KPI Examples: How Data Analysts Define KPIs That Drive Decisions

A KPI (key performance indicator) is a metric a team has agreed to watch because it shows whether an important goal is on track. Plenty of numbers can be measured; a KPI is the small set that is tied to a decision, has a target, and has an owner who acts when it moves. For data analysts, defining KPIs well matters more than calculating them: a precise query on a badly chosen metric still leads a team in the wrong direction.

This guide covers what separates a KPI from an ordinary metric, practical KPI examples by team, a step-by-step way to define one, and the mistakes that make dashboards full of numbers nobody uses.

KPI vs metric: what is the difference?

Every KPI is a metric, but most metrics are not KPIs. A metric is any quantified measurement, such as page views, tickets closed, or rows loaded. A performance indicator becomes "key" when it is directly connected to a business objective and someone is accountable for it.

• Metric: number of website sessions this week.

• KPI: trial-to-paid conversion rate, with a target, reviewed weekly by the growth lead.

A useful test: if the number dropped sharply tomorrow, would someone change what they are doing? If not, it is a supporting metric, not a KPI.

KPI examples by team

The right KPIs depend on the business model, but these examples show the pattern of linking each number to a goal.

Sales and revenue

• Monthly recurring revenue (MRR): recurring revenue normalized to a month, for subscription businesses.

• Win rate: deals won divided by deals that reached a decision.

• Average deal size: revenue from won deals divided by the number of won deals.

Marketing and growth

• Customer acquisition cost (CAC): sales and marketing spend divided by new customers in the same period.

• Conversion rate: the share of visitors or leads who complete a defined action, such as starting a trial.

• Payback period: months it takes for a customer's gross margin to cover their acquisition cost.

Product

• Activation rate: the share of new users who reach a first meaningful action within a set time.

• Retention by cohort: the share of each signup group still active after N weeks, best read with cohort analysis.

• Feature adoption: active users of a feature divided by all active users.

Operations and customer support

• First response time: median time until a customer gets a first human reply.

• On-time delivery rate: orders delivered by the promised date divided by all orders delivered.

• Customer satisfaction (CSAT): share of survey responses rated satisfied or better.

How to define a KPI step by step

Step 1: Start from the objective, not the data

Write the goal in one plain sentence: "Increase the number of trial users who become paying customers." Analysts often start from what is easy to query. Start instead from what the business is trying to change, then look for data that represents it.

Step 2: Choose a leading and a lagging indicator

Lagging indicators, such as revenue or churn, confirm results but arrive late. Leading indicators, such as activation in the first week, move earlier and give the team time to react. A healthy KPI set pairs at least one of each for an important goal.

Step 3: Write an exact definition

Most KPI arguments are really definition arguments. Document:

• Formula: numerator and denominator, written out.

• Inclusions and exclusions: test accounts, refunds, internal users, free plans.

• Time window: calendar week, rolling 28 days, or cohort-based.

• Grain and source: which table, at what level (user, account, order), and which system is the source of truth.

For example, "conversion rate" could mean paid accounts divided by trials started, or paid accounts divided by trials that ended. Both are valid; mixing them across teams is not.

Step 4: Set a baseline and a target

Calculate the KPI on historical data first, so the target is grounded in where the number actually is. Many teams use the SMART criteria (specific, measurable, achievable, relevant, time-bound) as a quick check that a target is usable.

Step 5: Assign an owner and a review rhythm

A KPI without an owner becomes decoration. Name who is accountable, how often the number is reviewed, and what happens when it crosses a threshold, such as a review meeting or an alert.

Step 6: Add guardrail metrics

Any single number can be gamed. Goodhart's law summarizes the risk: when a measure becomes a target, it stops being a good measure. Pair each KPI with a guardrail. If support is measured on response time, also track satisfaction, so the team does not answer quickly and badly.

A simple KPI definition template

Name:          Trial-to-paid conversion rate
Objective:     Grow paying customers from self-serve trials
Formula:       accounts that paid within 30 days of trial start
               / trials started in the period
Exclusions:    internal and test accounts, duplicate trials
Window:        weekly, by trial start date (30-day lag)
Source:        billing.subscriptions + product.trials
Owner:         Growth lead
Target:        set from the last quarter's baseline
Guardrail:     refund rate in the first 60 days

Name:          Trial-to-paid conversion rate
Objective:     Grow paying customers from self-serve trials
Formula:       accounts that paid within 30 days of trial start
               / trials started in the period
Exclusions:    internal and test accounts, duplicate trials
Window:        weekly, by trial start date (30-day lag)
Source:        billing.subscriptions + product.trials
Owner:         Growth lead
Target:        set from the last quarter's baseline
Guardrail:     refund rate in the first 60 days

Name:          Trial-to-paid conversion rate
Objective:     Grow paying customers from self-serve trials
Formula:       accounts that paid within 30 days of trial start
               / trials started in the period
Exclusions:    internal and test accounts, duplicate trials
Window:        weekly, by trial start date (30-day lag)
Source:        billing.subscriptions + product.trials
Owner:         Growth lead
Target:        set from the last quarter's baseline
Guardrail:     refund rate in the first 60 days

Name:          Trial-to-paid conversion rate
Objective:     Grow paying customers from self-serve trials
Formula:       accounts that paid within 30 days of trial start
               / trials started in the period
Exclusions:    internal and test accounts, duplicate trials
Window:        weekly, by trial start date (30-day lag)
Source:        billing.subscriptions + product.trials
Owner:         Growth lead
Target:        set from the last quarter's baseline
Guardrail:     refund rate in the first 60 days

A one-page definition like this prevents most of the "why does your number not match mine?" conversations.

Common KPI mistakes

• Too many KPIs. If everything is key, nothing is. Keep a handful per team and move the rest to supporting views.

• Vanity metrics. Totals that only grow, such as all-time signups, feel good but rarely drive a decision. Prefer rates and cohort views.

• Ratios without volume. A conversion rate jumping from 2% to 4% on twenty users is noise. Show the denominator next to the rate.

• Changing definitions silently. If the formula changes, version it and annotate the dashboard.

• Reading movement as cause. A KPI going up after a launch does not prove the launch caused it. Use experiments such as A/B testing when you need causal evidence.

How to present KPIs on a dashboard

Put the few KPIs at the top, each with its target, the previous period, and a trend line. Supporting metrics that explain movement go below. Keep definitions one click away. For layout choices, see these dashboard design principles for data analysts.

Frequently asked questions

How many KPIs should a team have?

There is no fixed rule, but most teams work best with roughly three to seven. Beyond that, attention spreads thin and reviews turn into status reports.

What is the difference between a KPI and an OKR?

An OKR (objective and key results) is a goal-setting framework for a period. KPIs are ongoing health measures. A key result is often a target on a KPI for a quarter.

Who should define KPIs, the analyst or the business?

Both. The business owns the objective and the target. The analyst owns the precise definition, the data source, and pointing out when a proposed KPI cannot be measured reliably.

Learn to turn business questions into metrics

If you want to practice framing business questions, defining metrics, and analyzing real data with feedback from practitioners, Coderhouse offers live online courses with one class per week:

• Data Analytics Course: 11 weeks for career switchers, analysts, and data-driven professionals, starting with how to frame ambiguous business questions and define metrics, then SQL, Python, visualization, and experimentation.

• AI Engineering Course: 8 weeks for experienced developers moving into AI engineering, covering LLM applications, retrieval, agents, evaluation, and deployment.