← All field notesDATA ANALYST · METRIC DEFINITION ALIGNMENT

AI Prompt: Data analyst metric-definition alignment memo

Two teams are using the same metric name but different logic. The dashboard is technically correct and politically useless until everyone agrees what the number means. This prompt turns the disagreement into a decision memo.

The prompt

You are a data analyst preparing a metric-definition alignment memo. Metric name: [e.g. active customer, churn rate, qualified lead, gross margin]. Stakeholders: [teams/owners]. Current definitions: [paste each team's definition, SQL snippet, spreadsheet logic, or description]. Data sources involved: [tables/tools]. Business decisions affected: [forecasting, compensation, board reporting, campaign optimisation, etc.]. Known edge cases: [refunds, trials, paused accounts, duplicates, cancellations, time zones]. Write a memo with: (1) The conflicting definitions in a comparison table (2) The practical business impact of each difference (3) Recommended canonical definition with inclusion/exclusion rules (4) Data source of truth and refresh cadence (5) Edge cases requiring stakeholder decision (6) A short approval request stakeholders can reply to. Keep it neutral and specific. Do not declare a winner without explaining tradeoffs.

What you’ll get back

A stakeholder-ready metric memo that turns definition drift into an explicit source-of-truth decision, including edge cases and approval language.

Tips for this one

  • Show the business impact, not just the SQL difference. Stakeholders care when they see how a definition changes revenue, churn, or bonuses.
  • Name edge cases separately. Most metric disputes hide in trials, refunds, time zones, duplicates, and reactivations.
  • Do not frame one team as wrong. Different definitions often served different past decisions; the memo is about choosing one future use.
  • End with approval wording. A metric definition is not final until the owners explicitly sign off.
99 more like this

Data Analyst prompt bible — $9.99 USD

100 prompts like this one — pricing, quoting, scheduling, hiring, hard emails. One-time payment. Yours forever.

Open the Data Analyst bible →