How to write an ICP a data engine can actually execute
A field by field template for an ideal customer profile that a data provider can run without guessing, including the exclusions most teams forget.
9 min read · Published September 13, 2026 · Updated September 16, 2026
The short answer
The short version.
- Most ICPs fail because they describe a feeling, not filters. A provider needs industry, size, geography, roles and exclusions.
- Exclusions do more work than inclusions. Vendors, competitors, agencies and junior titles are the usual leaks.
- Name the buying trigger and the person who signs, separately. They are often different people.
- A good ICP fits on one page and produces a rejection reason for every record it turns away.
Ask ten teams for their ICP and nine will send a paragraph about mid market companies that value efficiency. No engine, human or machine, can execute that.
This is the template we run against, field by field, with the failure mode each field prevents.
The seven fields that make an ICP executable
Industry, employee range, revenue range, geography, primary roles, secondary roles and exclusions. Anything missing becomes a guess.
Write each field as something a filter can test. Instead of growing companies, write 20 to 200 employees. Instead of North America, name the countries and, where it matters, the states or provinces.
| Field | Weak version | Executable version |
|---|---|---|
| Industry | Home services | Residential HVAC, plumbing and electrical contractors |
| Size | Small business | 10 to 150 employees |
| Geography | US and Canada | United States and Canada, excluding Quebec |
| Primary role | Decision maker | Owner, President, General Manager |
| Secondary role | Anyone senior | Operations Manager, Service Manager |
| Exclusions | None stated | Franchisors, suppliers, software vendors, staffing agencies |
| Trigger | Not stated | Hiring field technicians, multiple locations |
Exclusions are where lists are won
Every list has predictable contaminants: people who sell to your market rather than buy from it, companies too large to reach, roles that cannot sign, and firms you already work with.
Write the exclusion list before the inclusion list. It is shorter to write and removes more waste.
- Competitors and anyone reselling a similar product.
- Vendors, consultants and agencies serving your target industry.
- Job titles containing assistant, intern, junior or coordinator when you need a signer.
- Existing customers, open opportunities and anyone who has opted out.
Separate the person who answers from the person who signs
In owner operated businesses they are the same person. Above roughly 80 employees they usually are not, and a list built only for the signer will get you voicemail.
Name a primary role you want to reach and a secondary role that will be reached more easily and can introduce you. Both belong in the brief.
Demand rejection reasons
A provider that cannot tell you why a record was turned away is not screening, it is filtering a spreadsheet. Rejection reasons are the proof that your exclusions ran.
We record a reason for every rejected record and score every accepted one against the brief. The data quality page explains the screens, and the onboarding flow collects this template field by field.
Key takeaways
- Turn every ICP line into something a filter can test.
- Write exclusions first; they remove more waste than inclusions add.
- Distinguish the reachable person from the signing person.
- Insist on per record rejection reasons as proof of screening.
See the data behind the advice
Five ICP matched prospects with mobile numbers, direct dials and work emails, free. Same pipeline that fills a paid account, no credit card.
Frequently asked questions
What should a B2B ICP include for a data provider?
Industry in concrete terms, employee and revenue ranges, named countries or regions, primary and secondary job titles, explicit exclusions, and any buying trigger. Each field must be testable by a filter.
Why do lead lists contain the wrong companies?
Almost always because exclusions were never written down. Vendors, agencies and competitors serving your target industry match the same keywords your buyers do.
Keep reading
ICP
Seven ICP mistakes that quietly kill outbound
The targeting errors that make good lists behave like bad ones: vague industries, missing exclusions, wrong seniority, unbuildable signals and profiles written for humans rather than systems.
Buying
How to test a lead list vendor in one week
A repeatable seven day test for B2B lead data vendors: what to ask for in a sample, which fields to check, how to score match quality and what contract terms to insist on.
Buying
Where to buy B2B mobile numbers and direct dials for your ICP
A plain comparison of the four ways to buy B2B mobile numbers and direct dials, what each costs per usable record, and which one fits a calling team.