Here is something that does not get said often enough in the B2B startup world.
Most ICP documents are fiction.
Not intentional fiction. The founders who write them are smart, thoughtful, and genuinely trying to answer the question. But the ICP is built in a conference room, on the basis of assumptions, hypothesis, and whatever the founding team remembers about the last customer they won. It is then treated as a strategic document for the next 18 months.
The result: marketing messages aimed at a buyer who does not quite exist, sales sequences targeting accounts that will never convert, and a GTM motion that feels active but produces nothing.
The Assumption Problem
When I ask a founding team to describe their ICP, I usually hear something like: “Series A SaaS companies in the US, 50-200 employees, facing X problem.”
That sounds specific. It is not.
“Facing X problem” is doing all the work in that sentence, and it is the part we have done the least research on. We assume the problem exists. We assume it is urgent. We assume the buyer experiences it the way we think they do.
The ICP becomes a container for wishful thinking dressed in the language of strategy.
What an ICP Actually Requires
A real ICP is not a demographic description. It is a behavioral and psychological profile built on evidence. That evidence comes from three places:
1. Deals you won, analyzed for actual patterns. Not why you think you won, what the buyer actually said they were worried about before they signed. The words they used in the sales call. The objections they raised. The person who was the internal champion and why they stuck their neck out.
2. Deals you lost, analyzed with equal rigor. Who said no and why. If you do not know why you lost deals, you do not understand your market well enough to build a reliable GTM.
3. The buyer’s calendar, not the buyer’s job title. A VP of Marketing at a Series A SaaS company has 14 things on their plate this quarter. Your product is one of them, and probably not the most urgent. Until you understand the other 13 things, you cannot write a message that breaks through.
Where the ICP Trap Closes on You
The trap closes in month 3 or 4 of the GTM execution. Outbound has been running for 10 weeks. Reply rates are below 2%. The marketing team is running A/B tests on subject lines. The founders are wondering if the channel is the problem.
The channel is not the problem.
The ICP is wrong, and because the ICP is wrong, every message built on top of it is wrong. Subject line testing on a fundamentally mismatched message is rearranging deck chairs.
This is the moment when Crown steps in. Not to optimize what exists, to rebuild the foundation. We go back to the win/loss data, re-interview three customers, and identify the actual pattern. In every case, the real ICP is narrower than the stated ICP, and the actual buying trigger is different from the assumed trigger.
The Three-Week Fix
You cannot rebuild your ICP in a workshop. You rebuild it with data, and data takes time to collect properly. But you can course-correct a broken ICP in three weeks if you move quickly.
Week 1: Pull your last 12 deals, won and lost. Interview three won customers and two lost prospects, targeting the economic buyer, not the champion. Ask one question: “What would have had to be true for you not to buy this?”
Week 2: Identify the pattern. There will be one or two answers that appear in every won deal and are absent from every lost deal. That is your ICP signal.
Week 3: Rewrite your top-of-funnel messaging around the pattern, not the assumption. Test the new message against the old in parallel. The difference in reply rate tells you whether you found the real ICP or are still guessing.
The Go-to-Market Truth Most Teams Avoid
Go-to-market and product positioning is not a one-time exercise. The ICP you write at Seed is not the ICP you execute against at Series A. Markets move. Buyer priorities shift. The problem that was urgent in 2024 may be table-stakes in 2026 because three competitors have now solved it.
Crown builds this re-evaluation into our quarterly operating rhythm with every client. Not because we are particularly clever, because the data demands it.
The startups that build durable pipeline are the ones willing to challenge their own assumptions before their competitors challenge them first.
FAQ
How often should a B2B startup revisit its ICP?
At minimum, once per quarter at Seed and Series A stage. Every significant product change, funding event, or market shift should trigger an ICP review. The ICP is a living document, not a founding artifact.
What is the difference between ICP and buyer persona?
ICP defines the account (company size, stage, industry, buying trigger). Buyer persona defines the individual inside that account (title, motivations, objections, decision-making process). You need both, in that order. Get the ICP wrong and every persona document built on top of it is also wrong.
Can Crown help rebuild an ICP mid-GTM?
Yes, this is one of the most common engagements we take. We bring a structured win/loss interview process, pattern analysis, and message reframing. Most teams see meaningful outbound improvement within 30 days of completing the rebuild.
Crown helps B2B startups find the real ICP, build the right message, and own the pipeline. See what that looks like in practice at /our-work/ or learn about our go-to-market and product positioning approach.
