A flawless product explanation cannot rescue an AI company that has no clear first market. The first repeatable market is a narrow group of buyers with the same urgent problem, a reachable sales path, and enough willingness to pay that one sale can become ten.
In 2001, Dean Kamen was preparing to reveal the Segway in New York. The machine balanced itself, responded to the rider’s movements, and represented years of difficult engineering. Before the public launch, prominent investors and technology leaders had already seen its potential.
The unresolved question was smaller and more dangerous: who would buy it first, for which repeated journey, through what channel?
Steve Kemper documented the development and commercial uncertainty in Code Name Ginger. The Segway attracted extraordinary expectations before the company had established a repeatable market. The technology worked. The breadth of the vision made choosing a first customer harder.
That is the same moment an AI founder can reach after explaining a product perfectly.
The sentence that changes the room
The founder has covered the model, the workflow, the data, the interface and the technical advantage. Every question about how the product works has an answer.
Then someone asks where the first repeatable market is.
The room changes because the question moves from possibility to purchase. “Healthcare, logistics and financial services” names industries. It does not identify a market. “SMBs in Africa and Europe” describes a large population with different budgets, regulations, buying habits and reasons to act.
A usable answer sounds narrower.
It might be procurement teams at mid-sized distributors in Accra that still reconcile incomplete supplier records by hand. It might be small German agencies that lose several hours each week transferring client requests into delivery plans. The right segment depends on evidence, but the structure remains the same: a specific buyer, a recurring problem, an identifiable moment of urgency and a practical route into the account.
This is where strong product builders often hesitate. Narrowing the market feels like reducing the ambition. With limited runway, the opposite is usually true. A narrow market gives the team somewhere to learn.
Product clarity can hide market uncertainty
An AI demo can create a convincing sense of progress. The system takes messy input, produces a clean result and handles an edge case that would have consumed hours of manual work.
That proves capability. It leaves several commercial questions open.
Who feels the problem often enough to change behaviour? Who controls the budget? What event makes the problem urgent this month? Can the founder reach ten similar buyers without starting from zero each time? Does each customer need a different architecture, data source or approval process?
When those answers vary from one prospect to the next, the company may have several bespoke projects rather than one market. I have seen this tension across African, European and US opportunities. The same demo can earn interest in all three places while entering three completely different purchasing systems.
Geographic reach does not create repetition by itself. Neither does a signed pilot.
The harder test is what survives after the first sale. If the next customer needs a different problem statement, buyer, integration and pricing logic, the founder has learned something useful. They have not yet found the repeatable unit.
That is also why two credible requests can pull one team toward two companies. [Kojo’s two signed pilots](\/blog\/kojo-s-two-signed-pilots-six-months-of-runway-could-build-two-products-db9ae82b\/) looked like traction until the product consequences were placed side by side.
Find the repeated decision before the large category
I would leave the market-size slide alone for a moment and examine the last five serious conversations.
Write down who first described the pain, who would approve a purchase, what they do today and what happened immediately before they began looking for help. Use their actual words. “We need AI” tells you little. “The operations lead checks every supplier record before Friday’s approval meeting” gives you a workflow, an owner and a clock.
Then look for repetition.
Three buyers sharing an industry may still have unrelated problems. Three buyers in different industries may share the same operational trigger, budget owner and buying path. For an early AI product, that second group can be the more coherent market.
Next, remove the founder’s personal network from the acquisition story. If every opportunity depends on a former colleague, investor introduction or conference conversation, access remains founder-specific. Ask where the next ten comparable buyers already gather, which existing process exposes the problem, and who already has permission to discuss it with them.
This exercise can produce an uncomfortable answer: the first market may be smaller, less prestigious or closer to home than the market in the pitch deck. That is useful information while there is still runway to act on it.
Give the next meeting a testable answer
Segway’s engineering achievement did not automatically establish a mass market. The broad vision arrived before a narrow commercial foothold had proved how demand would repeat.
An AI founder can avoid that sequence by entering the next meeting with a claim that can fail.
Name one buyer. Name the repeated problem. Explain how the last three relevant conversations were reached. State what remained consistent and what still changed. Then describe the next test: five more conversations in the same segment, a paid workflow trial, or a second sale using the same product and buying path.
The answer does not need to sound enormous. It needs to show that the next customer will teach you more about the same company you are already building.
Comments
No comments yet.