Describe the situation in which the product matters
Start with comparable customers rather than a list of desirable logos. Ask what was true before purchase, why the buyer changed and what made implementation viable. Separate repeated conditions from special relationships and custom scope.
Write observable account criteria a researcher can reproduce, then record what only a buyer can confirm. A pricing page may establish enterprise tiers; it does not establish that manual entitlement work is painful. Include hard disqualifiers such as unsupported integration needs, unsuitable economics or no relevant operating condition. Review rejected opportunities to sharpen the profile.
| Decision | Evidence to use | What changes next |
|---|---|---|
| Operating fit | Workflow and product conditions visible from reliable sources | Create a testable problem hypothesis |
| Adoption fit | Technical, organizational and commercial constraints | Exclude cases the standard offer cannot serve |
| Decision fit | Relevant roles and possible reason to change | Ask the buyer rather than claim private intent |
Work through the decision
Illustrative ICP: B2B platforms adding enterprise plans, with differentiated access requirements and a feasible standard integration. A researcher can verify product tiers and technical documentation. They cannot verify dissatisfaction with the existing process without a conversation.
An account that matches the public conditions but has a mature internal solution may not be worth active pursuit. Preserve that rejection reason. The ICP should become more precise as discovery reveals which conditions actually matter, rather than accumulating every company that once responded.
The profile describes company size but not a buying problem
A list of sectors and headcounts can produce a large market while leaving relevance undefined. Add the workflow, consequence and adoption condition. If those cannot be described, begin with buyer research instead of declaring the ICP complete.
A concrete next step
Use two wins and two losses to draft the ICP worksheet. Ask another person to classify five unfamiliar accounts and discuss disagreements before acquiring a larger list.
Sources and research notes
- Schematic product documentationCompany product description
- GitLab sales development handbookCompany operating handbook
Primary sources reviewed October 6, 2026. The operating recommendations and worked scenarios are Daavid’s analysis. Illustrative numbers are assumptions, not measured client results. Company marks identify sources and prior experience; they do not imply a customer relationship or endorsement.
Schematic