Business Intelligence Foundations

What Are KPIs? Key Performance Indicators Explained

A practical guide to what KPIs are, how they differ from ordinary metrics, how to choose them and how they turn business data into a clear signal about performance.

By Adam Finer Updated July 2026 18-minute read

Businesses can measure almost anything: website visits, calls answered, orders placed, hours worked, revenue generated and hundreds of other values.

But measuring something does not automatically make it important. A key performance indicator, or KPI, identifies the measures that tell an organisation whether it is making meaningful progress toward an objective.

A KPI is a clearly defined, measurable indicator used to assess progress toward a specific business objective. It normally includes not only a calculation, but also a target, a time period, an owner and enough context to show whether performance is on track.

In simple terms, a KPI turns a number into a performance signal.

Business owners and managers
Aspiring BI analysts
Teams building dashboards
Students learning data and analytics
Adam Finer, founder of Learn BI Academy
Adam Finer Founder, Learn BI Academy

Adam teaches business intelligence, SQL, dashboard design and performance reporting. His approach to KPIs begins with the business objective and decision, not with a list of fashionable measures.

What does KPI actually mean?

KPI stands for key performance indicator. Each word matters:

Key

It must matter

The measure should be important enough to influence a real objective, priority or decision. If nobody would act when it changes, it may not be key.

Performance

It must assess progress

The value should help show whether a team, process, department or organisation is performing as intended.

Indicator

It is a signal, not the full explanation

A KPI highlights where attention may be required. Further analysis is often needed to understand why the result changed.

Together

A KPI connects data to a goal

It tells decision-makers whether performance against something important is improving, declining or remaining on track.

A KPI should not merely tell you what happened. It should tell you whether what happened matters.

Revenue, conversion rate and customer acquisition cost can all be KPIs, but none is automatically a KPI in every organisation. Their status depends on the objective being pursued.

For a company focused on profitable growth, revenue growth and profit margin may be key. For a support team trying to improve service, first-response time and customer satisfaction may matter more.

Why are KPIs important?

KPIs create focus. They help organisations reduce a large volume of available data to the smaller set of signals that decision-makers need to monitor.

1

Translate strategy into measures

KPIs connect broad objectives such as growth, efficiency or customer retention to observable results.

2

Align teams

Shared definitions and targets help people work toward the same interpretation of success.

3

Detect problems earlier

Regular monitoring can show when performance moves away from a target before the issue becomes harder to correct.

4

Support better decisions

Decision-makers can compare current performance with targets, previous periods and relevant benchmarks.

KPIs do not replace judgement. They provide structured evidence that helps people decide where to investigate and what action may be required.

KPIs vs metrics: what is the difference?

A metric is any measurable value. A KPI is a metric selected because it represents progress toward an important objective.

Metric KPI
Measures an activity, result or characteristic Measures progress toward a priority or objective
May be useful for context or diagnosis Should be important enough to trigger attention or action
May not require a formal target Normally needs a target, benchmark or expected range
Can be one of hundreds of available measures Should form part of a deliberately limited set

All KPIs are metrics, but not all metrics are KPIs.

Example: website traffic

Website sessions may be a useful metric. If the business objective is to increase qualified demand, however, the true KPI may be qualified leads generated, lead-to-customer conversion rate or revenue attributed to the website. Traffic helps explain activity; the KPI assesses the result that matters.

Metrics should not be treated as unimportant. Supporting metrics often help explain why a KPI moved. A KPI might show falling conversion rate, while supporting metrics reveal that the decline is concentrated on mobile devices or a particular landing page.

To understand how measures are combined and broken down inside reports, see What Are Metrics and Dimensions?

The anatomy of a well-defined KPI

A KPI is more than a name displayed on a scorecard. A usable KPI needs a complete definition.

ObjectiveWhat business result is the organisation trying to achieve?
KPI nameA clear, unambiguous description understood by its users.
FormulaThe exact calculation, including inclusions, exclusions and treatment of unusual cases.
Data sourceThe systems and fields used to calculate the value.
Target or thresholdThe desired result, acceptable range or point at which attention is required.
Time periodWhether the KPI is assessed daily, weekly, monthly, quarterly or over another period.
ComparisonTarget, previous period, forecast, benchmark or another meaningful reference point.
OwnerThe person or team responsible for monitoring the KPI and coordinating a response.
Review frequencyHow often performance and the continued relevance of the KPI are reviewed.

Without these details, two departments can use the same KPI name while calculating different values. “Customer churn,” for example, can vary depending on how a customer is defined, which period is used and whether reactivated customers are included.

Adam's advice

Create a KPI definition sheet before building the dashboard. Agreeing the formula and ownership in advance is much easier than discovering conflicting definitions during a management meeting.

Common types of KPIs

KPIs can be classified in several useful ways. These categories overlap: a KPI can be strategic, financial and lagging at the same time.

Strategic and operational KPIs

Strategic KPIs

High-level indicators used to assess progress toward organisational priorities. Examples include revenue growth, operating margin and customer retention.

