Compare recent wins, losses and customers who struggled after purchase. Look for shared buying triggers, alternatives, implementation conditions and successful outcomes. An ICP is useful when it helps your team make better targeting and qualification decisions, not when it merely describes an industry and company size.
Use outcomes, not only descriptions
An ideal customer profile should predict where your offer is useful and where a purchase is plausible. Industry, employee count and geography can help, but two companies with the same labels may have very different triggers, systems and decision processes. Start with accounts you know: recent customers who achieved value, deals you lost, and customers who bought but struggled.
The ICP guide shows which dimensions can matter. This exercise tests those dimensions against actual outcomes. Ask sales and customer teams to challenge the assumptions rather than defending the current target list.
Build a comparable sample
Select ten to twenty accounts if you have them. Include wins and losses from the same period and sales motion, plus a few customers whose implementation succeeded or failed. Record the buying trigger, prior alternative, expected outcome, stakeholders, urgency, budget process, required integration and what happened after purchase. Mark unknowns rather than filling gaps from memory.
Do not compare a small self-serve customer with a large enterprise procurement cycle as if they were identical. Segment the sample where the buying journey differs materially. A few interviews can explain why an apparently similar account made a different decision.
Look for conditions, not flattering patterns
A useful pattern might be: buyers with a specific regulation change and an existing data owner adopted quickly, while buyers without a data owner stalled. That is more actionable than “mid-market finance companies are best.” But beware of conclusions that simply describe where the sales team spent the most effort. If every win came from founder introductions, the pattern may reflect access rather than market fit.
The buying-trigger guide helps distinguish an urgent situation from a job title. Check post-sale outcomes too. A customer who signs quickly but never reaches value is weak evidence for an ideal profile.
Turn the hypothesis into a decision rule
Write a short rule your team can use: who to target, which trigger to look for, what makes implementation feasible, and which warning signs justify a lower priority. Test it on new enquiries for a month. Record accepted pipeline, sales progression and customer results. If the rule rejects good customers or accepts poor fits, revise it.
Apply this to your business
Put ten accounts in a table and highlight repeated conditions across the best outcomes. Write one targeting change and one qualification question from the pattern. Treat the result as a working hypothesis, not a permanent market boundary.
Frequently asked questions
How many deals are enough to validate an ICP?
There is no universal minimum. A small sample can reveal a hypothesis, but do not treat it as a final market truth. Keep checking new opportunities and customer outcomes.
Should we exclude every lost deal from the ICP?
No. A suitable buyer can be lost for timing, execution or competitive reasons. Separate fit from outcome before changing the profile.
Your target market looks broad but few customers are a strong fit?
Bring a sample of won, lost and struggling accounts. Mohit can help identify the buying conditions worth testing in your targeting and message.
Talk through the problem ↗See how we could work together
