After two years in which marketing events treated AI as the only subject worth discussing, the mood at the ANA Masters of B2B Marketing Conference in Chicago this month was notably calmer. Speakers from Shell, GE Aerospace, LinkedIn, SAP and Prudential Financial, none of them coordinating, arrived at the same reframing: AI is no longer the question but the conditions, and the question worth asking is what a brand is actually trying to say.

For B2B marketers the shift matters because it puts the discipline back at the centre. If AI is now ambient — “the air now”, as GE Aerospace’s chief brand officer Melissa Washko put it — then the scarce and valuable work is the human part: the story, the trust and the reputation that decide whether a brand is the one a buyer reaches for. And in B2B specifically, the speakers argued, the stakes make that more true, not less.

From “how do we use it” to “what do we want to say”

Washko’s own session barely mentioned AI at all. It was about craft, trust and the long game of building a brand, and that, she suggested, was the point: the field has spent two years fixated on how to use the tools rather than on the story it wants them to tell. Treating the technology as settled enough to recede into the background is itself the mature position.

Dean Aragon, chief executive of Shell Brands International and ANA board chair, made the same case from the main stage, noting how much more sober the room felt about AI than in previous years. The conversation, as he framed it, had moved from governing the technology to governing the way people interact with it — and, in an increasingly agentic world, served as a reminder of human agency rather than a surrender of it. For a marketing function that has spent two years being told it must change everything, that is a quietly significant recalibration: the tools are now stable enough that judgment and intent are the differentiators again.

Why AI raises the stakes in B2B, not lowers them

The more provocative argument was that large language models will reshape B2B more than consumer marketing, precisely because of what is at stake. Nobody consults an AI assistant to choose a coffee or a shampoo, Aragon observed, but a buyer weighing a $20m enterprise software contract almost certainly does — using it, in his phrase, as an analysis and scrutiny adviser. In high-consequence decisions, AI has become the first port of call rather than the last.

That relocates the moment of influence. By the time a salesperson enters the conversation, the buyer has already formed a view, shaped by what the assistant surfaced and synthesised long before any human pitch. For marketers it means the decisive ground is now the earliest and least controllable stage of the journey — which is exactly the territory where brand and reputation, rather than a well-timed sales call, do the work. The implication is not that AI diminishes marketing’s role but that it moves it upstream, to the formation of the view itself.

Reputation is the new visibility

LinkedIn’s Jann Schwarz gave the shift a structure. The old B2B playbook was built on visibility — dominate search, pay for attention, optimise for reach — and it is breaking down: by his account, roughly 83% of searches no longer end in a click, because AI answers the question before a buyer ever reaches a website. The new requirement, he argued, is not visibility but reputation, and it is built in three stages. A brand has to be discoverable in the sources AI itself draws on, through credible expert voices in trusted networks; credible enough, through visible proof and expertise, to earn a place on the shortlist; and then winnable.

The last stage is the one most marketers underplay. Winning a buying-group decision means warming up stakeholders who are not especially interested in the product — the hidden buyers worried about compliance, security or financial risk, who can make up half the people around the table. Schwarz pointed to platform data suggesting buyers more often choose the vendor they could defend if the decision went wrong than the one they were simply confident would perform. Winning that confidence is a brand job, not a sales job. It dovetails with research from LinkedIn and Bain, cited at the event, that 81% of successful tenders went to companies already known or aspirational to the buying group before the process even began. The preferred name is in the buyer’s mind before the brief exists, and unseating it, as Aragon put it, means beating the incumbent rather than settling for a draw.

The throughline is unglamorous and, for marketers, reassuring: AI changes the medium, not the job. The brands that win the high-stakes B2B decision are the ones already trusted and already known when the buyer — or the buyer’s AI adviser — starts looking. The agentic era, on this reading, is less a reason to hand the work to machines than a reminder of the one thing they cannot do, which is decide what is worth saying.