Ask any enterprise marketing head how their budget was structured five years ago and you’ll hear roughly the same answer. Some money for Google rankings. Some for paid reach on Meta and LinkedIn. A dashboard tracking clicks, impressions, cost per lead, pipeline. The whole machine rested on one assumption: people search, people click, and some of them buy.
That assumption is wobbling. I see it in conversations with marketing teams almost every week now. Their buyers have stopped opening ten tabs to compare vendors themselves. They type the question into ChatGPT or Gemini or Perplexity and let the machine do the comparing. On paper it looks like a minor change in behavior. In practice, it moves the decision to a place most funnels cannot even see.
Here is how I’d frame it. For a long time, winning meant being visible, so companies bought attention wherever it was sold. Then winning meant being findable, and an entire industry grew up around search rankings. What’s emerging now is a third stage, where the real contest is over which names an AI puts into its answer. Two or three brands make the cut. Everyone else, as far as that buyer is concerned, may as well not exist.
The mechanics matter here. A search engine gives you links and steps back. An AI assistant does the reading, does the weighing, and hands over a conclusion. So there is now a layer of software sitting between your buyer and your brand, quietly deciding whether you’re part of the conversation at all. Enterprises spent a decade perfecting everything that happens after the click. The trouble is that the shortlist increasingly gets written before any click happens.
And the numbers behind this aren’t fringe anymore. OpenAI said ChatGPT passed 900 million weekly users in early 2026. Google has folded AI Overviews into results seen by more than 2 billion people a month. Back in 2024, Gartner predicted traditional search volume would fall 25% by 2026, and at the time plenty of us thought that was bold. In some sectors, it undershot.
Which changes the question a CMO has to answer. “Do we rank?” is still worth asking. But the harder one is this: when someone asks an AI to name the best options in your category, does your company come up at all? Most enterprise budgets were never built with that question in mind.
Buyer behavior data backs this up. Semrush’s research puts the share of Google searches ending with no click to any website at close to 60%. Bain found four in five consumers now lean on these zero-click answers for a big chunk of their searches, and estimates organic traffic has fallen somewhere between 15 and 25% as a result. Sounds grim. Except Semrush also found the visitors who do arrive through AI answers convert at roughly four times the rate of regular organic visitors. Fewer people show up, but the ones who do are far closer to buying. Traffic and influence used to move together. They don’t anymore.
There is another awkward discovery waiting for enterprise teams: AI systems don’t score brands the way Google’s blue links did. They pull from reviews, from mentions on sites they consider credible, from structured data, from whether your company information stays consistent everywhere it appears. Plenty of companies with excellent rankings are finding out they barely register in AI answers. What these systems are really doing, as far as I can tell, is working out whether the wider internet vouches for you.
Measurement gets messy too. Sessions and last-click attribution still have their place, but they miss more of the journey every quarter, partly because AI referrals often land with no referrer and get lumped into direct traffic. Teams on top of this have started tracking how often they get cited in AI answers, what actually gets said about them there, and whether branded searches are climbing. Your analytics can tell you who showed up. It has no idea who got left off the shortlist.
The enterprises handling this well aren’t doing anything exotic. They’re cutting channels that were only kept alive by habit and putting real depth behind fewer ones. They’re writing content a machine can parse, not just a human. They’re chasing coverage on trusted third-party sites, because that’s the validation AI systems reward. And they’re building first-party audiences, since no algorithm update can take an email list away from you.
None of this means the fundamentals are dead. It means the buyer moved, again, and strategy has to follow. Enterprises spent twenty years fighting over positions on a results page. The fight now is over a sentence in an answer. The companies rethinking their marketing today aren’t chasing something shiny. They just noticed where their customers went.
(Views are personal)
