ICP checklist: how to check your ideal customer profile is well defined

Almost every B2B company that has ever run a strategy workshop has an ideal customer profile. Far fewer have one that is good for anything beyond a slide. The difference is not the length of the document. It is whether you can build a list of companies from it and write a message that lands on a real problem.

This article assumes you already have a profile, so we do not explain what an ICP is or how to build one from scratch. We cover that in our guide on how to build an ideal customer profile. Here you get five tests to find out whether your profile will survive contact with an outbound campaign.

In most B2B deals the person who feels the pain and the person who signs the contract are different people. A typical profile describes only the first, because that is the person who talks loudest about the problem.

The consequence is practical. A message to the person with the problem gets read, understood and then stalls, because they do not control the budget. A message to the person with the budget sounds abstract, because the problem it describes is not part of their working day.

The test: for every role named in your profile, write down what frustrates them and what they need to see before they say yes. An empty box, or a single role in the whole document, means the test is failed. There is often a third role as well: the person who uses the solution every day and can sink it, even though they have no formal say.

We agree these roles with the client at the kick-off workshop, before any prospect list is built. The whole process, from profile to booked meeting, is on our how it works page.

A profile built at a whiteboard describes the market you would like to serve. A profile built on closed deals describes the market that actually buys from you. They are rarely the same market.

This test is simple and a little uncomfortable. Take the customers you won in the last 12 months and mark each one as inside or outside the profile. Then do the same for those who churned or turned out to be unprofitable. If most of your closed deals fall outside the profile, the document describes someone other than your customer.

While you are at it, check what the industry and size criteria rest on. "Mid-sized manufacturers interested in growth" looks like a profile but tells you nothing. "Component manufacturers with a few dozen staff, where the operations director signs off" comes from real contracts and can be verified.

A profile from which you cannot list company names is not usable for outbound.

What matters is less the result than the point at which you get stuck. If after a dozen or so names you start adding companies that are "roughly similar", the problem is in the criteria: they are too vague or cannot be checked. A criterion is workable when you can verify it from publicly available information about the company, such as industry, headcount, markets served or the role of the decision-maker. "Companies that value quality" does not meet that standard.

This is exactly the operation you perform when building a prospect list. We build ours from public company and professional data rather than buying ready-made lists, so the quality of a list starts in the profile, not in the tool.

A profile with no exclusion criteria is not a profile but a wish list. Inclusion criteria say who you want on the list. Exclusion criteria protect your sales calendar from conversations with no chance of closing.

The check is short: write down five types of company you will deliberately not contact, and give a reason for each. If the list stops at one item, the profile describes a hope about the market.

Typical exclusions cover companies too small for the project to make economic sense and organisations that solve the same problem in-house. You also exclude markets whose buying cycle is longer than you can afford to fund. A separate category is segments that bought once and left quickly.

The cost of having no exclusions is measurable: the list grows, meetings appear, and sales still do not close, because the conversations are with companies outside your market.

If marketing and sales describe the ideal customer differently, there are two profiles in the business and neither binds anyone. Ask both sides separately to describe the profile in five points and lay the answers side by side. The gap is usually not about industry but about company size and the role of the decision-maker, which is what shapes the list.

The second part of this test is the date. Check when the profile was last updated, then write down what has changed in your offer since. A new service, a price change, a move into another market or a run of cancellations in one segment all shift the profile, so a two-year-old document describes a company you no longer run.

One last distinction that is easy to forget: the profile tells you who to look for, while buying signals decide when to send. That is a topic for another article.

An ideal customer profile is rarely completely wrong. More often it is simply untested: nobody has listed 50 companies from it, nobody has set it against closed-won deals and nobody has asked who we deliberately do not invite into a conversation. The five tests take a few hours and decide how well every later message lands. If you would like to put your profile up against our experience, get in touch.