Stop Asking AI About Your Brand. Ask It About the Problem.

The question determines the answer

There is a simple way to make almost any brand look impressive in AI.

Ask AI directly about it.

"What is Company X known for?" "Why should I choose Company X?" "What makes Company X different?"

These questions can produce detailed, reassuring answers. They can also create a misleading picture of how strongly AI actually associates the brand with its market.

At KVD, we believe measuring AI brand understanding requires a more demanding approach.

Do not start with the brand.

Start with the problem.

Branded questions tell you what AI knows

Branded queries still have an important role.

Ask an AI system to explain your company and you can examine what it associates with the brand: category, products, differentiation, proof points, audiences, competitors, sentiment and purpose.

This helps determine whether AI's representation corresponds with the brand defined in the Brand Book.

But there is an obvious advantage hidden inside the question: you have already given AI the brand name.

The real test begins when you take it away.

Problem-first questions tell you whether the brand matters

Imagine a company providing technology that dramatically reduces the time required for microbial testing.

We could ask:

"What does Company X do?"

But a potential customer may ask something quite different:

"How can we reduce microbial testing time?"

"Which technologies can improve sterility testing?"

"What companies provide rapid microbial detection?"

Now the brand has to earn its place in the answer.

KVD calls these problem-first queries. They test whether AI connects the problems customers experience with the brand intended to solve them.

That is a considerably more revealing measure of AI understanding. The KVD AI Brand Book methodology therefore examines branded queries alongside problem-first and category-level questions rather than relying on direct brand prompts alone. The_AI_Brand_Book.pdfPDF

Category questions reveal who AI thinks owns the market

The next level is the category.

Ask AI to identify leading companies, explain the competitive landscape or recommend providers within a particular category.

Which brands appear?

Which brands appear first?

What attributes are associated with each?

Which companies are treated as specialists, innovators or established leaders?

And perhaps most importantly: does AI place your company in the category you believe you occupy?

This begins to reveal something traditional brand tracking was never designed to measure: the competitive structure AI has constructed around a market.

Your strategy may identify five competitors. AI may identify ten different ones.

That difference matters.

Prompts are research instruments

This is why KVD does not view prompts simply as a way of "talking to ChatGPT."

In AI brand research, prompts become research instruments.

They need to be deliberately designed, repeated and compared. Different types of questions test different dimensions of the brand.

Branded queries examine representation.

Category queries examine association.

Problem-first queries examine relevance.

Competitive queries examine positioning.

Recommendation queries examine whether the brand surfaces when an AI system is asked to help someone make a choice.

No single prompt provides the answer.

The pattern across many prompts begins to reveal it.

One model is not enough

There is another complication.

There is no single AI interpretation of your brand.

Different models can produce different representations because they operate differently and may draw upon different information environments. The KVD methodology therefore calls for testing across multiple major AI platforms and comparing the results. The_AI_Brand_Book.pdfPDF

The objective is not to find one flattering answer.

It is to identify patterns.

Where is the brand consistently understood? Where does representation change? Which competitors repeatedly surface? Which intended attributes disappear? Which problems reliably trigger the brand?

This turns AI brand research from an anecdote into a methodology.

The harder question

For marketers, this creates an important distinction.

"What does AI say when we ask about our brand?"

is not the same question as:

"Does AI think of our brand when nobody mentions our name?"

The first measures recognition.

The second begins to measure relevance.

At KVD, we believe that distinction will become increasingly important as AI becomes part of how customers discover categories, investigate problems and compare solutions.

Because ultimately, brands do not simply want AI to know who they are.

They want to occupy meaningful territory in the machine's understanding of the market.

The next question is how much of that territory your brand actually owns.

We call it Share of AI Understanding.

About KVD KVD is a strategic marketing consultancy helping companies develop Brand Books and AI Brand Books for their brands. KVD is based in Cambridge, Boston.

Read the cornerstone guide: Brand Building for Humans and AI.