Building an ICP Fit Score in HubSpot From Data You Already Have
Published June 13, 2026
TL;DR: Most teams reach for paid intent data before they have used the firmographic signal already sitting in their own CRM. An ICP fit score is a transparent, rules-based measure of how closely an account matches your ideal customer profile, built from properties you already store: industry, employee count, region, and the product signals your own deals reveal. The work is not modeling, it is definition and data hygiene. A fit score is only as trustworthy as the fields it reads, so the same property completeness, deduplication, and lifecycle integrity that govern reporting govern scoring too. Build the score from a written ICP definition, weight a small number of high-signal attributes, keep it inspectable, and validate it against your actual closed-won accounts before you let it route anything.
Fit is not intent, and you need fit first
Two different questions hide inside “is this a good account.” Fit asks whether the account resembles the customers you serve well. Intent asks whether the account is in-market right now. Intent data is the expensive one teams rush to buy, but intent without fit routes your team toward in-market accounts you cannot serve profitably. Fit is the cheaper signal and the one you already own, because it is firmographic and it lives in your CRM.
Starting with fit is also more honest about what your data can support. You can compute fit from properties you already collect. Intent usually requires a third-party feed whose methodology you cannot inspect. For a team building its first account-based motion, a transparent fit score is the place with the highest return for the least effort.
A fit score is a definition problem, not a modeling problem
The instinct is to treat scoring as a machine-learning task. For most B2B teams that is premature. A weighted rules-based score that a human can read and defend will outperform an opaque model the team does not trust, because the team will actually act on a score they understand. The hard part is not the math. It is writing down what “ideal” means precisely enough to compute.
That definition is the same artifact a disciplined go-to-market motion needs anyway. If you cannot state your ICP as a small set of attribute thresholds, the problem is not the score, it is that the ICP is not yet defined. The score forces the definition into the open, which is most of its value.
The attributes worth scoring
A good fit score weights a small number of high-signal attributes rather than every field you have. More inputs dilute the signal and make the score harder to defend. Start with the firmographic properties HubSpot already supports on the company record:
- Industry. Map each industry value to a tier: target, adjacent, out-of-profile. This is usually the single strongest fit signal.
- Employee count or revenue band. Encode your real sweet spot as a range, not a floor. A company can be too large for your motion as easily as too small.
- Region. Score for the territories you actually serve and support, not the ones you would like to.
- Product or technology signal. If your own closed-won deals cluster around a platform or a use case, that pattern is fit evidence you already own.
Each attribute gets a weight that reflects how strongly it predicts a good customer, and the weights are written down and owned by a named person, exactly like a lifecycle stage definition.
Why the score is only as good as the data underneath it
This is where most fit scores quietly fail. A score that reads industry, numberofemployees, and country is only as trustworthy as those fields are complete and consistent. If industry is blank on a large share of companies, the score is silently scoring those accounts as out-of-profile when they may be your best fits. If a single account is split across two duplicate company records, its firmographics and its deal signal are fragmented, and the score sees half an account twice.
That is why an ICP fit score depends on the same foundation as everything else. Property completeness on the scored fields, deduplication so one account is one record, and association health so company firmographics actually reach the contacts being routed. These are the layers covered in our HubSpot Data Foundation Audit methodology, and they are prerequisites for scoring, not separate work. A fit score built on incomplete properties is the attribution problem in a new costume: a number that looks precise and is quietly partial.
Validate against closed-won before you trust it
A fit score is a hypothesis until you check it against reality. The validation is simple and it is the step teams skip. Pull your last several quarters of closed-won accounts and run the score against them. A trustworthy fit score should rate most of your won customers as high fit. If it does not, the score is wrong, or your ICP definition does not match who you actually sell to, and either way you have learned something before the score routed a single lead.
Run the same check against closed-lost and churned accounts. If high-fit accounts are churning, the fit definition is rewarding the wrong attributes. This back-test is the external signal that keeps a fit score honest, the same way reconciliation keeps a migration honest.
The takeaway
You probably do not need to buy intent data to start an account-based motion. You need to use the firmographic signal already in your CRM, encoded as a transparent fit score that a human can read and defend. Write the ICP definition first, weight a few high-signal attributes, build the score on fields you have verified are complete and deduplicated, and validate it against your real closed-won accounts before it routes anything. The score is only as good as the data underneath it, which is why fit scoring is a data-foundation problem before it is a targeting one.
If you want to score account fit but suspect the underlying fields are too incomplete to trust, that is the place to start. Start with a Data Foundation Audit, where property completeness and deduplication are the layers a fit score depends on.
Sources
- HubSpot Knowledge Base, “Create and edit properties”: https://knowledge.hubspot.com/properties/create-and-edit-properties
- HubSpot Knowledge Base, “HubSpot CRM default company properties”: https://knowledge.hubspot.com/properties/hubspot-crm-default-company-properties
- Internal account-fit scoring method and anonymized closed-won back-tests (no client name, no portal identifiers).