Operational KPIs

More frequent measures used to monitor processes and day-to-day performance. Examples include fulfilment time, defect rate and support response time.

Leading and lagging indicators

Leading KPIs

Signals that may indicate future results. They often measure activities or conditions that influence an outcome, such as qualified pipeline value, trial activation or preventive maintenance completed.

Lagging KPIs

Measures of outcomes that have already occurred, such as revenue, profit, customer churn or employee turnover.

A balanced KPI set often includes both. Lagging indicators show whether the outcome was achieved; leading indicators help teams influence what happens next.

Financial and non-financial KPIs

Financial KPIs include revenue, gross margin, operating costs and cash conversion. Non-financial KPIs include customer satisfaction, product quality, delivery time, staff retention and website conversion.

Input, process, output and outcome KPIs

  • Input KPIs measure resources used, such as advertising spend or training hours.
  • Process KPIs measure how work is performed, such as cycle time or error rate.
  • Output KPIs measure what was produced, such as units completed or leads generated.
  • Outcome KPIs measure the wider result, such as profit, retention or customer lifetime value.

KPI examples by department

The right KPIs depend on the organisation's objectives, business model and stage of development. The examples below are common possibilities, not a universal list.

Sales

Revenue growth, win rate, average deal size, sales-cycle length, qualified pipeline coverage and customer acquisition.

Marketing

Cost per qualified lead, lead-to-customer conversion, return on advertising spend, organic conversions and marketing-sourced revenue.

Finance

Gross profit margin, operating margin, cash runway, budget variance, debtor days and recurring revenue.

Customer service

First-response time, resolution time, repeat-contact rate, customer satisfaction and support backlog.

Operations

Order fulfilment time, on-time delivery, capacity utilisation, defect rate, downtime and cost per unit.

People and HR

Employee turnover, time to hire, absence rate, internal promotion rate and employee engagement.

A worked ecommerce example

Suppose an ecommerce company has the objective of increasing profitable online growth.

ObjectiveGrow online profit without inefficient acquisition spending
Outcome KPIContribution profit from online sales
Supporting KPIsConversion rate, acquisition cost and average order value
DimensionsChannel, campaign, customer type, product and device
DecisionWhere to invest, optimise or reduce spending

Website traffic and number of orders are still useful metrics, but they do not alone show whether growth is profitable. The complete KPI structure keeps the analysis connected to the objective.

How to choose the right KPIs

The most common selection error is beginning with available data rather than the business objective. A reporting tool can calculate thousands of measures. That does not mean the organisation should monitor all of them.

Define the objective

State clearly what should improve and why it matters.

Identify the decision

Clarify who will use the KPI and what decision or action it should support.

Select the outcome measure

Choose the strongest indicator that the desired result has been achieved.

Add leading and diagnostic measures

Include a small number of measures that help influence or explain the outcome.

Define the calculation

Document the formula, data source, frequency and business rules.

Set the target or benchmark

Specify what good performance looks like and how it was determined.

Assign ownership

Make clear who monitors the KPI and who can influence the result.

Review and refine

Remove KPIs that no longer support a relevant objective and update definitions when the business changes.

A practical KPI test

  1. Which objective does this measure support?
  2. Who uses it and what decision does it influence?
  3. Is the calculation understood and trusted?
  4. Is there a meaningful target or comparison?
  5. Can someone influence the result?
  6. Would a significant change trigger investigation or action?

If those questions cannot be answered, the measure may be interesting, but it is probably not yet a properly defined KPI.

How should KPI targets and benchmarks be set?

A KPI without context is difficult to interpret. Revenue of €500,000 may be excellent, disappointing or exactly as expected depending on the target, time period and size of the organisation.

Common comparisons include:

  • Target: the desired level of performance.
  • Previous period: the prior week, month, quarter or year.
  • Budget or forecast: the result expected in a financial or operational plan.
  • Baseline: the level recorded before a change or initiative.
  • Benchmark: an external standard, industry reference or comparable group.
  • Acceptable range: upper and lower thresholds within which performance is considered normal.

Targets should be challenging enough to guide improvement but credible enough to remain meaningful. An arbitrary target can create misleading conclusions or encourage undesirable behaviour.

How to report KPIs effectively

KPIs are commonly presented through scorecards, management reports and business intelligence dashboards.

A useful KPI display should normally show:

  • the current value
  • the target, threshold or benchmark
  • the variance from that comparison
  • the trend over time
  • the relevant reporting period
  • the last refresh date
  • filters or breakdowns needed for investigation

Adam's advice

Do not stop at a large number with a red or green arrow. Give the user enough context to understand the movement and enough interactivity to investigate it.

Colour should support interpretation, not replace it. Red and green can be useful for thresholds, but reports should also include labels, symbols or variance values so meaning does not depend on colour alone.

The role of KPIs in business intelligence

KPIs sit at the centre of the business intelligence process because they define what performance means.

