Key Performance Indicator (KPI)
What Is a Key Performance Indicator (KPI)?
A key performance indicator (KPI) is a measure selected to assess performance against an important product or business objective. It gives a team a consistent way to judge whether a result is improving, deteriorating, or staying within an acceptable range, and helps identify when a decision or investigation is needed.
Not every number on a dashboard is a KPI. The word “key” reflects a deliberate choice about what matters for a particular goal, team, and stage of the product. The same measure can be essential to one decision and background context for another.
What Makes a Metric a KPI?
A metric measures something. A KPI connects that measurement to an important performance question and a responsibility to review it. Total exports is a metric; the proportion of eligible accounts completing a required monthly report might be a KPI for a reporting product.
A usable definition makes the following explicit:
- Purpose: the objective or decision the indicator supports.
- Calculation: what is counted, the denominator, and any exclusions.
- Population and period: which users or accounts are included and over what window.
- Reference point: the baseline, target, or acceptable range used to interpret the result.
- Ownership: who checks the data and responds when the indicator changes.
For example, “activation” is an incomplete definition. “Percentage of new workspaces that connect a data source and share a report within seven days of creation” is specific enough to implement and discuss. The team still needs to agree how test accounts, deleted workspaces, and late events are handled.
KPIs, OKRs, and a North Star Metric
OKRs combine an objective with measurable key results for a defined period. A KPI can supply a key result's measure: a team might monitor activation continuously, then commit to improving it from a stated baseline to a target this quarter. The ongoing indicator and the time-bound improvement goal serve different purposes.
A north star metric is a central measure intended to represent the value a product delivers. A KPI set can include that measure, its inputs, and other important results or guardrails. A product does not need to force every performance question into one number.
KPI Example
Consider a hypothetical invoicing product. Its team wants customers to establish a repeat billing habit. It selects the share of new paying accounts that send a second invoice within 30 days as an early indicator, alongside later account retention.
A rising second-invoice rate may be encouraging, but the team checks whether the customer mix changed or an automatic test invoice inflated the count. It uses product analytics to inspect the underlying behavior before deciding that onboarding improved.
Interpreting an Indicator
A leading indicator is used as an earlier signal of a later result; a lagging indicator records a result that has already occurred. That relationship must be checked in the product's context. An action that correlates with retention is not automatically its cause.
Review whether the indicator still answers the intended question as the product changes. Keep definitions stable enough for comparison, and document material changes to the calculation. A favorable KPI should prompt an informed decision, rather than replace judgment about customer value, data quality, or unintended effects.