KPI vs Metric
A metric is any measurable number a business tracks — clicks, signups, page views, latency, response time. A KPI (key performance indicator) is a specific subset: the metrics that have been chosen as the most important indicators of progress toward strategic goals. Every KPI is a metric, but not every metric is a KPI.
Last reviewed on 2026-04-27.
Quick Comparison
| Aspect | KPI | Metric |
|---|---|---|
| Definition | A specifically chosen metric tied to a strategic goal | Any measurable number |
| Scope | Small set — perhaps 3–10 per team | Many — dozens or hundreds |
| Purpose | Drive decisions about strategic outcomes | Provide visibility, feedback, debugging |
| Frequency of review | Regular cadence — weekly, monthly, quarterly | On-demand or via dashboards |
| Selection | Deliberate; ideally tied to outcomes | Often emerges from instrumentation |
| Risk | Wrong KPIs steer the team in wrong directions | Too many metrics overwhelm; bad signal-to-noise |
| Examples | Monthly active users, gross margin, NPS | Page views, error rate, support ticket volume |
Key Differences
1. Subset versus universe
KPIs are a deliberately small set of metrics — typically a handful per team — chosen because they best reflect progress toward strategic goals.
Metrics are everything you measure. A web team might have hundreds of metrics across performance, traffic, conversion, errors, support — most aren't KPIs.
2. Selection versus emergence
KPIs are chosen. Someone decides this is the key indicator we'll watch. The choice is strategic and reviewed periodically.
Metrics often emerge from instrumentation. You build a system; it produces logs and counts; many metrics exist whether anyone's looking at them or not.
3. How they're used
KPIs drive decisions. They appear on executive dashboards, in board updates, in OKR reviews. Movement in a KPI triggers conversation about strategy and execution.
Metrics often live on operational dashboards. Engineers watch latency; support watches ticket volume. Many metrics are never escalated to executive attention because they're working as expected.
4. Choosing the right KPIs
A good KPI is: tied to a strategic goal, measurable consistently, sensitive enough to move with team actions, and not so easy to game that it becomes meaningless.
A bad KPI is one of: too easy to game (vanity metrics), too coarse to act on, too slow to move, or measuring the wrong thing entirely. "Total page views" is often a vanity metric; "engaged session" might be more meaningful.
5. How many KPIs is right
Too few KPIs mean blind spots. Too many mean none of them get sustained attention.
A common rule of thumb: 3–7 KPIs at the team level. The exact number depends on the team's scope and the strategic clarity of the goals.
6. KPIs change as strategy changes
Strategic shifts often require new KPIs. A company moving from "growth at all costs" to "unit economics" replaces growth-only KPIs with margin and retention measures.
Metrics often outlive strategies. The system is still measuring page views and error rates whether or not those numbers are KPIs this year.
When to Choose Each
Choose KPI if:
- Communicating priorities to executives, the board, and the whole team.
- Driving accountability against strategic goals.
- Reviewing performance at quarterly business reviews.
- Anywhere a few key numbers should drive attention and action.
Choose Metric if:
- Operational dashboards for engineering, support, marketing.
- Debugging and diagnosis when things go wrong.
- Detailed performance analysis.
- Anywhere visibility into details supports daily work without driving strategy.
Worked example
A SaaS company tracks dozens of metrics: page views, signup conversion rate, signup-to-active conversion, monthly active users, churn rate, support ticket volume, NPS, page latency, error rates, and so on. From this, the executive team selects four KPIs for the year: monthly active users, net revenue retention, gross margin, and customer satisfaction. All four are metrics; only those four are KPIs. The rest are watched at the team level for daily operations.
Common Mistakes
- "More metrics is always better." Adding metrics adds noise and dilutes attention if not framed within KPIs.
- "Every metric is a KPI." KPIs are a deliberately small subset.
- "KPIs and OKRs are the same." Different frameworks. OKRs are objectives and key results; KPIs are continuous indicators of performance. They can complement each other.
- "Set KPIs once and stick to them." Strategies change; KPIs should evolve to match.