Ideal customer profile software that starts from the deals you won
Most ICP tools ask you to describe your ideal customer, then score against your own description. Telepath derives the profile from your closed-won deals, weights each trait by how strongly it shows up across your wins, and scores every open deal with enough data to score on how closely it resembles them; low-data deals are left blank, not guessed. The result is written into HubSpot as a single property your reps can see.
What ideal customer profile software does
Every B2B company has an ideal customer profile somewhere. Usually it is a slide: four or five traits written down in a workshop and not revisited since. The problem is not that the slide is wrong. It is that a slide is not a scoring system. “Mid-market SaaS, 200 to 500 employees” cannot help a rep decide which of forty open deals to work on a Tuesday morning, because it has no weights and so cannot rank anything.
ICP software does three jobs:
- Derives or captures the profile. What your best customers actually have in common: industry, size, region, deal shape, buyer role, how long the cycle ran.
- Weights the traits. If industry alignment carries far more of your model’s weight than company size does, the model says so, with a number.
- Scores against it. Every open deal or target account with enough data to score gets a score showing how closely it resembles the pattern, along with which signals matched and which were missing.
Job three is the one people buy. Jobs one and two are where tools differ, and where the value comes from.
The four kinds of ICP tool
Almost every product in this category is one of four things. They are marketed with similar language, so it is worth knowing which one you are evaluating.
add data to records you already have: firmographics, technographics, headcount, funding. They make your existing definition better informed. They do not tell you whether the definition was right.
tell you which companies are researching your category right now. That is a timing signal, not a fit signal. A company can be actively in-market and still be a bad customer.
score on how much someone is using your product or engaging with your content. That measures attention rather than fit. A poor-fit prospect who opens every email will outscore a perfect-fit prospect who does not.
compute the profile from your closed-won deals and weight the traits by how strongly they show up across the deals you have won. The model is derived, not declared.
Most stacks need more than one of these. Only the fourth answers “who should we be selling to?” from evidence rather than opinion, and it is the kind Telepath is.
Declared or derived: the question to ask any ICP tool
A lot of tools, including good and expensive ones, let you assign the weights by hand. You pick the signals that matter, give each a multiplier, and tune it over a few weeks until the results look right. That is faster and more consistent than a spreadsheet, and it is worth being clear about what it is: your hypothesis, applied at scale. If you set the weights, the model inherits your assumptions, including the wrong ones.
- You pick the signals that matter, give each a multiplier, and tune it over a few weeks until the results look right.
- your hypothesis, applied at scale
- If you set the weights, the model inherits your assumptions, including the wrong ones.
- Feed in the deals you closed, let the patterns across them decide what matters, and accept the answer even when it is uncomfortable.
- The most useful output is usually the trait you did not expect to matter.
- the profile it produces can tell you that the slide was wrong
A written-down ICP never can. There is a longer argument for this in why every definition of ideal customer profile describes a guess.
2 to 4
distinct win patterns in most closed-won datasets, not one
What Telepath produces
The two to four distinct customer patterns in your closed-won data, each described by the traits that define it, rather than one blended profile that describes none of them.
How much each trait carries, as a share of your model's weight, computed from your deals rather than assigned in a workshop.
A 0 to 100 measure of how closely each open deal resembles the deals you have already won, with the signals that matched and the ones that were missing.
The T-Score lands on each deal record as a single custom property, telepath_t_score, synced and scored nightly against your current model, so reps see it where they already work.
On the Vision plan, ICP Drift Intelligence shows when your open pipeline stops matching your win patterns, dimension by dimension.
A deal without enough data to score is left unscored rather than guessed, and a segment too small to be statistically real is flagged as such.
Every figure is computed from your data in code. The method, from closed-won export to weighted profile, is written up step by step in how to build a weighted ICP from closed-won data, and the full pipeline is on the how it works page.
Who this is for
B2B founders, sales leaders and RevOps teams with at least 10 closed-won deals who want their ICP derived from evidence rather than asserted in an offsite. Teams on HubSpot get the score written into the CRM; any CRM that exports a CSV can run the analysis. Under 10 won deals, the patterns are not statistically real yet, and the tool says so rather than inventing one.
Personal contact data never reaches the analysis: names, email addresses and phone numbers are stripped server-side before anything runs, and the column check shows you exactly what was removed.
How to start
Start with the free ICP report: upload a CSV of your closed-won deals and receive your segments, weighted criteria and win patterns as a PDF, usually within a few minutes. No account, no credit card. If the profile it finds is one you want every open deal assessed against inside HubSpot, with Telepath telling you when it has moved so you decide when to refit, the Reader plan does that from £500 a month.
Ideal customer profile software FAQs
Software that turns a description of who you sell to into something a sales team can act on: a weighted profile and a score for every account or open deal showing how closely it matches that profile. The tools differ mainly in where the profile comes from. Some ask you to type it in. Telepath derives it from the deals you have already closed.
A CRM lead score is configured: someone chooses the signals and the points each one earns, and the CRM adds them up. Telepath’s ICP is learned from your closed-won deals, so the weights come from what your wins actually have in common, including the traits nobody thought to put on the slide. The two can sit side by side; they answer different questions.
Resemblance. A T-Score of 80 means an open deal closely matches the pattern of deals you have won before; a score of 20 means it does not. It is a fit measure computed from your own history, and every score comes with the signals that matched and the ones that were missing. It describes the deal’s fit against your history and says nothing about the deal’s outcome.
For the free ICP report, no: upload a CSV of closed-won deals from any CRM or spreadsheet. For live pipeline scoring on a paid plan, HubSpot is the integration that is live today (Starter, Professional or Enterprise). Salesforce, Pipedrive, Microsoft Dynamics and Zoho are on the roadmap.
A minimum of 10. Between 30 and 500 gives the sharpest results. Below roughly 50, treat the output as directional; the report itself says where the sample is thin rather than papering over it.
Personal contact columns such as names, email addresses and phone numbers are detected and stripped automatically before analysis begins, and company names are removed too. The analysis works on the shape of your wins, not the names in them. Your data is never sold and never used to train AI models.
The ICP report is free, with no account and no credit card. Ongoing pipeline scoring inside HubSpot starts at £500 per month on the Reader plan, with Vision and Oracle above it. The pricing page lists every plan in full.