How to Create an Ideal Customer Profile: A Step-by-Step Guide

Most B2B companies have an ICP document. Very few have an ICP that anyone uses.

The gap between the two is almost always the same: the document was built from a brainstorm, not from data. A marketing leader assembled the team, someone wrote "Series B SaaS companies with 50–200 employees" on a whiteboard, and everyone nodded. The document was formatted, filed, and forgotten — because it described who the team hoped to sell to, not who actually bought, stayed, and expanded.

An ideal customer profile built from evidence works differently. It describes a specific set of accounts with such precision that your SDR can build a list from it, your PMM can test messaging against it, and your CS team can flag early which accounts look like good fits before the first QBR.

This guide walks through how to build that profile from scratch — or rebuild one that isn't working.

40%

---

Step 1: Pull your best existing customers — and define what "best" means

Before you build an ICP, you need a definition of success. "Best customer" is not the largest logo or the one your CEO likes to mention in board meetings. It is the account that:

• Closed fastest relative to deal size
• Requires the least ongoing support
• Has expanded revenue over time
• Has the highest likelihood of renewal
• Refers other accounts unprompted

What to do:

1. Export your full customer list from your CRM.
2. Score every account on the five criteria above. You do not need a formula — a simple 1–3 rating per criterion is enough to separate the top 20% from the rest.
3. Identify your top 15–20 accounts by composite score. These are your ICP exemplars.

Output from this step: A ranked list of 15-20 exemplar accounts. A scoring rubric you can reuse every quarter.

---

Step 2: Extract firmographic patterns from your exemplars

Firmographics are the observable, structural attributes of a company -- the facts you can find in a database without talking to anyone. They are the starting filter for ICP, not the end of it, but they are what makes the profile actionable for prospecting.

What to extract for each exemplar:

• Employee count at time of first purchase (not current -- companies change)
• Annual revenue at time of purchase (estimated from funding data or public sources)
• Funding stage (seed, Series A-B-C, bootstrapped, PE-backed, public)
• Industry vertical (be specific -- "SaaS" is not a vertical; "HR tech SaaS" is)
• Geography (HQ country, primary market served)
• Ownership structure (VC-backed, founder-led, PE-backed, public subsidiary)
• Technology stack (what other tools do they use -- pull from BuiltWith, LinkedIn, job postings)

Output from this step: A firmographic data table — one row per exemplar, one column per attribute.

---

Step 3: Identify behavioral and situational triggers

Firmographics describe what a company looks like. Triggers describe why they buy now. The trigger is the event or condition that moved them from "we should probably think about this" to "we need to solve this by end of quarter."

Without triggers, your ICP is a list of companies who might eventually buy. With triggers, it is a list of companies who are likely to buy this quarter.

Firmographics find the pond. Triggers find the fish that are biting.

What to do:

1. For each exemplar account, ask your AE: What was happening at that company in the 60-90 days before they first engaged with us?
2. Cluster the answers. Common trigger categories for B2B SaaS:
   • Growth event: headcount doubled, new market entered, recent funding round
   • Leadership change: new CMO/VP Marketing/CRO hired in last 6 months
   • Process breakdown: existing tool or manual process visibly failing at scale
   • Competitive pressure: a competitor launched something that exposed a gap
   • Deadline: board review, rebrand, product launch, fiscal year planning
3. Identify your top two or three trigger patterns. These become the "signal" your SDR team watches for in prospecting.

Output from this step: A list of 2–3 primary buying triggers with a description of each.

---

Step 4: Map the buying committee — roles, not just titles

B2B purchases are rarely made by one person. A typical ICP has a buyer map: the economic buyer who signs, the champion who pushes internally, the end user who sets requirements, and the blocker who has to be convinced.

Getting this wrong wastes sales cycles — your team sells to the champion for three months, then loses when the CFO sees the contract for the first time.

