What Is an Ideal Customer Profile? Definition and Worksheet
A plain definition of an ideal customer profile, what belongs in one and what does not, a one-page worksheet with a disqualifier column, the difference between fit and timing, and a rule for whether you need one profile or two.
An ideal customer profile (ICP) is a written description of the type of company that gets the most value from what you sell, buys it with the least friction, and stays. It describes a company, not a person: industry, size, location, tools in use, and the problem that makes them buy. The short answer to "what is an ideal customer profile" is a fit filter. It tells you who is worth your time, not who is ready to buy this month; that second question is covered below.
Why does an ideal customer profile matter?
Because outbound time is the scarcest thing a small team has, and most of the market is not buying. Research from the Ehrenberg-Bass Institute with Professor John Dawes puts the share of B2B buyers in market at any given time at about 5%, with the other 95% months or years away from a purchase (Ehrenberg-Bass Institute). If you contact companies that could never be good customers, you spend that scarce time twice: once on the message, and again on the deal that drags or churns.
A labelled example. Suppose a four-person agency sells HubSpot onboarding. Without a profile, their list holds every company using HubSpot: a solo consultant, a hospital group, a Series B software company. The consultant cannot pay, the hospital takes a year to sign anything, and the software company buys in a week and refers two friends. The profile is the sentence "software companies with a small sales team that bought HubSpot in the last year", and the list shrinks to companies that behave like the third one.
What goes into an ideal customer profile?
A useful profile fits on one page and has five parts. Most templates stop after the first two.
| Part | What to write | Example (hypothetical agency above) |
|---|---|---|
| Firmographics | Industry, company size, location, revenue band or funding stage | B2B software, 11 to 50 staff, EU or UK |
| Situation | The tool, change or problem that creates the need | Bought HubSpot in the last year; pipeline still lives in spreadsheets |
| Buying setup | Who decides, how long it takes, how they pay | Founder or head of sales decides; signs within two weeks; pays by card |
| Evidence of fit | Where this type has already won for you | Three of the last five closed deals; shortest sales cycle; zero churn so far |
| Disqualifiers | Traits that look like a fit but end badly | Agencies (they resell you), enterprise (procurement), companies with no sales hire yet |
The disqualifier row saves the most time, and almost every published template leaves it out. Write it from your lost and churned deals, not from guesswork.
How do you create an ideal customer profile?
Four steps; an afternoon with a dozen customers, an hour with three.
- Sort your customers by how good they were for you. Not by revenue alone. Rank on how fast they closed, how little support they needed, whether they renewed and whether they referred anyone. The top third is your raw material.
- Write down what the top third shares. Fill the firmographics and situation rows from your CRM or your memory. If you cannot see a pattern, you have a customer list rather than a profile. That is fine; be deliberately narrow for the next ten deals so a pattern can form.
- Ask three of them one question. "What was happening in the company the month you decided to look for this?" The answer is the situation row, in the customer's words, and it is usually more specific than anything you would have guessed.
- Write the disqualifiers from the bottom third. Who ghosted after a demo, who churned in month two, who asked for a discount before the first call. Name the trait they share, not the companies.
Then write the whole thing as one sentence you can say out loud, in the format Clearbit uses in its template examples: "B2B software companies with 11 to 50 staff in the EU that bought HubSpot in the last year and have no RevOps hire" (Clearbit). If the sentence has more than three qualifiers, cut the weakest.
Ideal customer profile vs buyer persona vs buying signal
All three describe "who to sell to", and they answer different questions.
| Ideal customer profile | Buyer persona | Buying signal | |
|---|---|---|---|
| Describes | A type of company | A role inside that company | An action by a specific person or company |
| Answers | Is this account worth pursuing? | Who do I talk to there, and what do they care about? | Is this account deciding now? |
| Built from | Your best and worst customers | Interviews with the people who bought | What you observe in public or in a conversation |
| Changes | Yearly, or when the product changes | When the buying committee changes | Daily |
| Example | B2B software, 11 to 50 staff, uses HubSpot | Head of sales, measured on pipeline, hates admin | That head of sales posts "our CRM data is a mess" on LinkedIn |
The persona matters because the room is crowded. Gartner's survey of 632 B2B buyers, run in August and September 2024, found buying groups ranging from five to 16 people across as many as four functions (Gartner, 7 May 2025). Forrester's 2026 buyer research puts the typical buying decision at 13 internal stakeholders and nine external influencers (Forrester, 21 January 2026). An ICP gets you to the right building; the persona tells you which door; the signal tells you whether anyone is home.
