Most scale-ups treat ICP as a founding decision – something you figure out early, write down in a strategy doc, and then execute against. And for a while, usually the first million or so, that approach works well enough. As you’re still close to every deal at this stage, you know your customers personally, and you can tell when something ‘feels off’.
But that won’t work forever. When you’ve got a team, a pipeline, and a growth target, your ICP is usually not something you’re actively interrogating. It’s just the filter your BDRs are using to build lists. By that point, most scale-ups aren’t running on a real ICP, but on the legacy memory of one.
The companies that successfully scale from £1 million to £10 million+ treat ICP not as a decision, but as a discipline. One that has to be revisited at every meaningful stage of the journey.
Getting that right is harder than it sounds, but getting it wrong is more expensive than most founders expect.
The most dangerous place to be: product-market fit with the wrong customer
Picture this – you’ve got customers paying, retention looks healthy, the product is being used, but when you try to scale and hand over to a sales team, push into new segments or raise another round, something breaks.
Natasha Ratanshi-Stein, the founder of Surfboard (acquired by Dialpad in 2024), describes this with uncomfortable precision in her Making The Grade episode. Before the acquisition, Surfboard hit £100K ARR three times. Each time, they pulled back because the product wasn’t ready, or the customers weren’t quite the right fit, or the signal wasn’t as clean as it looked.
“We had product-market fit with early customers, but it wasn’t the right ICP. When you do the math of how to get to a million, three million, nine million ARR with that, it just doesn’t add up.”
There’s a clear distinction between problem-market fit and product-market fit. Early customers, especially those who come in through founder relationships or a willingness to take a punt on something unproven, validate that the problem is real. They don’t validate that you’ve found the customer you can scale with.
According to Natasha, the signal you’re looking for isn’t just that customers are buying, but that you can displace a competitor at full price, without the founder in the room, without promising a feature that doesn’t exist yet, and with a clear articulation of the ROI. Until all of those are true, you haven’t actually found your ICP.
That distinction matters enormously when you’re deciding whether to hire salespeople, invest in outbound, or scale a GTM motion. Move too early, before the ICP is truly validated, and you’re just accelerating confusion.
Narrowing when every instinct tells you to widen
Most scale-ups facing the pressure of a growth target respond to ICP ambiguity by widening their definition of who they’re selling to – more segments, more personas and more use cases.
While you might think a bigger addressable market means more opportunity, the opposite is often the case.
Pete Crosby, founder of Revelesco and one of the most experienced revenue coaches in the UK startup ecosystem, sees this pattern repeatedly:
“I’m a real fan of narrowing uncomfortably small on an ICP to begin with. Sales hates it. It’s uncomfortable. It’s intuitively wrong, but it actually works.”
The reason it works is that clarity compounds. When your ICP is tight, your messaging sharpens, which means your BDRs know exactly who to call and what to say, and your AEs can qualify faster and more accurately. That all leads to better retention, case studies and referrals.
Paul Fifield, one of the UK’s most experienced GTM leaders, adds an operational dimension that most ICP conversations miss entirely. Defining the ICP is one thing, but protecting it is another, and it’s where most scale-ups quietly fall apart.
The failure mode he sees repeatedly is straightforward: a woolly ICP definition, a team of SDRs hired to drive outbound, and list-building delegated to the people least equipped to do it well. His answer is structural:
“You never let anyone – reps, sales leaders, or SDRs build their own lists. That should be done centrally, probably by someone in your RevOps team.”
The ICP, in other words, shouldn’t just live in a strategy document, but should govern the mechanics of how pipeline is built, at every level of the organisation.
Even once you have your list, you want to make sure you don’t collapse your ICP into a single buyer profile and blast the same message at everyone who fits it. What actually works, Paul argues, is going deep on the four or five distinct personas within your ICP and understanding how each of them experiences the problem you’re solving, and crafting messaging that speaks to each specifically.
That is the difference between an ICP that narrows your market and one that sharpens your execution across it.
ICP is what makes or breaks your US expansion
Nowhere is ICP clarity more consequential and commonly underestimated than in international expansion. And nowhere has the lesson been learned more painfully, and more repeatedly, than in the UK-to-US journey.
Natalie Johnson has done this journey twice. The first time with the strategy to hire a team in San Francisco with experienced salespeople, assuming the UK motion would travel across the pond.
This is such a common assumption, but it couldn’t be further from reality. Your UK customers are completely different to your US customers, and their buying behaviours and risk appetite are often worlds apart. What eventually worked for Natalie was doing the ICP work properly before committing to boots on the ground. Understanding not just who the customer is, but where they’re concentrated, what their buying process looks like, what a winning case study needs to say to move the next tranche of prospects.
“America is a big old place. You can spend a lot of time going in a very wrong direction if you’re not really clear about what you’re looking for.”
The narrower the market, she adds, the more important focus becomes. A single country with hundreds of millions of potential contacts and dozens of viable cities is not an argument for a broad ICP. It’s an argument for a tighter one.
What ongoing discipline actually looks like
None of this means ICP should be in constant flux. Chopping and changing your target customer in response to every difficult quarter isn’t the answer. The discipline isn’t perpetual reinvention; it’s structured interrogation at the right moments.
Those moments tend to cluster around the same inflexion points:
- When you’re building your first sales team
- When you’re entering a new geography
- When growth stalls after a period of momentum
- When you’re preparing to raise
Each of these is a forcing function for the question: is our ICP still right, or have we been executing against an assumption that’s quietly expired?
The companies that do this well tend to keep ICP ownership close to the top of the organisation, led by the CRO or the founder. They treat customer win-loss data not as a reporting exercise but as ICP input. They have a clear answer to what makes a customer winnable, retainable, and expandable, and they update that answer as the evidence changes.
Sam Jacobs, founder of Pavilion, advises founders building towards an exit to stay focused on ICP long past the point where it feels necessary:
“If you’re sub a hundred million and you’re pursuing more than three different personas, there’s some reflection that needs to be done.”
The scale-ups that reach ten million with a clean, well-validated, actively maintained ICP have a fundamentally different business than the ones that got there by saying ‘yes’ to everyone. They have better unit economics, better retention, and a much clearer story to tell investors, new hires, and the market.
And most importantly, the ICP conversation never really ends. The companies that treat it as finished are the ones that find out that it wasn’t…usually at the worst times imaginable.
—
These insights are drawn from Season 2 of our Making The Grade podcast hosted by Tom Glason. Listen to more episodes here.
