Define Your Ideal Customer Evidence

What you will complete

Complete the ICP evidence model: one offer, one purchase context, observable fit criteria, exclusions, evidence sources and unknown-data handling.

Choose one offer and one qualification decision

Lead qualification becomes unreliable when one model tries to serve several offers with different buyers and sales motions. Choose one offer for this course. State what the buyer is considering, which problem the offer addresses, and what decision the system must prepare.

The boundary should be narrow enough to test. “Qualify every contact” is too broad. “Prepare inbound requests for the workplace-safety training package so a sales reviewer can choose priority review, standard review, nurture, referral, or decline” is inspectable.

Separate an account from a lead. An account may fit your market while the current request does not fit the offer. A person may be relevant to the work but unable to represent the buying process. Your model should preserve those distinctions instead of forcing them into one label.

Write criteria as observable evidence

Use conditions another reviewer can check: supported geography, industry, operating need, organization type, role, request type, existing relationship and required timing. Avoid labels such as “good company,” “serious buyer” or “high intent” unless you define the evidence behind them.

Record exclusions separately. A competitor request, vendor solicitation or unsupported use case should have its own outcome and reason code. Missing evidence is not an exclusion.

Worked example: Harborline Safety Training

Fictional example: Harborline Safety Training sells manager-led workplace-safety programs to Ontario employers. Its course example covers inbound website forms, referrals and event enquiries. Harborline can serve employers in Ontario with a current training need. A reviewer must inspect requests from outside Ontario, sole learners seeking a public course, vendors, competitors and existing customers.

Harborline records the evidence source for each criterion. A submitted Ontario location is direct evidence. An inaccessible company website produces an unknown, not a failed company-size criterion. An existing customer match overrides normal assignment and sends the request to the account owner.

This is a teaching case, not a Flowgrammer client result. The criteria are deliberately simple so you can see how each later module changes the same system planner.

Complete section 1 of the workbook

1. Name one offer and the exact inbound decision the system will prepare.
2. Write four to eight fit criteria, each with an allowed evidence source.
3. Add explicit exclusions, escalation conditions and a separate action for unknown values.
4. Ask a colleague to apply the rules to one recent lead. Rewrite any criterion that depends on private intuition.

Download the Lead Qualification System Planner workbook or read the workbook instructions.

Common failure modes

• Combining several offers whose fit rules conflict.
• Treating a senior title as buying intent.
• Guessing company size, budget or authority from weak public data.
• Using missing information as proof that the account is a poor fit.
• Ignoring customers, partners and open opportunities during matching.

Your first model can remain incomplete. Label the gap and create a review action instead of filling it with false precision.

Check your decision