How Do You Know Your Ideal Customer Profile Is Too Broad?

How do you know your ideal customer profile is too broad?

The clearest sign is that you cannot name a company it excludes. If your ICP would accept almost any business that could theoretically buy from you, it is a market description rather than a profile. A real ICP makes some perfectly good companies ineligible, and saying that out loud should feel slightly uncomfortable.

The second sign is that your team cannot use it. Ask a salesperson whether a specific inbound lead fits, and if they have to think for more than a few seconds, the profile is not doing its job. An ICP that requires interpretation is a document, not a decision tool.

Most of the ICPs I am shown are too broad, and almost none of them are wrong. That is the trap. They accurately describe companies that could buy, which feels like success, while failing to describe the companies you should chase, which is the only thing the exercise was for.

Why does a broad ICP feel safe?

Because narrowing feels like turning off revenue, and nobody wants to be the person who removed half the market from the plan. A broad profile keeps every option open, offends nobody in the leadership meeting, and cannot be proven wrong in the short term. It is the path of least resistance.

There is also a founder-specific version of this. Early on you take whoever pays, which is correct. The mistake is carrying that improvisation forward as strategy. The set of customers who happened to say yes in year one is not the same as the set of customers you should be built around in year three, and confusing the two is extremely common.

The cost of a broad profile is real but delayed, which is why it survives. You do not feel it this quarter. You feel it four quarters later, when your messaging has gone generic, your pipeline is full of slow deals, and nobody can explain why the win rate keeps drifting down while activity stays high.

What does a too-broad ICP do to your messaging?

It forces every sentence up a level of abstraction. When your page has to speak to five different kinds of buyer, the only words that work for all of them are the vague ones. That is how a homepage ends up promising to help teams work smarter, which is a sentence that has never persuaded anybody of anything.

Watch what happens to your proof as well. Specific proof only works on a specific reader. If your ICP spans agencies, hospitals and logistics companies, your case studies have to be rotated or genericised, and a genericised case study persuades nobody because the reader cannot see themselves in it.

This is usually diagnosed as a copywriting problem and it is not. I have rewritten plenty of pages where the real fault was upstream, in a profile so wide that no honest sentence could be written. If your writers keep producing bland copy after three attempts, look at the brief, not the writers. The same confusion often shows up as a positioning issue, which I wrote about in the piece on knowing your positioning is wrong.

What does it do to your sales conversations?

It makes discovery longer and shallower. When you do not know what kind of company you are speaking to, every call starts from zero. You ask general questions, get general answers, and arrive at a proposal that could have been written for anybody. The buyer notices, even if they do not say so.

It also makes internal agreement harder on the buyer's side, which is where deals actually die. Gartner's sales survey found that 74 percent of B2B buyer teams demonstrate unhealthy conflict during the decision process. You cannot remove that conflict, but you can arrive with language that already fits how that specific kind of company argues internally. Generic language leaves the buyer to do that translation themselves, and most of them will not bother.

The symptom to watch for is the proposal that needs a long explanatory call attached to it. If your champion cannot forward your document and have it understood without you in the room, you have handed them something written for a market rather than for their company. That is where deals stall, and it looks like procurement when it is really positioning. I went into the stalling pattern separately in the article on deals stalling at procurement.

How does a broad ICP show up in your numbers?

Look at win rate by segment rather than in aggregate. A broad profile almost always hides one segment that wins often and several that barely convert. The aggregate looks mediocre and stable, which invites the conclusion that sales needs to try harder, when the real answer is that most of the pipeline should never have entered it.

The second number is sales cycle length by segment. Deals outside your true profile take longer, because more has to be explained and more people need convincing. If one group closes in six weeks and another takes five months, you are running two different businesses and reporting them as one.

The third is churn by acquisition source and segment, which is the one that settles arguments. Customers who were a poor fit tend to leave, and they leave after somebody has already booked the revenue. When a founder resists narrowing, the churn table is usually the exhibit that changes their mind, because it converts an abstract debate into money that already walked out.

How narrow is narrow enough?

Narrow enough that you can name ten real companies that fit and explain why each one qualifies. If you cannot list ten by name, the profile is too abstract. If you can list a thousand, it is too broad. Somewhere between those is a profile you can actually build a go-to-market motion around.

The useful dimensions are rarely the ones in the template. Industry and headcount are easy to write down and usually weak predictors. What predicts fit is closer to situation: what the company has already tried, what just changed internally, what system they currently run, what the buyer is personally accountable for. Those are harder to collect and much more useful.

My own rule of thumb is that a good profile contains at least one qualifier that would make a salesperson wince, because it disqualifies deals they would otherwise work. If nothing in the document costs you anything, nothing in it is doing any work.

What do you actually write down?

Three things: who qualifies, who explicitly does not, and the trigger that makes them ready now. The exclusions matter most, because they are the part people skip. A profile without a disqualification list is just a description of everyone you would happily take money from.

Keep it to one page and write it in the language your team already uses. Not personas with names and stock photographs, but the sentences a salesperson could say on a call. If it has to be translated before it is usable, it will not be used, and an unused ICP is worse than none because it creates the illusion of alignment.

Include the trigger explicitly. A company that fits your profile perfectly but has no reason to act this quarter is not a prospect, it is a future prospect, and treating the two the same is how pipeline forecasts become fiction. Fit and timing are separate tests and both have to pass.

How do you narrow without losing revenue you already have?

You narrow where you spend, not where you sell. Existing customers stay customers. Inbound leads outside the profile still get answered. What changes is where the marketing budget, the content calendar and the outbound effort point, and that can be redirected without cancelling anything.

This distinction defuses most of the resistance. Nobody is proposing to turn away a good deal that arrives. The proposal is that you stop spending money trying to manufacture deals in segments that convert badly, and move that spend to the segment that already converts well. Framed that way, narrowing is a reallocation rather than a sacrifice.

Give it a full sales cycle before you judge it, and decide the review date in advance. Narrowing usually makes the top of the funnel smaller before it makes the bottom bigger, and a team that panics in week six will reverse the decision right before it starts working. The same discipline applies to channels, which I argued in the piece on how long to give a channel.

When is a broad ICP actually the right answer?

When you genuinely do not know yet. A company in its first year with twenty customers does not have enough evidence to narrow responsibly, and premature narrowing based on three data points is its own failure mode. In that situation, breadth is a deliberate search strategy rather than an avoidance of the question.

The condition is that you say so out loud and set an end date. Broad because the evidence is not in yet is a legitimate position. Broad because narrowing is uncomfortable is not, and the two look identical from the outside unless somebody writes down which one is true.

There are also genuinely horizontal products where the buyer role matters more than the company type. Even then, the profile narrows on something, usually on the situation the buyer is in. Horizontal is not the same as undefined, and the products that get this right are usually extremely specific about the moment they are bought.

What should you do next?

Pull your last twenty closed-won deals and your last twenty closed-lost ones, and look for what the wins share that the losses do not. Not industry, but situation. That comparison usually takes an hour and produces a sharper profile than any workshop.

Then write the one-page version, including the disqualification list, and read it to one salesperson. If they can use it to judge a live lead in under ten seconds, it is finished. If they hesitate, it is still too abstract and needs another pass.

If you are looking at a pipeline full of deals that all feel slightly wrong and cannot work out what they have in common, reach out. Sorting the wins from the noise is usually a short exercise, and the answer is often already sitting in your own closed-won list.