KPI drop diagnosis and action plan
Investigate why a key metric moved: a structured hypothesis tree, the data cuts to run, and a prioritized plan to confirm and fix.
Turn a business goal into a short set of well-defined KPIs, with formulas, data sources, guardrails and the traps to avoid.
Help me define the right metrics for this goal.
Goal: {{goal}}
Business or team context: {{context}}
Data I can realistically get: {{data_sources}}
Reporting cadence: {{cadence}}
Deliver:
1. One north-star metric and why it reflects the goal better than the obvious alternatives.
2. Three to five supporting KPIs, split into leading indicators (move first, controllable) and lagging indicators (outcomes).
3. A definition table with these columns: Metric, Precise formula (numerator, denominator, time window), Data source, Owner role, Target-setting approach, Cadence.
4. Two guardrail metrics that would reveal if we hit the main KPI by harming something else.
5. Ways each KPI can be gamed or misread (vanity effects, denominator games, Simpson's paradox, seasonality) and how to prevent that.
6. What I should not measure yet and why.
7. A suggested first dashboard layout: what goes at the top, what trends, what segments.
Be specific; avoid generic KPIs unless you explain why they apply here. If my goal is fuzzy, ask up to three questions to sharpen it, but still propose a first version.Precise formulas with numerator, denominator and window avoid the metric definition arguments that derail reporting. Add your current numbers to get target suggestions.
Licensed CC0 (public domain) Version 1 · history
Investigate why a key metric moved: a structured hypothesis tree, the data cuts to run, and a prioritized plan to confirm and fix.
Turn a dataset description into a step-by-step exploration plan with checks, hypotheses and the first charts to make.
Set up a proper cohort retention analysis, from defining cohorts and the activity event to producing the table and reading it correctly.
Convert a stakeholder's fuzzy ask into a scoped analysis brief with the real question, definitions, data needed and success criteria.
Systematically diagnose why a metric suddenly changed, ranking likely causes and giving the exact checks to confirm each one.
Condense a long, messy ticket history into a crisp escalation note with impact, timeline, what has been tried and what you need.
Comments (0)
Sign in to join the discussion.
No comments yet. Be the first to share how it worked.