Filterable criteria drawn from who actually buys — firmographics, technographics, triggers, and disqualifiers.
Build our ICP from HubSpot closed-won data. We sell to B2B marketing teams.
Headcount turned out to be a weak predictor — team composition is the real signal, with accounts having two or more content or growth roles closing at 3.1x the base rate. Agencies and companies under twenty people drive most churn and belong in disqualifiers.
An ideal customer profile template is useless if the output is an adjective. This one produces criteria you can filter a list on — size bands, technologies, triggers, and explicit disqualifiers — derived from the accounts that actually closed rather than from who you wish would buy.
A definition of the accounts worth pursuing, expressed as testable criteria: industry and size, technology in use, structural traits like team composition, observable triggers, and the disqualifiers that should stop a rep from working an account.
An ICP describes the company you should sell to; a persona describes the person inside it you talk to. You need both, and conflating them produces targeting that is either too broad to filter or too narrow to build a list from.
Replace every adjective with something observable. "Fast-growing" becomes a headcount growth band; "technically sophisticated" becomes named technologies in the stack; "understands our category" becomes a role that exists on the team. If a criterion cannot be looked up, it cannot be operationalized.
Because saying no is faster than qualifying yes. Explicit disqualifiers — wrong buying process, incompatible stack, size below the point where your value appears — save more rep time than any positive criterion.
The agent reads closed-won and closed-lost patterns from HubSpot or your CRM, combines them with the positioning and competitor set in workspace Knowledge, and produces criteria mapped to the fields your list-building tools can filter on. Output feeds the cold email and sequence templates so qualification happens before anyone writes.
Closed-won and closed-lost history if you have it; a description of your best ten customers if you do not.
It looks for traits that correlate with closing, not just traits your customers share.
Each criterion arrives with the field or source you would filter on.
Save the ICP so every outbound template qualifies against it automatically.
Every criterion is filterable, with a named data source
Derived from closed-won patterns rather than aspiration
Includes explicit disqualifiers, which save the most rep time
Feeds the outbound templates so qualification is automatic
B2B SaaS companies, 50 to 500 employees, running HubSpot and a product-led motion, with at least two content or growth roles on the team, currently hiring for growth — and explicitly not agencies or companies under twenty employees. Every clause there is something you can filter a list on.
Start from accounts that closed and retained, find the traits they share that prospects who churned or never closed did not, then express those traits as observable criteria. Working from aspiration instead of outcomes produces a profile nobody can use.
Narrow enough that a rep can disqualify in under a minute, wide enough to support your pipeline target. If both cannot be true, the constraint is usually pricing or positioning rather than the profile.
Every two quarters, and immediately after a pricing or product change. Profiles drift quietly as your product improves and your win patterns shift.