She asked one question before we got to anything else.

“Do you have a clear ICP?”

She had built and sold two companies. She didn’t ask about the product roadmap. She didn’t ask about the team, the funding, or the tech stack. She went straight to the ICP. The Ideal Customer Profile: the specific description of the exact type of customer your product is built for.

We had an answer. It just wasn’t a good one.

“Parents who want their children to do better in math.”

She nodded slowly. The kind of nod that isn’t agreement.

“That’s not an ICP,” she said. “That’s a category.”

The difference matters more than most founders realize.

A category is everyone who could theoretically want what you’re building.

An ICP is the specific person who will pay for it today, refer it tomorrow, and forgive you when you get something wrong next month.

CB Insights analyzed 431 failed VC-backed companies and found that 43% failed due to poor product-market fit. Not bad execution. Not bad timing. Not bad people. Wrong customer.

The sea is wide. Most startups don’t drown because they sailed badly. They drowned because they had no idea where they were going.

So we did the exercise.

We listed every possible ICP on a whiteboard. The international school parent pays London rates for IGCSE tutoring. The middle-income family in Kuala Lumpur watching their child fall behind in SPM math. The tuition center owner is looking to run more classes with fewer teachers. The homeschooling parent. The working single mother. The anxious father of a Grade 9 student in Manila.

The list was long. That was the problem.

We rated each one on three axes.

Market size: how many of these customers actually exist?

Pain: how urgently do they feel it?

Ease to convert: how quickly can we get them from awareness to paying?

A broad ICP scores well on market size and almost nowhere else. The pain is diffuse. The conversion path is unclear. You spend your resources trying to speak to everyone and end up resonating with no one.

The narrower the ICP, the sharper the signal.

Defining an ICP is the cartography work of early-stage building.

You are in the sea without a map. Product-market fit is somewhere out there, but the ocean is bigger than your runway. You cannot sail everywhere.

Every week you spend talking to the wrong customer is a week you did not spend with the right one. Every feature you build for a diffuse audience is a feature that solves nothing completely.

The map does not tell you exactly where the treasure is. It tells you where it is most likely to be, based on what you already know about the water, the currents, and the shape of the land. That is what a scored ICP gives you. Not certainty. Direction.

The instinct against narrowing is understandable. It feels like leaving money on the table.

It isn’t.

Nearly half of all startups have less than 10% of their ICP as actual customers. The problem is rarely that they built the wrong product. It is that they never decided who they were building it for. They kept the whiteboard full. Every row felt like optionality. Every row was actually noise.

We left that conversation with a shorter whiteboard and a clearer target.

The narrowing felt like a loss at first.

You can always expand the map once you know where you are.