Category Design - Knowledge Base
Category Design for AI Companies
AI companies have a specific category problem. The technology is genuinely new. The capabilities are real. But the market framing almost always defaults to something older - a familiar category, a known competitor, an existing budget line. That default kills pricing power and buyer urgency before the conversation starts.
The AI category problem
Most AI companies are built on a genuinely new capability. The founders can see what is possible that was not possible before. The product does things that existing solutions cannot do. And yet the market persistently places the company inside an existing frame - automation, analytics, data management, enterprise software - and applies the criteria from that frame to evaluate the new thing.
This is not buyer stupidity. It is how markets work. Buyers need a reference point to evaluate something unfamiliar. If the company does not provide that reference point clearly and forcefully, the market will find one. It will almost always be the wrong one.
The wrong reference point creates a specific set of problems. The comparison set is wrong: buyers compare the company to incumbents it should not be competing with. The budget is wrong: the purchase is drawn from a budget line that does not reflect the actual value being delivered. The price expectation is wrong: buyers apply pricing logic from the old category to the new one. All of this happens before a single sales conversation.
Why AI makes category design more urgent, not less
There is a tempting logic that says: AI is moving so fast that category definitions are unstable, so it is better to wait for the market to settle before committing to a category. This logic is wrong in both directions.
First, the market does not wait for companies to define their category. It categorises constantly, using whatever information is available. A company that waits is not preserving optionality. It is ceding the categorisation decision to the market, to competitors, and to analysts who have their own frameworks.
Second, the speed of AI development makes category ownership more valuable, not less. In a world where execution speed is compressing - where what takes one company eighteen months to build, another can build in six - the durable advantage is not the product. It is the category. A company that has defined the frame, named the problem, and built market belief that it leads this category has something that cannot be copied in a product sprint.
The companies that will lead AI categories in five years are the ones that are defining those categories now. Waiting for the market to settle is waiting for someone else to set the terms.
The generic AI positioning trap
There is a pattern that is becoming common among AI companies. The product is real and valuable, but the positioning defaults to capability language: we use AI to do X faster, better, more accurately than the alternative. The problem with this framing is that it is true of many companies simultaneously.
When multiple companies claim the same capability improvement, the category becomes AI itself. The buyer's comparison set is every AI company. The decision criteria become price and integration ease. The conversation that should be about a specific problem the company uniquely owns becomes a procurement negotiation about features.
This is the generic AI positioning trap. It is not solved by adding more specificity to the capability claims. It is solved by moving upstream - to the problem the company owns, the category it leads, and the frame that makes it the logical choice rather than one of many options.
An AI company that owns a specific problem in a specific market, with a specific category name that buyers can use to explain what they bought and why, is in a fundamentally different commercial position than an AI company that is competing on capability claims in a market where capability is becoming abundant.
The investor narrative problem for AI companies
The category problem for AI companies is also a capital problem. Investors apply category logic to valuation multiples. A company in a clearly defined, growing category with a credible claim to leadership gets valued differently from a company that is doing interesting AI things in an undefined space.
The question investors are actually asking - even when they phrase it as a product question or a market size question - is often a category question: what is this company leading, who will it beat, and why will it be the one that captures most of the value in this space?
An AI company that cannot answer that question clearly is not just leaving money on the table in its current round. It is making the next round harder, because the story has not become more legible. Category clarity is what makes each successive fundraise easier rather than a reset.
The moment of recognition for AI companies
For AI companies, the recognition moment often comes through the investor relationship. A lead investor who has backed multiple AI companies starts to see the pattern: the product is differentiated but the market is not seeing it that way. Deals are taking too long. The comparison set is wrong. The pricing is under pressure.
When the investor, the board, or a key enterprise customer says this out loud - that the company needs a clearer category position, not a better product - that is the moment to act. The internal permission to make a real change is present. The urgency is real. Category design work at this moment has force.
What Venturoxx does for AI companies
Venturoxx works with funded B2B software, AI and data companies - typically Series A to C - when the category position is unclear, contested, or actively working against the company's commercial interests.
The Diagnostic finds the real problem: is it the category frame, the positioning within a correct category, the GTM alignment, or something else? The Blueprint rebuilds the position from the category up: the problem the company owns, the category name, the POV, and the GTM system aligned to it. The Category Control maintains that position through the live decisions that companies make every day - product launches, partnership announcements, pricing changes, investor conversations - any of which can quietly erode a category if they are not evaluated against the frame.
The work is done by Richard Poolman. Three decades of building first commercial engines for category-defining enterprise software companies in Europe provides the operating judgment that distinguishes this from advisory work built on AI hype rather than category design practice.