The profile without signals gives you a correct list with no order. Signals without a profile give you a timely list full of companies that will never buy from you. If you are new to the second half, the lesson on what a buying signal is covers it.
Where does an ideal customer profile mislead?
Four ways, all common in companies under 200 people.
Built from too few customers. Three customers that happen to be fintech companies do not make fintech your market. Treat an early profile as a hypothesis and write the date on it.
Built from the loudest customers. The accounts that email most are not always the ones that pay most or stay longest. Rank on the four measures in step one, not on who you remember.
Built once. A profile written at launch describes the people who tolerated a launch product. After a year the product and the buyer have both moved. Reread it every quarter and rewrite it when the top third changes.
Used as a timing tool. This is the expensive one. A company that matches every row can still be two years from buying, which is what the Ehrenberg-Bass figure above means in practice. The profile says "worth a conversation", never "worth a conversation today".
How to use an ideal customer profile this week
Three actions, in order.
- Fill the one-page worksheet above. Do it from memory if your CRM is thin; the point is to have something to be wrong about. Read the disqualifier row out loud to whoever does your outbound.
- Rebuild one list through the filter. Take the list you were going to work this week and strike every company that fails a firmographic row or hits a disqualifier. Keep the removed names in a second tab; if any of them buys next quarter, the profile needs editing. For the right person inside the companies that remain, the guide to finding decision makers on LinkedIn gives the exact filters.
- Add timing on top. Watch the companies that pass for one public action: a post about the problem in your situation row, a comment under a competitor's announcement, a new hire in the role that usually buys. Intentsly's LinkedIn signals do this step continuously and score every person found against the ideal customer profile you give it, so the list is sorted by fit before you open it. When we launched, we aimed our own signals at a narrow profile, founders complaining about cold-outreach burnout, and booked 14 qualified demos in 14 days with zero manual outreach. The narrow profile is what made every reply worth taking.
Related lessons
This lesson sits alongside what a buying signal is, which covers the timing half of the pair, and what intent data is, which covers what happens when signals are collected and scored over time. The Learn index lists new lessons as they are published, including trigger events and lead scoring, both of which assume you have a profile to score against.
If you want the timing half in a form you can use tomorrow, the free Buying Signal Playbook lists twelve signals worth chasing and eight to ignore, with the LinkedIn trigger and a script for each. Fill the worksheet first; the playbook is far more useful pointed at a narrow profile than at the whole market.
FAQ
What is an example of an ideal customer profile?
A one-sentence version: "B2B software companies with 11 to 50 staff in the EU that bought HubSpot in the last year and have no RevOps hire." It names an industry, a size band, a location, a situation and a disqualifier. A full profile adds who decides, how long signing takes, which past customers fit, and which traits predict a bad deal.
What is the difference between an ideal customer profile and a buyer persona?
An ideal customer profile describes the type of company worth pursuing: industry, size, situation, disqualifiers. A buyer persona describes a role inside that company, such as a head of sales, and what that person is measured on and worried about. You use the profile to choose accounts and the persona to choose who to contact and what to say.
How many ideal customer profiles should a company have?
One per product line or pricing tier, and ideally one in total if you have fewer than ten customers. A second profile is justified when a group of good customers differs from the first on at least two rows of the worksheet and buys for a different reason. If the only difference is size, you have one profile with a wide size band.
How often should you update an ideal customer profile?
Reread it quarterly and rewrite it when the top third of your customers no longer matches it, or when the product changes what problem it solves. Early-stage companies should expect to rewrite it after every ten or so deals, because the first version is a guess about who will tolerate a new product.
Is an ideal customer profile a sales or a marketing document?
Both, and it only works when they share the same one. Marketing uses it to decide which companies to reach and what problem to lead with; sales uses it to qualify or disqualify quickly and to order the week's outreach. If the two teams keep separate versions, the leads marketing passes over will be the ones sales rejects.
