Most early-stage founders have an ICP problem.
The customers exist. A few of them are genuinely great: they use the product, pay on time, refer people, and don’t file a support ticket every Tuesday. But when you ask the founder how those customers arrived, there’s not consistent answer. There’s no pattern.
Your best customers are already telling you something
The mistake is treating early customer acquisition as noise: random, unrepeatable, not worth examining. It isn’t. The fact that five people decided your product was worth paying for, through whatever combination of timing, network, and luck, is signal.
The question isn’t “how do we find more customers.” It’s “what made these specific customers the right ones, and how do we find more of them on purpose?”
The interview you should already be having
Before you write a targeting framework or fill out an ICP template, talk to your best customers. Don’t just send a survey. An actual conversation: 30 minutes, recorded. These conversations also give you the raw material for mapping your customer journey — the same trigger moments and objection language you’ll use everywhere else.
Ask them three things:
- What were you trying to solve when you found us?
- What were you using before, and why wasn’t it enough?
- When you realised it was working, what specifically had changed?
That last question is the one most people skip. It gets you the outcome language: the words your customer uses to describe success, not the features they list when you ask what they like about the product.
What you’re looking for in the answers
You’re not looking for the things they say they want. You’re looking for the situation they were in when they started looking.
There’s a difference between “I needed better analytics” and “I was going into a board meeting and had no idea whether the last quarter’s campaigns had done anything.” The first is a feature request. The second is a moment: a specific pressure, a specific deadline, a specific feeling. That’s the thing worth targeting.
Write down the situation, not the feature. Do it for all three interviews. Look for overlap.
If two out of three customers described a version of the same pressure before they found you, you have your ICP trigger. That trigger: the thing that made them start looking: is what your messaging needs to speak to.
The ICP document that actually works
Forget the persona template with a photo of “Marcus, VP of Marketing, 38, drives a Tesla.” That document doesn’t help anyone write copy or pick a channel.
A useful ICP has four parts:
Who they are. Industry, company size, role. Be narrow. “B2B tech companies” isn’t an ICP. “Seed-stage B2B tech founders with a few dozen paying customers who are preparing for a Series A conversation” is closer.
The trigger. What happened in their business, in their role, in their week, that made them start looking for a solution? The more specific you can make this, the more useful it is.
What they’ve already tried. If they’re looking for you, they’ve usually already ruled something out. Understanding why the alternatives didn’t work tells you exactly what to promise and what not to.
What a good outcome looks like. What does their situation look like six months from now if this works? This is the thing your content should be describing.
The test
Write your ICP down in one paragraph. Then send it to your three best customers and ask: does this sound like you?
If they say yes, you have an ICP. If they say “kind of, but…”: that “but” is the most useful feedback you’ll get this quarter. Don’t ignore it.
And if you’d rather not wait to talk to customers: look at the ones you already have. The pattern is already there. You just haven’t looked at it yet.
Once the ICP is clear, auditing your positioning is the next step — because a tight ICP without matching positioning still won’t convert. From there, a messaging guide is what turns the positioning into sentences your team can actually use. And if you’re wondering how a content strategy fits on top, that’s where it starts making sense.