Brands have always been built for people. Now increasingly, machines are becoming part of how people decide what to buy, use and trust.

That matters because almost everything we know about brand is grounded in human behaviour.

Brands create meaning around products, services and organisations. Brand strategy defines what we want to be known for, builds associations around it and works to make those associations mentally available when people come to choose.

AI introduces a different participant in that choice.

It is increasingly helping people find products, compare offers, locate services, understand information and identify expertise. So what happens when a discipline designed to influence human understanding also needs to work through machines?

‍

Machines don't choose like people do.

AI can interpret what a brand says and how others describe it. But it doesn't experience a brand's meaning, story or symbolism in the way people do.

When someone asks AI for a recommendation, it starts with a different problem: what is relevant to this question?

‍

Where can a young person in regional NSW get mental health support without a referral? ‍

Who has experience helping councils adapt infrastructure to climate change?

‍

To respond, AI connects the question with relevant organisations, services, people and information, looking for enough evidence to support an answer.

For people, brand creates meaning and mental availability. For machines, that meaning has to survive a different process: being interpreted in relation to a specific question and the available evidence.

‍

Can brand meaning travel into the answer?

The challenge is whether the associations we aim to build around a brand can travel into the answer an AI provides. Research suggests that AI recommendations aren't limited to what's explicitly asked. Models can infer implicit preferences while needs, constraints and distinctive positioning cues can change which brands they surface.

AI puts more emphasis on clearly connecting what you offer to an audience, their needs and the benefits you provide. But what you want to be known for still matters. If those distinctive associations are consistently reinforced by evidence, they can shape not only whether you appear in an answer, but how you are understood when you do.

This suggests a new question for brand strategy: what questions should your brand be the answer to?


Building an evidence architecture

Start with the questions that sit at the intersection of what audiences need, what the organisation wants to be known for, and where it has something genuinely valuable to offer.

There is a useful precedent in knowledge architecture. When designing ontologies and knowledge graphs, practitioners use competency questions to define what a system needs to be capable of answering, then work backwards to the information and relationships required.

Brands could do the same.

Once you know the questions that matter, you can ask: what would need to be known about us for our brand to be a credible answer?

For a consultancy to be recognised for helping councils adapt infrastructure to extreme weather, it isn't enough to claim climate expertise. AI might need to connect the organisation with relevant specialists, services, projects, research, outcomes, clients and external recognition.

No single proof point establishes the association. The system of reinforcing evidence does.

This is the role of evidence architecture: creating, connecting and making visible the evidence that supports the propositions and associations you want your brand to be known for.

Some of that evidence will live in communications. Some will exist elsewhere: in products and services, people and expertise, research and data, partnerships, customer outcomes and what credible third parties say about you.

The job isn't to create more content or repeat the same claim more often. It's to build the evidence from which the right connections can be made. ‍

‍

A new job for brand management

Brand management is already an orchestration job. AI expands that job again.

Evidence architecture reaches beyond communications. If you want to be recognised for expertise, you need to demonstrate it. If you want to own an issue, you may need to create distinctive knowledge around it. If you want a service recommended for a particular need, its proposition needs to be clear. If you want to be seen as authoritative, credible sources beyond your organisation may need to reinforce that view.

Brand management therefore starts to reach further into knowledge, services and products, data, partnerships and organisational capability.

Brands aren't becoming machine-first. But they are no longer be human-only.

We still need to create meaning, trust and mental availability for people. But we may also need to make that meaning legible to the machines increasingly helping them choose. That means asking not only what do we want to be known for? But what questions should we be the answer to, and what evidence will earn us a place in that answer?