What to do:

1. For each exemplar, identify every person who was involved in the purchase decision: who was on calls, who reviewed contracts, who had final approval.
2. Record their title, department, seniority level, and their primary concern in the deal.
3. Identify patterns:
   • Who is the economic buyer (signs the contract)?
   • Who is the champion (advocates internally)?
   • Who is the end user (uses the product daily)?
   • Who is the blocker (raises concerns or slows the deal)?

Output from this step: A buying committee map -- four roles, with typical titles, departments, and primary concerns for each.

---

Step 5: Document the pain in customer language

The ICP is not finished when you know who the company is. It is finished when you know how they describe the problem they hired you to solve -- in their words, not yours.

This matters because your messaging, your outreach, and your discovery questions all need to match the vocabulary your buyer uses when they are in pain. A PMM who says "we have message drift across channels" and an AE who says "we need better brand consistency" are describing the same problem. One of those phrases will land in a cold email; the other will not.

What to do:

1. Pull direct quotes from:
   • Win interview transcripts
   • Sales call recordings (first discovery calls)
   • Customer onboarding calls
   • G2 / Capterra reviews from your category
   • LinkedIn posts by people in your ICP title
2. Look for the phrases that recur -- the specific words they use to describe the symptom, not the solution.
3. Write three "voice of customer" sentences that could plausibly appear in a Slack message from your ICP buyer on a bad day.

The companies that close fastest are almost always the ones where the AE's discovery question matches the exact language the prospect used to describe their problem internally. That match is not luck -- it comes from building the ICP on real quotes, not internal assumptions.

Output from this step: 3-5 direct voice-of-customer phrases that describe the problem your ICP has before they buy you.

---

Step 6: Write the ICP document

With data from the previous five steps, you now have everything needed to write a useful ICP. The document has four sections -- no more.

Structure:

1. The company snapshot (two to three sentences) A paragraph description of the exemplar company -- specific enough that a reader could identify a matching prospect. Reads like a description of a real company, not a job description.

2. Firmographic filters (bullet list) The hard criteria for list-building: size range, funding stage, industry, geography, stack signals, ownership structure.

3. Buying triggers (bullet list) The 2-3 situational events that predict a near-term purchase. Specific enough to turn into LinkedIn Sales Navigator filters or news alert keywords.

4. Buying committee (table or short paragraph) Economic buyer, champion, end user, blocker -- with typical titles and their primary concern in a deal.

ICP document completeness check

Output from this step: The ICP document. One page. Shared with sales, marketing, and CS before the week is out.

---

Step 7: Validate with your sales team before publishing

An ICP built without sales input is a marketing artifact. An ICP built with sales input is an operating tool.

Before the document goes live, run it past two or three of your best AEs. Not to get approval — to get correction. Ask them: Does this describe the accounts that actually close? Are there patterns here that don't match what you see in the field?

The answers will surface two common problems: the ICP is too narrow (excludes accounts that actually buy) or too broad (includes accounts that look right but always stall).

Output from this step: A validated ICP document, with at least one round of sales input incorporated.

---

Step 8: Schedule a 90-day ICP review

ICPs go stale. The company you were best at selling to eighteen months ago may not be the company you are best at selling to today -- the product has changed, the market has shifted, and your sales motion has evolved.

What to do:

1. Put a 90-day ICP review on the calendar today, before you publish the document.
2. The review agenda is simple: compare the last quarter's closed-won accounts against the ICP. How many were inside the profile? How many were outside it? What patterns in the outliers suggest the profile needs updating?
3. Assign ownership -- one person is responsible for running the review and updating the document.

Output from this step: A 90-day review meeting on the calendar with an assigned owner.

---

Using Stratridge to accelerate this process

• Positioning Audit: The eight-lens diagnostic surfaces where your ICP assumption is leaking — particularly in audience specificity and competitive frame.
• Strategic Context: Store the ICP as a memory pillar so every Stratridge tool — battle cards, launch playbook, copy studio — is grounded in the same profile.
• Win/Loss Review: The objection patterns from lost deals are often the fastest way to tighten an ICP that is too broad.