ICP Template: Fill It In and Approve It Before Outreach

An ideal customer profile (ICP) is a written description of the companies you want as customers. Most ICPs fail in one of two ways. They are too vague to filter anything ("growing B2B companies"), or they are a set of guesses that nobody checked. This guide gives you a worksheet that avoids both. Every field has a place for evidence. Every row gets a score. Nothing counts until a person signs it off.

If you want the longer reasoning behind each field, read How to Define an Ideal Customer Profile. This page is the working document. For a short definition, see the ICP glossary entry.

Who this is for

Use this worksheet if:

• You are a founder, sales lead or marketer who has to say who you sell to, and you want that written down in a form others can check.
• You have some closed deals, won or lost, in a CRM or a spreadsheet. A handful is enough to start.
• You plan to use the profile to drive outreach, ads or account lists, and you want a sign-off step first.

It is less useful if you have no customers and no deal history at all. You can still fill it in, but most rows will be hypotheses, and the worksheet will label them that way. That is fine. It is better than presenting guesses as findings.

What you need before you start

Gather three things. As a rough guide, allow an hour or two, though your own time may differ.

1. A list of your won companies and your lost companies. One row per company, not per deal. Outlevo's win/loss analysis counts companies, not deals, so a customer who bought three times is one won company. [[1]](#sources) Counting this way stops one large account from looking like a pattern.
2. Your own notes on why you won and lost. Loss reasons from your CRM are the most useful input. If you keep deals in a spreadsheet, a CSV import maps columns to domain, name, country, size, industry and deal fields such as stage, amount, outcome and loss reason. [[2]](#sources) It is worth setting your spreadsheet up with those columns now, even if you never import it.
3. Access to a few public pages from your best customers. Their site, a customer page or a job post. You will quote from these as evidence.

The template

Copy this table into a document or spreadsheet. Make one copy per segment. If you sell to two different kinds of company, fill in two worksheets rather than blending them.

How to fill each column

Field and What to write. Keep entries short. If an entry needs a paragraph, the field is not specific enough yet. "50 to 200 employees" can be checked against a company. "Mid-sized" cannot.

Evidence to attach. Each row needs evidence you can point to, not an impression. Write the source and the date next to it. Outlevo's Readiness Card applies the same rule to individual accounts: it needs at least three observations, each with a source and a date. [[3]](#sources) Borrow that rule for your template. A row with fewer than three dated, sourced observations is thin.

Evidence status. Use one of three labels:

• Verified observation: you or a colleague saw it in a record or on a page, and you can show where.
• Inference: you concluded it from other facts. Write the facts it came from.
• Hypothesis: a guess you want to test.

These labels follow the ones the free assessment applies, which marks each fact as a verified observation or an AI inference. [[4]](#sources) Outlevo also labels guesses as hypotheses and never presents them as findings. [[3]](#sources) Do the same in your own document. If you quote a customer's website, copy the exact words. Outlevo keeps facts read from a website only when their quote is found on the page. [[3]](#sources) A quote you cannot find again is not evidence.

Match count. Write it as "X of Y won companies." X is how many won companies fit the row. Y is how many won companies you have in total. Do the same for lost companies in rows 9 and 10. Count companies, not deals. [[1]](#sources)

Score. See the next section.

The scoring column

The score answers one question: how much should anyone trust this row? It is not a score of how good the segment is. Use three values.

• Supported. The row has at least three observations, each with a source and a date, and the evidence status is verified observation. The match count is filled in. This threshold is this template's own suggestion, not an Outlevo rule.
• Thin. The row has one or two observations, or the status is inference. Someone can still work from it, but the reviewer should read it closely.
• Guess. The row has no sourced observation. Its status must read hypothesis.

Then add one more check for the whole worksheet, which looks at your sample size. Outlevo labels small samples, and fewer than 10 won companies is very low confidence. [[1]](#sources) If your worksheet draws on fewer than 10 won companies, write "very low confidence" at the top of the page. This does not make the worksheet useless. It tells the next reader to treat every row as something to test. For more on working with few deals, see How to run a win/loss analysis on few deals.

This template sets no further confidence bands above that line. If you have 10 or more won companies, write the count and let your reviewer decide how much weight to give it.

Steps

Work through these in order. Plan a few hours for the first pass, though your own time may differ.

1. Create one worksheet per segment. Start with the segment you believe is strongest. Name it in row 1.
2. List your won and lost companies. One line per company, with domain, name, country, size and industry. Add stage, amount, outcome and loss reason for any deal you have. If you use HubSpot, Outlevo reads deal pipeline, stage, amount and closed-lost reason, which are the same fields to pull by hand. [[5]](#sources) If you use Salesforce, it reads accounts and opportunities and their stages. That connection is a pilot. [[6]](#sources) A spreadsheet works just as well for this exercise. To import one later, see CSV and Excel import and export.
3. Fill in row 2, countries, yourself. Write the countries you actually sell into and can show from your own records. Do not copy a list from a tool or a competitor. Outlevo asks which countries a customer sells into and never proposes markets. [[1]](#sources) Follow the same discipline: this field records a decision you made, with evidence from your own sales.
4. Fill the firmographic rows (3 and 4). Size and industry come from the list you built in step 2. Write the range or names, then the match count.
5. Fill the story rows (5 to 7). Trigger, buyer role and problem need the most care, because they are where wishful thinking creeps in. For each one, attach a dated source. If you have none, write the entry as a hypothesis and score it "Guess."
6. Fill the pattern rows (8 to 10). Look at the won list, then the lost list. Write what the won companies share, what the lost companies share, and which traits should rule a company out. If you have loss reasons in your records, quote them. If you do not, say so in the evidence cell.
7. Fill row 11, data you can check. For each field, note whether it can be verified from public information. A trait nobody can check from the outside will be hard to use in a list or a filter. Mark it so the reviewer knows.
8. Label and score every row. Set the evidence status, match count and score. Be strict. Some rows may come out as "Thin" on a first pass. That is acceptable.
9. Write the sample-size note and hand over. Add the count of won companies, the confidence label if it applies, and the date. Then go to the review below.

