Measure the Handoff and Finish the System Plan

What you will complete

Complete the Lead Qualification and Sales Handoff System Planner and grade it against a 14-point capstone rubric.

Choose operating measures before outcome claims

Start with measures the system directly affects: time from arrival to review, records with sufficient evidence, MQL acceptance, returns for missing information, unassigned leads, routing corrections, overrides, duplicate actions and failures by step.

Track false positives and missed leads as reviewed examples, not unsupported percentages. Segment by source, offer and model version when enough records exist. A rise in MQL volume is not useful if sales acceptance falls or review slows.

Do not claim revenue impact without a baseline and an attribution method. The first proof can be operational: a reviewer receives the right evidence, records a clear decision and leaves no lead without an owner or next action.

Assign policy and system ownership

Set a review date and required evidence. Change one policy version deliberately, rerun affected tests and document the result. Do not let quiet prompt edits or CRM workflow changes alter the qualification contract without review.

Worked example: Harborline operating review

Fictional example: Harborline records its current median review time, missing-field rate and count of leads without a named next action before implementation. Revenue operations owns daily queues and weekly failure review. The revenue leader owns ICP and threshold changes. Sales supplies acceptance, return and override reasons.

At the first scheduled review, Harborline compares results by source and model version. Several referral leads were returned because a required field was unavailable. The team can choose to make the field optional for trusted referrals, add an approved enrichment step or keep manual review. It cannot claim the model improved revenue from that observation.

The system expands only after the core cases pass, reviewers can explain decisions, and fallback queues have owners.

Complete section 7 and assemble the capstone

1. Record current baselines using a defined population, period and source.
2. Name the owner and calculation for each operating measure.
3. Set the review cadence, change-control rule and pause authority.
4. Attach the ICP model, scoring rules, stage contract, responsibility map, routing matrix and test pack.
5. Grade every capstone criterion. Repair any zero before approving technical discovery.

Finish section 7 and the capstone rubric.

Capstone acceptance

The workbook scores seven sections from zero to two points each. A decision-ready submission earns at least 12 of 14 points with no zero. It names one offer, observable evidence, explicit unknown handling, human SQL acceptance, complete routing fallbacks, representative tests, baselines and accountable owners.

A passing planner is a business and operating specification. It does not prove technical feasibility, legal compliance, security approval or production performance. The next decision may be to proceed to technical discovery, repair source data, run more examples, narrow scope or stop.

Choose the appropriate next step

If one inbound qualification workflow is clear and the capstone passes, use it to scope an AI Lead Qualification System or AI Automation System. Bring the current workflow, recent sanitized examples and the point where leads wait or lose ownership.

If several workflows compete or the qualification problem is still unclear, an AI Success Audit can compare priorities before a build. Keep operating the manual review path until the tested replacement is approved.

Check your decision