ICP vs persona
The ICP describes the company you want to sell to. The persona describes the people inside it. Conflating them produces campaigns aimed at a job title in the wrong kind of company — the most common targeting error we see.
Firmographics
Industry, size, geography, growth stage, funding and structure — the attributes a researcher can filter on without judgement.
Technographics
What they run, what they integrate with, what they'd have to replace. For most B2B SaaS this is a sharper fit signal than headcount.
Business model fit
How they make money determines whether your product touches a cost line or a revenue line — and therefore how urgent it feels and who owns the budget.
Pain
Write the pain in the buyer's language, drawn from real calls. If your pain statement uses your product category's vocabulary, it's a positioning statement, not a pain.
Triggers
The events that turn a tolerated problem into a funded one.
- Funding rounds
- Relevant hiring
- Leadership changes
- Stack or vendor changes
- Growth thresholds
- Regulatory change
Buying conditions
Budget ownership, procurement burden, security review requirements and realistic cycle length. These decide whether a good-fit account is actually reachable this quarter.
Disqualifiers
Explicit exclusions: company types that churn, stacks you can't support, sizes where the economics fail. Disqualifiers usually improve results faster than adding criteria.
Persona mapping
For each ICP segment, map who feels the pain, who owns the budget, who blocks and who benefits — then give each one its own entry point.
Scoring
A simple weighted score across fit and trigger strength is enough. Complex models tend to encode assumptions you haven't tested yet.
Segmentation
Split the ICP into two or three segments that behave differently. Each becomes a separate campaign hypothesis with its own messaging.
Testing the ICP
Outbound is the fastest ICP test available. Reply quality, objection type and meeting-to-opportunity conversion tell you within weeks whether the definition holds.
Example ICP framework
Use this structure as a worksheet — fill each line with a criterion a researcher could act on unaided.
- Segment name and hypothesis
- Firmographic filters
- Technographic signals
- Business model conditions
- Primary pain in buyer language
- Trigger events
- Buying group and entry persona
- Disqualifiers
- Estimated account universe size
- What result would disprove this segment
FAQ
Narrow enough that you could name twenty accounts that fit perfectly and explain why. If the definition includes most of your market, it isn't an ICP — it's a category.
Monthly while outbound is new, quarterly once segments stabilise. Revisions should be driven by conversations and closed-won patterns, not by opinion.
Not sure this is your constraint?
Pressure-test your ICP