PM KPI Guide
(2026 Edition)
6 KPI types every PM should know, a 6-point checklist for choosing KPIs, and the 6 common KPI mistakes that lead to well-measured failure.
Build Metric Judgment Daily — Free →6 Types of KPI
1. Outcome KPI
What you actually care about — user value or business impact delivered.
💡 Example: Users completing 3 lessons in first week (activation outcome).
2. Output KPI
Things you shipped. Easier to measure but weaker signal of value.
💡 Example: Lessons launched this quarter, features shipped.
3. Leading Indicator
Predicts future outcomes. Moves before the lagging indicator.
💡 Example: Day-1 retention predicts Day-30 retention 3 weeks out.
4. Lagging Indicator
Measures past outcomes. Authoritative but slow to react.
💡 Example: Annual Revenue, D90 retention — real truth, delayed.
5. Guardrail KPI
Must NOT degrade, even if primary KPI wins.
💡 Example: Churn rate, crash rate, support ticket volume — when optimising conversion.
6. Vanity KPI
Looks impressive but doesn't drive decisions. Avoid as primary metrics.
💡 Example: Page views, total signups, followers — often gamed or decoupled from value.
6-Point Checklist for Choosing a KPI
Does this metric move when users get real value? (If not, it's a vanity metric.)
Can it be gamed without creating value? (If yes, add guardrails.)
Is it measurable with the instrumentation we have? (If not, it's aspirational, not real.)
Does the team agree on the definition? (Ambiguous definitions create false alignment.)
Does it respond quickly enough to drive decisions? (Metrics that only update quarterly are poor KPIs.)
Is there a clear path from PM decisions to this metric? (If actions don't affect it, it's not yours.)
6 Common KPI Mistakes
Picking output KPIs because they're easy ('features shipped') instead of outcomes ('metric moved')
Only tracking the primary KPI — guardrails catch regressions that primary misses
Changing KPIs every quarter — if it wasn't a real strategy, it's a rotating target
Defining KPIs loosely ('engagement') — ambiguity kills alignment
Falling into Goodhart's Law — team games the metric without creating value
No segmentation — aggregate KPIs hide cohort-level problems
FAQ
What's the difference between KPI and north star metric?
The north star is your most important KPI — the one that captures your core value delivery. Other KPIs (input metrics, guardrails, operational metrics) support it. All north stars are KPIs, but not all KPIs are north stars. Most product teams have 1 north star + 5–10 supporting KPIs.
How many KPIs should a PM track?
5–10 for a focused product area. More than 15 and you're not tracking — you're reporting. The PM who tracks 30 metrics weekly is often less effective than the one who tracks 7 religiously. Depth in few beats breadth across many.
What's the biggest KPI mistake?
Optimising for what's measurable rather than what matters. Page views are easy to measure; genuine value delivered is hard. PMs who default to measurable but shallow metrics end up optimising for activity rather than outcomes. Always start from: 'what does value look like to the user?' and work backwards to metrics — not the other way around.
Build Metric Intuition in 2 Minutes a Day
Daily scenarios on KPI choice, diagnosis, and guardrail design.
Start Free Trial →