A shortcut if you want a first draft

The free ICP assessment reads up to 5 public pages of a website. [[4]](#sources) It returns facts labelled as verified observations or AI inferences, and up to 5 experimental segment hypotheses. [[4]](#sources) You can use that output as a draft for rows 1 and 3 to 7. Treat each segment as experimental: Outlevo labels AI-proposed ICP segments experimental until a person approves them. [[1]](#sources) The review in the next section mirrors that approval on paper. It does not replace approving the segments in Outlevo itself.

If you import past deals from a CSV or your CRM, Outlevo returns a win/loss analysis that proposes ICP hypotheses. [[7]](#sources) Those arrive as hypotheses too. Paste them into your worksheet with the status "Hypothesis," then work through the evidence column as above.

What to check

Run this review before anyone approves the worksheet. Ideally the reviewer is not the person who filled it in. A fresh reader notices a missing source faster than the author does.

Row-level checks

• Does every row have an evidence status?
• Does every "Supported" row have at least three observations, each with a source and a date? [[3]](#sources)
• Can you open each source and find it again? For quoted website text, is the quote on the page? [[3]](#sources)
• Is every guess labelled as a hypothesis, and not written as a fact? [[3]](#sources)
• Is the match count filled in as "X of Y," counting companies and not deals? [[1]](#sources)

Worksheet-level checks

• Is the number of won companies written at the top? If it is under 10, does the page say "very low confidence"? [[1]](#sources)
• Does row 2 list only countries you chose and can show from your own records?
• Do the disqualifiers in row 10 appear in the lost list, or in your churn notes? A disqualifier with no example is an opinion.
• Does the lost pattern in row 9 contradict anything in the won pattern in row 8? If it does, resolve it. Either the segment is too broad, or one of the patterns is wrong.
• Does each segment worksheet stand alone? If two segments share rows, they may be one segment.

What the reviewer decides

For each row, the reviewer picks one of three outcomes: approve, send back with a note, or keep as a hypothesis. Rows kept as hypotheses are allowed to stay in the document. They should not drive outreach on their own.

Approve it before outreach uses it

The approval step is the point of the whole exercise. A worksheet that nobody signs is a draft, and a draft should not decide who gets contacted.

Add a block at the bottom of each worksheet:

• Approved by (name and role)
• Date
• Version (for example, v1)
• Rows approved, rows kept as hypotheses, rows sent back
• Sample-size note carried from the top

Then set a rule for your team: only an approved version is used to build lists, write messages or set targeting. A change to any row makes a new version and goes through the same review.

This mirrors how Outlevo handles outreach. Nothing is sent outside a policy that a person approved. Policies set who can be contacted, on which channels, how often and with what budget. [[8]](#sources) Every external action is checked against the approved policy before it runs, and recorded in a ledger. [[8]](#sources) Kill switches pause a campaign, a channel or the whole workspace at once. [[8]](#sources) Treat your worksheet as the basis for any policy you set. If the worksheet is not approved, there is nothing sound to build a policy on.

If you connect a CRM, a similar gate applies to the data. For HubSpot, nothing is written until an owner or admin approves the field mapping. [[5]](#sources) Outlevo writes only to its own properties, including fit score, ICP version and reason. [[5]](#sources) The version number on your worksheet is your team's own record. Connecting and syncing a CRM uses no credits. [[5]](#sources)

Costs to expect if you use Outlevo for the follow-up

The worksheet itself costs nothing to fill in, and a CSV import or export uses no credits. [[2]](#sources) If you later qualify accounts against your approved profile, qualification costs 3 credits and light qualification costs 1 credit. [[9]](#sources) The Free plan is USD 0 per month. [[9]](#sources) See the pricing page for the other plans.

Next step

Do these three things today:

1. Copy the table above and fill in rows 1 to 4 for your strongest segment, using your won and lost company list.
2. Mark each row with an evidence status and a score. Write the number of won companies at the top.
3. Give the worksheet to one other person with the review checklist. When they sign it, you have a profile that outreach can use.

If you want a draft to react to, run the free ICP assessment on your site and paste its facts and segment hypotheses into the worksheet with the right labels. For the reasoning behind each field, go to How to Define an Ideal Customer Profile. To see how approval works inside the product, read How Outlevo works.

Sources

1. ICP from evidence
2. CSV and Excel integration
3. Readiness Cards
4. Free ICP assessment
5. HubSpot integration
6. Salesforce integration
7. Release 2026-10-19
8. How Outlevo works
9. Outlevo pricing