HOW BRANDS GROW (IN THE AGENTIC AGE) Part 1: First to Mind.

The battle for tomorrow's shortlist may begin long before anyone asks AI to create one.

TL;DR

Everyone's fighting to get their brand onto AI's shortlist. But the more important battle may be happening before the buyer even knows what to buy.

The industry has found its latest obsession: getting brands recommended by AI.

Companies are tracking prompts, citations and rankings. New dashboards promise to tell companies whether ChatGPT knows they exist. Brands are being encouraged to measure how often they appear in AI-generated recommendations and what they can do to improve their standing.

There's good reason to pay attention. If a potential customer asks an AI assistant to recommend suppliers, being absent from the answer could matter.

But there's a problem with making that moment the centre of the strategy.

It assumes the buyer already knows what they're looking for.

Consider the CEO of a growing SaaS company who asks ChatGPT why, despite increasing revenue, the business seems to be getting harder to manage.

They're not looking for financial software. They're not evaluating suppliers or preparing a shortlist. They're trying to understand why the processes that worked when the company was smaller are starting to break down.

The conversation explores fragmented financial information, time-consuming reconciliations, disconnected banking and accounting systems, and the difficulty of maintaining a clear picture of cash flow. AI introduces approaches the CEO may not have considered, challenges some of their assumptions and perhaps draws on examples of companies that have tackled similar problems.

Some brands might be mentioned. Others won't be.

Months later, the CEO decides it's time to rethink how the company manages its finances. By then, their understanding of the problem has evolved, along with their view of what a credible solution might look like. Perhaps their perception of which companies are relevant has changed too.

The buying journey didn't begin when they started looking for a supplier. It began when they started trying to understand the problem.

And that raises a question the current debate about AI visibility is largely overlooking.

What happens when AI participates in shaping the buyer's understanding before the buyer even knows what category to search for?

The 95% who aren't buying

Marketing science has spent decades explaining why brands need to reach people who have no immediate intention of buying.

In 2021, Professor John Dawes of the Ehrenberg-Bass Institute articulated the 95/5 rule. In B2B categories with long replacement cycles, only a small proportion of potential buyers may be actively purchasing in any given quarter. The familiar 95/5 split is an illustration, not a universal ratio, but its commercial implication is important.

Most future customers aren't looking for you today.

They will, however, encounter business problems, consider different approaches and develop opinions about the companies operating in their market.

A brand's future growth depends partly on being remembered when a relevant buying situation arises and being accessible when buyers are ready to act. Marketing science describes these as mental and physical availability.

That distinction is important because it explains why capturing existing demand is only part of the job. Brands also need to build and refresh associations among people who may not become customers for months or years.

AI introduces another environment in which some of that learning may take place.

The interesting question isn't simply whether AI can recommend your brand when someone is ready to buy. It's whether AI-mediated conversations might influence the understanding and associations that buyers carry into that decision.

The conversation before the category

Imagine a financial institution struggling with customer onboarding. Its leadership team knows the experience is poor, but it hasn't established whether the underlying problem is identity verification, process design, legacy technology or customer communication.

Or consider a growing SaaS company whose finance team is spending more time reconciling information across systems than helping leadership make decisions. The founder knows the current approach isn't working, but hasn't established whether the problem is banking, accounting, internal processes or the way those systems connect.

Neither company begins with a neatly defined supplier category.

The problem comes first.

Historically, people have worked through these questions with colleagues, consultants, analysts, publications, industry networks and search engines. AI assistants now offer another way to explore them, often through extended conversations that allow users to test explanations and follow new lines of enquiry.

An AI assistant can connect a business problem with possible approaches, introduce an unfamiliar solution category or draw attention to examples of organisations addressing similar challenges.

The resulting answer may include brands. It may cite independent research or use a particular company's perspective to explain an issue. Or it may discuss the subject without mentioning any provider at all.

For a brand, the strategic question extends beyond whether it appears in a list of recommended suppliers. It concerns whether the brand is relevant, credible and represented in the information people encounter while working out what their problem actually is.

The contest may begin before the buyer knows what category to search for.

That's a different challenge from improving a position in an AI-generated ranking. It asks brands to think about the problems they want to be associated with, the evidence supporting those associations and the information available when buyers begin exploring a subject.

How brands get into people's heads

Mental availability concerns the likelihood that a brand comes to mind in a buying situation.

Professor Jenni Romaniuk's work on Category Entry Points explains how buying situations contain cues that help people retrieve brands from memory. These cues might involve needs, motivations, occasions or circumstances that make a particular category relevant.

Brands develop mental availability by building and refreshing associations with those cues. Advertising, distinctive creative assets, customer experience, PR and other forms of communication can all contribute.

AI doesn't overturn those principles. But it may change some of the encounters through which people learn about categories and brands.

An assistant might reference a company's research when explaining an industry challenge, describe an approach through a particular provider's work or cite an independent case study. These encounters could potentially introduce a brand to someone who had no reason to search for it directly.

Could that contribute to what the person subsequently remembers?

Possibly. But the distinction between machine visibility and human memory is fundamental.

Being mentioned by AI is not the same as being remembered by a human.

A citation isn't a memory structure. An AI recommendation isn't proof of mental availability. And appearing in an answer doesn't establish that someone's future purchasing behaviour has changed.

We don't yet have sufficient evidence to make those claims.

What we can do is investigate the environment in which those encounters occur. Which problems and category entry points matter to future buyers? What information does AI surface around them? Which brands appear, in what context, and supported by what evidence?

The causal effect on human memory remains an open question. The strategic relevance of understanding those conversations does not depend on pretending we've already answered it.

The shortlist is rarely a blank sheet

The importance of what happens before an active purchase is not a new discovery.

