AI Prompt for Choosing KPIs

Choose the right KPIs with this AI prompt. A framework that picks metrics that actually drive business decisions. With fill-in variables and real examples.

The prompt

ROLE: You are a business intelligence consultant who helps companies escape vanity metrics.

CONTEXT: Business: [BUSINESS], [STAGE]. Business model: [MODEL]. Strategic goals: [GOALS]. Metrics currently tracked: [CURRENT_METRICS]. Decisions KPIs must support: [DECISIONS].

TASK: Recommend the right KPI set.

FORMAT: 1. KPI Framework explainer (input vs output vs outcome metrics, 3 lines each). 2. Recommended KPIs: 8-12 total, grouped (Growth, Revenue, Efficiency, Customer, Team), each with: definition, why it matters (one line), target-setting guidance. 3. Vanity Metrics to Drop (from my current list, with why each misleads). 4. KPI Tree: show how lower-level metrics roll up to the top 3 business outcomes. End with the one metric you would watch if you could only pick one, and why.

CONSTRAINTS: Every KPI must connect to a decision; metrics nobody acts on are decoration. Max 12 KPIs; more means none get attention. Each definition must be unambiguous enough that two people compute it identically.

Variables to fill in

Replace each [variable] in the prompt with your own details:

[BUSINESS] [STAGE] [MODEL] [GOALS] [CURRENT_METRICS] [DECISIONS]

Example input

BUSINESS=DTC subscription coffee, STAGE=Series A; MODEL=$24/month subscription; GOALS=profitable growth; CURRENT_METRICS=followers, site visits, revenue; DECISIONS=marketing spend, pricing, hiring

How to use this prompt

List every metric you track today, run the prompt, then cut your dashboards down to the recommended set within two weeks.

Tips for the best results

If a metric never changed a decision in 6 months, delete it; dashboards accumulate cruft.
Pair every growth KPI with a quality KPI (e.g., signups with retention) to prevent gaming.
Review the KPI set annually; the right metrics for finding product-market fit differ from those for scaling.