Audit a metric definition guide for ambiguity

Review internal metric dictionaries and data definitions for edge-case loopholes. Prevent conflicting dashboard reports across product, sales, and finance teams.

You getRisk Summary, ambiguity Breakdown and standard Boilerplate Rules

Prompt

Prompt
Act as a senior data governance lead. Audit the metric definitions pasted below to find edge cases, ambiguities, and calculation loopholes.

- Check whether time boundaries, timezone standards, and cancellation timing are clearly specified.
- Identify how edge cases like refunds, trial users, paused accounts, or internal test records are treated.
- Highlight where product, finance, and marketing teams are likely to report contradictory numbers using the current wording.

Output format:
1. Risk Summary: 2 sentences on the overall reliability of the dictionary
2. Ambiguity Breakdown: Table with columns Metric Name | Identified Loophole | Recommended Clarification
3. Standard Boilerplate Rules: 3 operational rules to append to all metric definitions

Here is the metric definition guide:

Example output

Example for: internal definitions for Monthly Active User, Churn Rate, and Expansion MRR

  1. Risk Summary:
    The current definitions lack standardized event timestamps and handling for non-revenue accounts, which will lead to reporting discrepancies between billing and analytics systems. Establishing explicit exclusion criteria will resolve immediate cross-team disputes.
  2. Ambiguity Breakdown:
  • Monthly Active User (MAU) | Does not specify whether passive logins or background webhook triggers count as activity. | Define active criteria as an explicit user-initiated write event within a rolling 30-day window (UTC).
  • Churn Rate | Fails to state whether churn is recognized on cancellation request date or subscription expiration date. | Measure churn strictly on the end-of-service date, categorizing prior requests as pending churn.
  • Expansion MRR | Does not address whether currency conversions or tier restructuring without price changes count. | Attribute expansion exclusively to net seat additions or plan upgrades using fixed-rate baseline currencies.
  1. Standard Boilerplate Rules:
  • All timestamp measurements must default to UTC midnight boundaries.
  • Test, internal, and employee accounts must be excluded by default via standardized database tags.
  • Refunded transactions must retroactively debit the original transaction month rather than the settlement month.
Written by AI from this prompt. Your results will vary.

Copy your company metric dictionary or KPI definitions and paste them directly after the prompt in Claude, ChatGPT, or Gemini. Ask the model to draft SQL-ready logic tests to validate the cleaned definitions.

Make it repeatable

Keep using this prompt without pasting it in every time.

Keep exploring

More in Data Analysis