In its 2025 Buyer Experience Report, 6sense drew on a main survey of nearly 4,000 B2B buyers across North America, Asia-Pacific and EMEA. It reported that, in 95% of the purchases described by respondents, the eventual winning supplier was already on the buying group's Day One shortlist.

This is entirely separate from Dawes's 95/5 rule. The two figures describe different aspects of B2B buying.

We interpret Day One as the beginning of the buyer's recognised purchasing process, rather than the beginning of all previous learning and experience. The report does not precisely define that boundary.

What the finding illustrates is the importance of the knowledge and preferences buyers bring into supplier evaluation. Familiarity, reputation and previous experience can shape which organisations are considered before a formal selection process gets underway.

The research doesn't establish that AI created those preferences. Prior experience with providers was often important.

But it reinforces a commercially significant point.

By the time a supplier becomes aware of an opportunity, much of the groundwork may already have been laid.

There is also evidence that AI has become part of active B2B information gathering. Gartner surveyed 645 B2B buyers in August and September 2025 and reported in May 2026 that 45% had used generative AI during a recent purchase, primarily to gather information about vendors and products. Buyers continued to rely on human interactions to validate AI-generated information.

Gartner's research concerns active buyers, not people exploring problems months before a purchase. It therefore cannot tell us what happens in the period before a purchase begins.

Taken together, these findings give us two reasons to pay attention: prior knowledge matters in B2B purchasing, and AI is already participating in some of the research buyers undertake.

The question is what role AI might play in the period before those two things meet.

Your brand is no longer the only one telling your story

Of course, it never was.

Customers, journalists, analysts, competitors and professional communities have always contributed to a brand's reputation. What a company says about itself is only one part of what people come to believe.

AI introduces another intermediary into that relationship.

An assistant may assemble an answer from material published by the brand, independent research, customer accounts, editorial coverage and other accessible sources. The resulting explanation is not a controlled brand communication. It is a synthesis produced by a system whose sources and behaviour can vary.

That matters because brands cannot assume that their preferred positioning will be reproduced simply because they've published it.

The work begins with clarity. A company needs to explain which problems it addresses, what makes its approach distinctive and what credible evidence supports its claims. Its own website, research, product information and other published material are important parts of that foundation.

But the information environment extends beyond the brand's own channels.

Independent research, customer testimony, credible editorial coverage and third-party case studies can corroborate claims and place a company within the wider context of its category.

That doesn't mean independent citations guarantee inclusion in AI answers. Different systems retrieve and synthesise information differently, and the presence of a source doesn't establish why an answer was produced.

The opportunity is broader than accumulating mentions.

A brand needs to be credible in the places where its story is being assembled, not just persuasive in the places where it tells that story itself.

For marketers, that changes the questions worth asking. Is the brand associated with the problems it genuinely solves? Can those associations be supported by evidence? Are credible sources available beyond the company's own claims? And when AI systems discuss those problems, how accurately do they represent the brand's relevance?

This is brand-building work expressed through a changing information environment. It demands strategic clarity, credible proof and an understanding of how the wider market describes the company.

It cannot be reduced to a technical optimisation exercise.

First to Mind

The temptation is to measure what is immediately visible.

Prompt rankings, citation counts and recommendation frequency can provide useful observations about how AI systems represent a brand. They may reveal gaps, inconsistencies and opportunities to improve the information available.

But those observations should not be mistaken for the outcome.

The outcome we're interested in is whether future buyers know the brand, associate it with relevant buying situations and consider it when a need arises.

Established brand-building activity remains essential to that task. AI-mediated learning introduces a new area to investigate, not a replacement for advertising, creativity, customer experience or the other influences on human memory.

First to Mind is about how brands become known and relevant before buyers are ready to buy, including the potential role of AI in that process.

The practical response is to identify the problems and category entry points that matter, build credible associations around them and observe how those associations are represented in AI-mediated conversations.

That gives brands a starting point for learning without pretending that visibility metrics prove an effect on memory or sales.

The battle before the shortlist

Most of the industry's attention is focused on what happens when someone asks AI to recommend a supplier.

Who appears? What gets cited? Which brands make the shortlist?

Those questions matter. They're the subject of Part 2 of this article.

But by the time a buyer asks them, they may already have developed a view of the problem, an expectation of what a solution should look like and some familiarity with the companies that might provide it.

AI could become another influence on that accumulated understanding. How much it matters, and under what conditions, remains to be established.

Yet concentrating exclusively on the recommendation moment risks overlooking the conversations that precede it.

Most of the industry is asking how to get a brand into the answer.

We're asking what happens before the question exists.

The battle for tomorrow's shortlist may begin long before anyone asks AI to create one.

That's First to Mind.

When buyers become ready to investigate suppliers, compare alternatives and make decisions, a second challenge comes into focus.

First to Find.

Next in the series: How Brands Grow in the Agentic Age — Part 2: First to Find.

Wild Ducks. Never Tame.

Research and further reading

1. John Dawes, Ehrenberg-Bass Institute (2021) — The 95:5 Rule. https://johndawes.info/the-955-rule/

2. Jenni Romaniuk, Ehrenberg-Bass Institute (2022) — Category Entry Points in a Business-to-Business World. https://marketingscience.info/news-and-insights/category-entry-points-in-a-business-to-business-b2b-world

3. Gartner (20 May 2026) — B2B buyers' use of generative AI and human validation. https://www.gartner.com/en/newsroom/press-releases/2026-05-20-gartner-survey-finds-sixty-nine-percent-of-b-two-b-buyers-turn-to-sales-reps-to-validate-ai-generated-insights

4. 6sense (2025) — B2B Buyer Experience Report. https://6sense.com/science-of-b2b/buyer-experience-report-2025/