Sales
Finance
Marketing
CRM
Operations
Prepare and model reliable data
Calculate agreed KPIs
Compare performance with targets and trends
Investigate causes and decide what to do

Data pipelines keep the information current. Data models structure it. Dashboards present it. KPIs decide which signals deserve attention.

A business intelligence analyst often helps stakeholders translate broad goals into precise definitions, calculations and reporting requirements. The analyst should not invent business priorities alone; KPI selection requires collaboration with the people responsible for the objective.

Common KPI mistakes

Tracking too many KPIs

When everything is labelled key, nothing is key. Use supporting metrics without promoting every measure to KPI status.

Choosing what is easy to measure

Available data may not represent the result that matters. Begin with the objective rather than the dashboard fields.

No target or comparison

A value alone rarely shows whether performance is good, bad or expected.

Unclear definitions

Ambiguous formulas create conflicting figures and reduce trust in the reporting environment.

Only using lagging measures

Outcome KPIs reveal what happened but may provide little guidance about what can still be influenced.

Ignoring unintended behaviour

Targets can encourage people to optimise the measure at the expense of the underlying objective.

No clear owner

A KPI can remain permanently red when nobody is responsible for interpreting and responding to it.

Never reviewing the KPI set

A measure can remain on a dashboard long after the business priority that justified it has changed.

Beware of vanity metrics

Vanity metrics look impressive but provide little evidence of meaningful progress. Social followers, raw traffic or app downloads may be useful context, but they become dangerous when presented as success without linking them to engagement, customers, revenue or another genuine outcome.

What are the limitations of KPIs?

KPIs simplify complex performance into a manageable set of measures. That is useful, but simplification also creates risks.

  • A KPI can show that performance changed without explaining the cause.
  • Poor-quality data can produce a precise but incorrect value.
  • A narrow target can encourage short-term behaviour that damages long-term outcomes.
  • Aggregated results can hide differences between products, customers, regions or teams.
  • External events may affect results in ways the responsible team cannot control.
  • A KPI can become outdated as strategy and operating conditions change.

This is why KPIs should be treated as signals for discussion and investigation, not unquestionable verdicts.

How AI is changing KPI monitoring

Artificial intelligence can help analysts document calculations, detect unusual movements, generate narrative summaries and identify which dimensions contributed most to a change.

For example, an AI-assisted reporting process could explain that revenue fell primarily because repeat-customer orders declined in one region, rather than simply stating that revenue was below target.

AI does not decide what should matter to the business.

It can calculate, compare and explain a KPI only after the objective, definition, data and business context have been established.

AI can also produce confident but incorrect explanations when data or definitions are unreliable. Human review, governance and transparent KPI logic remain essential. Read more in How AI Is Changing Business Intelligence.

The core idea

A KPI is not simply an important-looking number.

It is an agreed performance signal that connects a measurable result to an objective, a target, an owner and a decision.

Frequently asked questions

Questions about KPIs

What does KPI stand for?

KPI stands for key performance indicator.

What is a KPI in simple terms?

A KPI is a measurable signal that shows whether a business, team or process is making progress toward an important objective.

What is the difference between a KPI and a metric?

A metric is any measurable value. A KPI is a metric deliberately selected because it represents progress toward a priority or objective. All KPIs are metrics, but not all metrics are KPIs.

What are examples of KPIs?

Examples include revenue growth, gross profit margin, customer retention, sales conversion rate, on-time delivery, first-response time and employee turnover. The right KPI depends on the organisation's objective.

What makes a good KPI?

A good KPI is relevant to an objective, clearly defined, based on reliable data, compared with a target or benchmark, assigned to an owner and capable of supporting a decision or action.

How many KPIs should a business have?

There is no universal number. Each dashboard, team or objective should use a deliberately limited set that keeps attention on what matters while retaining supporting metrics for diagnosis.

What are leading and lagging KPIs?

Leading KPIs indicate activities or conditions that may influence future performance. Lagging KPIs measure outcomes that have already occurred. A balanced KPI set normally includes both.

Are KPIs the same as OKRs?

No. An OKR combines an objective with a set of key results used to describe desired progress. KPIs are ongoing measures of performance and may be used inside or alongside an OKR framework.

Where to Go Next

Continue with the concepts and skills most closely connected to KPI reporting.

What Is Business Intelligence?

See how organisations bring data together and use KPI dashboards to monitor performance and support decisions.

What Are Metrics and Dimensions?

Understand how measures are calculated and broken down inside business intelligence reports.

7 Business Intelligence Terms Every BI Beginner Should Know

Build the vocabulary needed to understand data models, dashboards and reporting projects.

How to Learn Business Intelligence

Follow a structured route through business concepts, SQL, modelling and dashboard creation.

The Ultimate Guide to Looker Studio

Learn how to turn metrics and KPIs into clear, interactive reports.

BI Analyst Starter Program

Develop practical skills in SQL, data modelling, dashboards and business-focused portfolio projects.