Ideal customer profile working session
Choose a useful market boundary using evidence, exclusions, and a test plan.
Use when the team is spreading effort across too many account types or mistaking a few large wins for a repeatable market.
Before the working session
- Export a consistent period of won, lost, no-decision, and retained accounts. Include deal size, cycle length, implementation effort, and support load where available.
- Ask sales, success, and product to name the conditions that made value easy or hard to deliver. Record disagreements before the session.
Run the process
Build a segment table
Choose two or three segmentation dimensions that could change the motion: company size, workflow maturity, trigger, or required integration. For each segment, compare volume, qualified conversion, time to value, retention signal, and cost to serve. Put sample size beside every rate.
Separate fit from reach
Score problem intensity, ability to deliver, buyer access, and economic potential separately. A segment can fit the product but be too expensive to reach through the current channel.
Write observable gates
Translate the preferred segment into conditions a researcher can verify: trigger, team structure, systems, minimum operating complexity, and explicit exclusions. Put unknown accounts into a research queue.
Run one controlled test
Select a fixed account sample and one message. Track qualified conversations, disqualification reasons, and time spent. Compare against the current motion; decide whether to expand, revise, or stop on a set date.
Fictional example: A workflow SaaS team sees strong interest from regional service firms but slow deployment at very large enterprises. It tests 40 firms with 100–500 employees, a recent multi-location expansion, and a named operations owner. It records qualified meetings and implementation objections for six weeks before changing targeting.
Failure checks
- “Anyone with the problem” is too broad to guide prospecting.
- Revenue alone hides costly onboarding and poor retention.
- A vertical label without a trigger does not explain urgency.
Working worksheet
ICP DECISION RECORD Decision date / owner / review date: Job and accountable buyer: Trigger that makes the job urgent: Segment definitions and sample sizes: For each segment: wins / losses / no-decisions / cycle / time to value / retention signal / effort: Evidence quality and missing data: Fit gates observable before discovery: Exclusions and reason: Reachable channel and message hypothesis: Test sample / duration / owner / quality threshold / effort guardrail: Expand if / revise if / stop if: