Sign in or create a free account to save this Card to your Deck.
Practice Card · Evidence of progress
Research the empirical ideal customer profile
YOUR BEST-FIT BUYERS LEAVE AN EVIDENCE TRAIL.
An empirical ideal customer profile is a working decision standard derived from observed buying, delivery, retention, and economic outcomes. It specifies which buyers tend to benefit from the offer, when their problem becomes pressing, and which conditions predict poor fit. It is different from a demographic persona or a list of industries that could theoretically purchase. Strong ICP research compares both successful and unsuccessful accounts, checks data quality, includes the buyer's own language, and documents exceptions. The finished output should improve the next targeting and qualification decision, not merely produce a polished audience description. AI accelerates synthesis, while the team retains responsibility for the interpretation and commercial judgment.
Expanded teaching
Expanded teaching
Prerequisites and inputs
Choose one offer with enough history to inspect. Gather a usable sample of won and lost opportunities, customer outcomes, renewals, service and support records, and contribution margins where available. Note which accounts required unusual founder attention or repeated scope exceptions. Obtain and use buyer interview language only where it is permitted. Before analyzing, check that deal stages, loss reasons, industry labels, and account identifiers mean the same thing across records; inaccurate CRM classifications produce confident but spurious patterns.
A disciplined six-step method
1\. Define the outcome that makes a customer good for both sides: fit, realized benefit, margin, retention, and manageable delivery. 2. Segment recent accounts by observed outcome rather than only revenue. 3. Compare buying triggers, business context, decision authority, urgency, and constraints across the groups. 4. Examine why apparently attractive accounts became expensive or unsuccessful. 5. Draft positive fit conditions and explicit disqualifiers. 6. Validate the draft against new prospects and revise it when exceptions reveal an incomplete rule. Do not build a profile from only your favorite customers.
The distinction that makes the research useful
A company size or job title can help locate a buyer, but it rarely explains why the buyer acts now. The trigger event matters: an operational backlog that threatens delivery, a compliance deadline, a costly recurring failure, or growth that has exceeded existing systems. Disqualifiers matter just as much. A buyer may fit the industry and size while lacking budget, implementation ownership, necessary data, or a problem the offer can solve profitably. The profile should help staff explain both acceptance and decline decisions.
How to interpret AI patterns
Ask AI to summarize candidate correlations and cite the underlying records, then examine whether the result survives reasonable challenges. Small samples, missing close reasons, inconsistent project costing, and a skew toward memorable clients can all produce false confidence. Treat an AI-generated segment as a hypothesis. Have commercial and delivery owners inspect representative cases, including inconvenient exceptions. Never infer sensitive personal characteristics or let an unreviewed model make consequential decisions about people.
What to measure in the next cycle
Test the ICP against the next set of qualified inquiries. Track the proportion that meets documented criteria, stage conversion, time spent on rejected leads, scope changes, and client fit after delivery begins. Also review whether promising unusual accounts were excluded for the wrong reasons. A narrower list that brings fewer leads but better client outcomes may be a stronger system than one that maximizes contact volume.
The first deliverable
Produce a versioned ICP Decision Sheet with positive criteria, trigger events, disqualifiers, supporting evidence, exceptions, and the owner who revisits it. Use that one document to inform Offer, Message, Lead Qualification, and Sales Presentation. If the next salesperson cannot make a defensible fit decision without the founder, the practice is unfinished.
Diagnostic inquiry
Questions to ask, not conclusions
Do our documented ICP criteria explain which recent customers benefited and stayed profitable, and which apparently attractive prospects became bad-fit accounts?
Questions that distinguish possible explanations
- What evidence distinguishes repeat buyers from expensive or high-friction customers within the same industry?
- Which events created a decision deadline for customers who bought, rather than merely expressed interest?
- What recent counterexample would force us to adjust the proposed exclusion criteria?
Practical application
Practical application
Assemble a representative recent sample of wins, losses, renewals, and difficult delivery accounts. Clean the records, tag trigger and outcome patterns, draft an ICP Decision Sheet, and test it prospectively for one sales cycle before making it a hard automatic filter.
Ask Atria
Talk through this Card
Talk it through with us
If this Card describes your week, a free conversation is the fastest way to see what it means for your company and what to do next.
Book a free conversation