In his latest column, Alex Lamaro-McCrindle, digital lead, Miroma Founders Network, looks at the agentic future, and how to market to agents…

A couple of weeks ago I was at Founders Forum Global. The conversation was familiar territory, AI in marketing, changing customer behaviour, the compression of the buyer journey. Good discussion, relevant topics of the moment.

Then, almost as an afterthought, someone from Meta said something that pricked my ears.

"Next we need to figure out how to market to agents."

It landed as a half joke. One of those comments thrown out at the end of a panel that gets a knowing laugh and then disappears into the next conversation. But I wrote it down. Because underneath the throwaway delivery was an articulation of a conversation we've been having with a few different businesses, albeit not phrased in such an explicit way.

We have spent decades learning how to market to humans. We have studied their behaviour, mapped their journeys, built personas, and obsessed over every touchpoint. But what happens when a growing number of those humans start delegating their research, their consideration, and their purchasing decisions to AI agents?

Who are you marketing to then?

To be clear, we are not there yet. Most people using AI today are doing something far simpler, typing a question and reading an answer. The idea of delegating a purchasing decision to an agent feels, for most people, a long way off.

But bell curves have a habit of moving faster than we expect. The early adopters of agentic behaviour are not evenly distributed. They are concentrated in the places that tend to predict what everyone else does next. Technology businesses. Financial services. Senior decision makers at high growth companies. If your buyers live in those worlds, the curve is already more advanced than the general population numbers suggest.

To understand why this matters, it helps to think about what an agent actually does on behalf of a buyer.

A human researching a purchase will search, browse, compare, ask colleagues, read reviews, and gradually form a view. Moments where marketing can show up, influence, and persuade. Even if much of that journey is hidden from you, the human is still moving through it in a way that is recognisably familiar.

An agent changes that entirely. Give it a goal and it researches, evaluates, and returns a recommendation. The consideration phase, the part of the journey where opinions are formed and shortlists are built, happens inside a system you cannot see and cannot influence in the moment.

The buyer didn't go dark. They just sent someone else into the dark on their behalf.

So what are the implications of agents delivering recommendations on your marketing?

This is the question that the Meta comment put right at the top of my mind.

Marketing has always been about influencing a decision maker. We have built entire disciplines around understanding human psychology, emotional triggers, and behavioural nudges. We obsess over creative, storytelling, and brand identity because humans respond to those things. They make us feel something.

Agents don't feel anything. They evaluate. They synthesise. They score against criteria. An agent researching solutions on behalf of a buyer doesn't care about your brand film, your tone of voice, or your award-winning campaign. It cares whether you appear in authoritative sources. Whether independent voices speak well of you. Whether your positioning is clear and consistent enough to be accurately represented in a synthesised recommendation.

In short, the agent is asking one question on your buyer's behalf. Can I find credible, consistent, independent evidence that this brand is the right answer?

If the answer is no, you don't make the shortlist. And the buyer never knew you existed.

So what do you actually do about it?

The instinct will be to look for a new tool or a new channel. That tooling is coming, there are already early-stage platforms tracking brand mention and sentiment across LLM outputs (definitely check them out). The more important shift right now is strategic and practical.

First, audit what agents are actually saying about you. This costs nothing. Systematically prompt the major LLMs with the questions your buyers are likely asking. Who are the leading providers of X. What should I look for when choosing Y. Compare these options. Do this monthly and track the changes. It is the most direct view you currently have into your agent visibility.

Second, obsess over third party representation. Agents weigh independent sources heavily because they are designed to. Trade press, analyst coverage, review platforms, Reddit threads, these are the infrastructure of your agent visibility. If you are not present and well represented in these sources, you are not in the conversation.

Third, sharpen your positioning until it is impossible to misrepresent. Vague brands get averaged into nothing when an agent synthesises a category. A sharp, specific, consistently repeated point of view is far more likely to survive that process. If you cannot describe what you do and who you do it for in two clear sentences, an agent probably can't either.

Fourth, make your content machine readable. Agents digest content differently to humans. Clear, well structured, factually dense content that answers specific questions directly is far more likely to be extracted and used than narrative heavy brand copy. Structured data on your website helps agents understand exactly what you do, who you serve, and what category you belong to.

None of this is as new as it feels.

Before digital gave us the illusion of perfect visibility, marketers built great brands and won markets without tracking a single click. They did it through authority, reputation, consistency, and presence in the places their buyers trusted.

Those are exactly the qualities that win with agents. The fundamentals haven't changed. The audience has.

So when that person from Meta threw out their half joking comment at the end of a panel, they were probably expecting a laugh and a change of subject. What they were describing was the next evolution of one of marketing's oldest challenges.

Understanding who you are really talking to.

The answer, it turns out, is no longer just the person sitting behind the screen. It's everything that informs the system sitting in front of them.