AI is growing your market and shrinking your segment

Traditional B2B segmentation assumes that companies with similar characteristics will behave in similar ways. AI is weakening that assumption.

Two companies can be the same size, in the same industry, with the same tech stack and the same buyer, and still have completely different problems, expectations and alternatives. Because they’re at different stages of AI maturity.

In some markets, AI maturity may now predict buying behavior better than the firmographics we’ve always used to define segments. That makes it more than a messaging consideration. It’s a segmentation variable. And it can change both how a company buys and whether it’s addressable at all.

Segmentation is supposed to predict behavior

We don’t segment companies by employee count, industry, geography, growth rate or tech stack because those characteristics are interesting in themselves.

We do it because they’re proxies. They help us predict:

• What problems a company is likely to have
• What it will prioritize
• What solutions it will consider
• What it expects from a product
• How it buys
• What it might pay

The assumption underneath is simple: companies that look alike will behave alike often enough for the grouping to be commercially useful.

AI is making it less reliable.

Two identical companies, two different realities

Take two professional-services companies with 500 employees each. Same industry. Same size. Same geography. Same buyer: the Head of IT.

• Company A is deciding whether employees should be allowed to use ChatGPT.
• Company B has rolled out enterprise AI. Employees are building their own agents, and IT is dealing with permissions, unreliable source material and governance.

On paper, they’re one segment. In reality, they don’t have the same problem. They don’t have the same urgency. They won’t respond to the same message. They don’t expect the same product. And they may not even consider the same category of solutions.

Demographically, they’re the same buyer. Commercially, they’re increasingly different segments.

AI adds another axis to your segmentation

Your existing segmentation doesn’t disappear. You still look at industry, size, geography, role and use case.

But now another variable cuts through all of them: AI maturity.

You don’t need an elaborate maturity framework. A simple progression is enough to see it: exploring, adopting, operationalizing, AI-native.

The labels aren’t the important part. What matters is that the same company profile behaves differently at each stage.

It affects much more than messaging

The obvious conclusion is: “We should tailor our messaging to different levels of AI maturity.”

Yes. But that’s the least interesting consequence. AI maturity can change almost everything about how a segment buys.

• The problem
  The same knowledge-management problem can be about findability for one company and about AI reliability for another.
• Urgency
  “Prepare your data for AI” means little to a company that’s only experimenting. Once AI is deployed, it becomes an immediate operational problem. And urgency decides whether a market will buy.
• Product expectations
  APIs, MCP, AI agents, permissions, integrations and automation can move from nice-to-have to table stakes.
• Alternatives
  More technically mature organizations increasingly consider building instead of buying. Your real competitor might be their own team.
• Value
  What solving the problem is worth depends on what the company is already doing with AI.
• Product fit
  Most importantly, your product might simply stop being a good fit at a certain level of maturity.

Which leads to the uncomfortable part.

Parts of your segment may no longer be addressable

Suppose you’ve always counted 5,000 companies as your target segment.

Look at them through AI maturity, and you might find something like this:

• Some aren’t mature enough to have the new problem yet.
• Some are exactly at the stage where the problem becomes urgent.
• Some have moved far enough that they expect capabilities you don’t have.
• Some have become capable enough to solve it themselves.

Suddenly, 5,000 companies with the same firmographics don’t add up to a 5,000-company addressable segment. It’s the same lesson as sizing your market beyond TAM, SAM and SOM: the number that matters is the companies that can and will buy from you now.

This is the difference between AI maturity as a messaging variable and AI maturity as a market variable.

AI can create markets while shrinking segments

There’s a lot of talk about AI creating enormous new markets. That’s true.

But something else is happening at the same time. AI can grow the overall market while making individual addressable segments smaller.

Because adoption isn’t happening evenly. The market fragments by how far organizations have come, what they’ve already put in place, what problems that has created, and what they’re capable of doing next.

So a huge “AI market” doesn’t automatically mean a huge, coherent pool of buyers for your product.

The variables you segment on need to change

Don’t throw away firmographics. A 50-person company and a 50,000-person enterprise still behave very differently.

But ask whether your segmentation variables still predict buying behavior.

Instead of stopping at “Who are they?”, add: “Where are they in the technological change that’s affecting this market?”

And then:

• What problem does that create for them right now?
• What do they expect from a solution?
• What alternatives do they have?
• Is our product actually a good fit at this stage?

Segment on what predicts buying behavior

Segmentation was never really about employee count, industry or geography. Those were useful because they helped us predict how companies would behave.

If AI maturity now predicts that behavior better, your segmentation needs to reflect it.

The companies in your ICP may still look the same on paper. That doesn’t mean they’re still the same market.

Where I can helpMarket & Customer Intelligence