Key takeaways
- An ICP describes the company you can help; a persona describes the person you talk to.
- Split criteria into must-haves, nice-to-haves and disqualifiers before you assign any points.
- Test the scale against won and lost deals, then tune it. The first version is a guess.
ICP fit is how closely a company matches the kind of customer you can sell to, serve well and keep (ICP stands for ideal customer profile). To define it, pick a short set of company facts a rep can check quickly, sort them into must-haves, nice-to-haves and disqualifiers, then score them and test the score against deals you won and lost.
Most teams have an ICP somewhere, usually in a slide that says “mid-market, B2B, growing.” That is a description, not a definition. You cannot score it, and two reps reading it will pick different accounts. This guide turns it into something you can apply the same way every time.
ICP vs persona
These two ideas get mixed up, so separate them first.
- ICP describes the company: its industry, size, setup and situation.
- Persona describes the person inside that company: their role, goals and daily problems.
A company can be a perfect fit while the person you emailed is the wrong one. A great persona match at a company that can never buy is still a dead end. Score fit at the company level first. Then use personas to decide who to contact, which is covered in mapping the buying committee.
Three kinds of fit criteria
Good fit criteria come in three groups. You need at least one from each.
Firmographic
Firmographics are basic company facts, like demographics for businesses.
- Industry or business model
- Employee count or revenue band
- Geography and whether it sits in your territory
- Company structure, such as single office vs many locations
Technographic
Technographics describe the tools and systems a company uses.
- A system your product connects to or replaces
- A category of tool that signals they already work in a certain way
- Missing tools that suggest they still do the work by hand
Situational
Situational criteria describe how the company operates right now. These are often the most predictive and the hardest to find.
- Team structure, such as having a dedicated sales development team
- How complex their process is
- Whether they feel the problem you solve today
Keep situational fit separate from buying signals. A team of 20 reps is a fit fact. Hiring 10 more reps this quarter is a timing signal. If you blend them, you lose the ability to explain why an account ranks where it does. The buying signals guide covers the timing side.
Must-haves, nice-to-haves and disqualifiers
Before you assign a single point, sort every criterion into one of three buckets.
- Must-haves. Without these, a deal is very unlikely. Example: the company has an outbound sales team.
- Nice-to-haves. These make a deal more likely or larger, but are not required. Example: they already use a tool you integrate with.
- Disqualifiers. If any of these is true, the account is out, no matter how good everything else looks. Example: they are in a region you cannot sell to, or they are a direct competitor.
This step matters more than the math. Disqualifiers keep good-looking accounts from wasting your week. Must-haves stop you from scoring a company highly just because it ticks many small boxes.
Aim for three to five must-haves, three to five nice-to-haves and a short list of disqualifiers. If you have twenty criteria, most of them are not changing anyone’s decision.
Turn criteria into a points scale
Now make it scoreable. Here is a simple approach on a 0 to 100 scale. Treat these numbers as a starting point, not a standard. Tune them to your own deals.
- Check disqualifiers first. If any is true, the score is 0 and you stop.
- Give each must-have a share of 60 points. If one is missing, cap the total score low, for example at 30.
- Give each nice-to-have a share of the remaining 40 points.
- Allow partial credit only where it is honest, such as a company slightly outside your size band.
Write down how each criterion is checked and where the answer comes from. “Has an outbound team” should mean something specific, like “at least three people with SDR or BDR titles on their professional profiles.”
A worked example
Imagine you sell a tool for sales development teams. Your illustrative scale looks like this:
| Criterion | Type | Points | How to check |
|---|---|---|---|
| B2B company selling to other businesses | Must-have | 20 | Website and customer pages |
| 100 to 2,000 employees | Must-have | 20 | Company profile |
| Has a team of 3+ SDRs or BDRs | Must-have | 20 | Job titles on professional profiles |
| Uses a CRM you integrate with | Nice-to-have | 15 | Job posts and tech mentions |
| Sells into multiple regions | Nice-to-have | 15 | Office locations and job posts |
| Hires sales roles regularly | Nice-to-have | 10 | Careers page history |
Now score two fictional accounts.
A 400-person logistics software company sells to freight carriers, has eight BDRs, mentions your integrated CRM in a job post and has offices in two regions. It meets all three must-haves (60) and the first two nice-to-haves (30). It rarely hires sales roles, so it gets 0 for the last one. Score: 90.
A 150-person consumer subscription brand has a strong marketing team but sells directly to consumers and has no outbound reps. It misses two must-haves. Under the cap rule, its score stays low no matter what else is true. That is the point: a familiar name does not make it a fit.
Check the definition against won and lost deals
Your first scale is a guess. Test it before you trust it.
- Pull a list of recent deals you won and deals you lost or that stalled.
- Score each company with your new scale, using only what you could have known at the start.
- Look for patterns. Do won deals mostly score high? Do lost deals score lower?
- Study the surprises. A won deal with a low score may reveal a missing criterion. A lost deal with a high score may show a must-have that is not really required.
- Adjust the weights, then repeat with the next batch of deals.
If you only have a handful of deals, do not over-adjust. Look for clear patterns, talk to the AEs who ran the deals, and revisit the scale each quarter.
The rule to remember
Fit should be something a rep can check in a couple of minutes and explain in one sentence. Sort criteria into must-haves, nice-to-haves and disqualifiers first, keep the points simple, and let real outcomes tell you when to change them.
Once fit is defined, combine it with timing to tier your territory.
Common questions
What is the difference between an ICP and a TAM list?
Your ICP describes the kind of company you can sell to and serve well. A TAM list is the set of real, named companies in your territory that could buy from you. In practice, you build the TAM list by applying your ICP criteria to a raw list of companies, so the ICP is the filter and the list is the result.
How often should you update your ICP definition?
Review it on a set rhythm, such as once a quarter, and whenever something big changes: a new product, a new market, new pricing or a run of surprising wins or losses. Small changes to weights are normal. Rewriting the whole definition every few weeks usually means it was never tested against real deals.
Can a company fit your ICP but still be a bad prospect right now?
Yes. Fit describes whether the company could be a good customer. Timing describes whether there is a reason to talk now. A strong-fit account with no recent change, or one that just signed a long contract with a competitor, may be better watched than worked. Keeping fit and timing separate lets you tell these cases apart.
Who should own the ICP definition?
Usually sales leadership and marketing agree on it together, with input from the people closest to deals, such as AEs, SDRs and customer success. Reps should be able to challenge it with evidence, like a pattern of won deals the criteria would have missed. One shared, written version matters more than who holds the pen.
