms keep their data; publishers own what audiences do on their properties after exposure. In a measurement stalemate, the party with observable behaviour wins. |
| The same gap undermines your own pricing if you ignore it | Every licensing negotiation and sponsored-content renewal eventually hits the "and then what?" question. Publishers who cannot answer it are defending CPMs; publishers who can are selling documented demand. |
| The window is temporary | The measurement vendors will eventually close some version of this loop. The advantage belongs to publishers who package their evidence while the gap is still open. |
The industry has spent two years arguing about whether AI visibility matters, and the buy side answered with its budget: 73% bought the dashboards.
03What publishers should do
04The bottom line
The industry has spent two years arguing about whether AI visibility matters, and the buy side answered with its budget: 73% bought the dashboards. Now they are discovering the dashboards stop at the mention, and no vendor can take them further. Publishers are the one party in this chain holding the data that describes what happens after — and most are leaving it unpackaged while they argue about CPMs. Until the tooling catches up, a publisher with clean first-party outcome evidence is selling something the majority of CMOs have already budgeted for and currently cannot buy. Briefs this clearly written do not stay unanswered for long.