Ideal Customer Profile: A Worksheet From Your Won Deals

An ideal customer profile (ICP) is a written description of the company most likely to buy, stay, and grow with you, built from firmographic fit (industry, size, revenue) and the situational triggers that make now the right time to buy. It describes the company, not the person.

This guide gives you a worksheet built from your own closed-won deals, a short table on how an ICP differs from a buyer persona, and the fields to add so your CRM can filter for ICP fit instead of relying on memory. For how a qualified lead becomes a deal once it matches, see lead qualification frameworks.

What is an ideal customer profile (ICP)?

An ICP works at the company level, not the person level. It combines two kinds of information: firmographic fit (traits that stay roughly stable, like industry, employee count, revenue band, and the tools already in place) and situational triggers, the events that make a company ready to buy right now, such as a new location, a headcount jump, a compliance deadline, or a tool nobody trusts anymore.

Most ICP definitions stop at firmographics, which is why they end up describing thousands of accounts that will never buy this year. Adding the trigger is what turns a list of look-alike companies into a list of companies actually in motion.

How is an ICP different from a buyer persona?

An ICP and a buyer persona answer different questions, and folding them into one document is one of the most common ways an ICP stops being useful. Even inside a company that matches your ICP on paper, a B2B purchase is rarely decided by one person: several people typically weigh in before a deal moves forward, which is exactly why the company-level fit an ICP describes still needs the person-level detail a persona provides.

Keep both, but keep them separate: one page names the accounts worth chasing, the other explains how to talk to the people inside them.

How do you build an ICP from your last 20 won deals?

Skip the workshop where everyone guesses at the “ideal” customer from memory. The most reliable ICP comes from deals you already won, not deals you lost and not deals still open, since neither has proven it can close and stay.

1. Pull your last 20 closed-won deals straight from the CRM (fewer is fine if you haven’t closed 20 yet: ten is enough to start a pattern). Skip pilots, friends-and-family deals, and anything won on a discount so steep it wouldn’t repeat.
2. Record five facts for each one: industry, employee count band, the event that triggered the purchase, deal size, and how long the deal took to close.
3. Sort the five facts into two buckets. Firmographic fit is what stayed the same across most rows: industry, size band, the tools already in place. Situational triggers are what changed right before the deal opened: a new location, a leadership change, a deadline, a tool that stopped working.
4. Find the pattern that shows up across most of the list, not the one outlier account that closed for reasons that won’t repeat. If your 20 deals split cleanly into two different patterns, you likely have two ICPs, not one: write them up separately rather than averaging them into a vague middle.
5. Write the pattern as fields, not prose. A sentence in a slide deck can’t be filtered on; a picklist can.

What does a filled-in ICP worksheet look like?

Example: a hypothetical 12-rep B2B SaaS company in Cincinnati, Ohio, sells scheduling software. Reviewing its last 20 closed-won deals turns up this pattern:

• Firmographic fit: field-service companies (HVAC, landscaping, and electrical contractors) with 25 to 150 employees and no dedicated IT staff.
• Situational trigger: the company just added a second location or a second crew, and the spreadsheet or aging scheduler that worked for one crew stopped working for two.
• Deal size and cycle: consistent, mid-range deal sizes and a short sales cycle, because the trigger already created urgency before the first call.

The resulting ICP is specific enough to filter on: field-service companies, 25–150 employees, expanding to a second location or crew. A “reason for switching” field, filled in on every new lead, tells a rep in seconds whether the account matches, without anyone re-reading the deal history to check.

How do you turn ICP criteria into fields you can filter on?

An ICP that lives only in a slide deck gets re-derived from memory every quarter. Once the pattern is clear, add it to the CRM as fields on the lead and deal record, not as a note:

• An industry or vertical picklist, matching the language your own deals actually used, not a generic list borrowed from a data vendor.
• An employee band field, so a filter can separate a 30-person company from a 3,000-person one instantly.
• A trigger type field (new location, leadership change, compliance deadline, tool switch) filled in from the first qualifying call.

Once those fields exist, a saved list filtered on all three turns “which of our leads actually match the ICP” from a judgment call into a query anyone on the team can run.

What mistakes weaken an ICP?

• Building it from every deal, not just the wins. Lost and churned deals belong in a separate analysis of what to avoid: mixing them in blurs the pattern you’re trying to find.
• Firmographics only, no trigger. A company that matches on paper but has no reason to buy right now still won’t convert; the trigger is what makes the list actionable today.
• Treating the ICP as permanent. A product change, a new pricing tier, or a shift upmarket changes who wins, so revisit the worksheet when the product does, not on a fixed calendar alone.
• Averaging two patterns into one vague ICP. If the 20 deals split into two distinct groups, write two ICPs rather than a blend that fits neither well.
• Confusing the ICP with the persona. One document trying to do both usually does neither well: see the table above.

How do you know whether the ICP is working?

Track win rate, deal size, and sales-cycle length for ICP-fit deals against everything else, using the same fields you just added. If ICP-fit deals close faster, close more often, or hold their price better than average, the pattern is doing its job. If the numbers look the same either way, the fields probably need another pass at step four: the pattern hasn’t actually been found yet.

Run this as a saved report against won deals on a fixed cadence (quarterly is reasonable for most B2B teams) rather than a one-time check nobody revisits. For where the ICP fits into building pipeline in the first place, see how B2B lead generation feeds the pipeline.

How does Senitix CRM help you apply an ICP?

An ICP is only useful if the fields it depends on live where a rep is already working. Senitix CRM lets you add custom fields for industry, employee band, or trigger type directly to the lead and deal record, and save a list filtered on all three at once instead of relying on someone remembering the pattern.

Reports built on won deals show win rate and cycle length by segment, so “is the ICP working” has an answer instead of a guess. See what a lead and deal record hold in the reporting section of the feature catalog, check how your own accounts and industries line up on the industries page, then use your ICP to build a target account list. Compare plans and start on Free to build your first ICP fields.

Frequently asked questions

How many deals do I need to build an ICP?

Twenty closed-won deals is enough to see a repeating pattern rather than an outlier. Ten is a workable starting point for a newer team; fewer than that usually isn’t enough to tell a real pattern from a coincidence, so treat an ICP built on five or six deals as a draft to revisit.

What if my product serves more than one type of customer?

Build a separate ICP for each distinct pattern rather than averaging them together. Two clear, narrow ICPs, each with its own firmographic fit and trigger, are more useful to a sales team than one broad one that half-describes both.

Should an ICP include company size, industry, or both?

Both, plus the situational trigger. Industry and size alone describe a company that could buy someday; the trigger is what tells a rep the company is worth calling this week rather than filed away for later. The worksheet above pulls all three from the same twenty deals, so building the ICP in one pass is what keeps them from drifting apart.

How often should you update an ICP?

Whenever the product, pricing, or target market shifts (a new tier, a move upmarket, a new integration that opens a vertical) rather than on a fixed calendar alone. Even without a change, revisit the worksheet quarterly to confirm the pattern still holds against your most recent wins.

Is an ICP the same as a target account list?

No. An ICP is the criteria; a target account list is the result of applying those criteria to real companies. Build the ICP first, then use it to name the accounts: see how to build one in building a target account list.