KPI vs. Metric: What's the Difference?

KPI vs metric is a question that sounds like semantics and affects what ends up on your dashboard. A metric is any number you measure. A KPI is a metric that you have chosen, because it tracks progress toward an important goal, and that you manage with a target, an owner and a review. All KPIs are metrics, and most metrics are not KPIs. Confusing the two leads to dashboards crowded with numbers and no sense of what matters. This guide explains the difference, shows how a metric earns KPI status, works through an e-commerce funnel with checked numbers to separate KPIs from supporting metrics, offers a decision rule and lists the mistakes that follow from mixing them up. It belongs to business analytics for finance and operations, follows what are KPIs and 25 financial KPIs, and leads to how to choose KPIs that drive decisions and KPI scorecards. The distinction matters in budgeting and forecasting, where only a few measures carry targets.

The difference at a glance

A useful phrase: metrics tell you what is happening, KPIs tell you whether you are winning.

How a metric becomes a KPI

A metric earns KPI status by passing a few tests.

1. It connects to an objective the leadership cares about. If improving it would not matter to the goals, it is a metric.
2. A target can be set. You can say what good looks like and by when.
3. Someone can influence it. The owner can name actions that would move it.
4. It is reliable. The definition is written down and the data is trusted.
5. It drives a decision or action. Changes in it lead to responses.

If a metric fails the tests, keep it as a supporting number. Many metrics are valuable for diagnosis even though they never become KPIs. When the KPI moves, you look at the supporting metrics to find out why.

A worked funnel

An online business wants to grow profitable revenue. It tracks the following over a month (hypothetical):

Check: 100,000 × 0.08 = 8,000 add-to-carts; 8,000 × 0.3125 = 2,500 orders; 2,500 / 100,000 = 2.5 percent; 2,500 × 60 = $150,000; 2,500 × 24 = $60,000; 30,000 / 1,500 = $20 per new customer.

All of these are metrics. Which are KPIs? Think of the objective: grow profitable revenue. The KPIs might be contribution after marketing ($60,000 - $30,000 = $30,000), customer acquisition cost ($20) compared with first-order contribution, and repeat purchase rate (not in the table). They are tied to the objective, they have targets, and managers can influence them. Sessions, add-to-cart rate and checkout completion are supporting metrics. They are not unimportant: if conversion falls, the funnel metrics tell you whether the problem is traffic quality, product pages or checkout. But putting all twelve on the leadership dashboard buries the three that matter.

Notice also that conversion rate sits in between. For the marketing team it might be a KPI, since they influence it directly through landing pages and traffic mix, and for the executive team it is a supporting metric. Whether a number is a KPI depends on whose objectives you are looking at. A hierarchy helps: company KPIs, department KPIs and supporting metrics, each linked to the one above. See KPI scorecards.

The cost of confusion

Treating every metric as a KPI has real costs. Suppose a leadership team reviews 40 numbers in a monthly meeting and spends an average of 90 seconds on each: that is 60 minutes, about the whole meeting, with nothing left for decisions. A team that reviews 8 KPIs for 3 minutes each and keeps the supporting metrics for exceptions spends 24 minutes on KPIs and has half an hour for the actions that follow. The difference in the use of senior time over a year, at 12 meetings, is small in hours and large in effect.

The opposite mistake is to track only KPIs and discard the supporting metrics. When a KPI turns red, you need the metrics beneath it to diagnose. Keep them available in a drill-down or an appendix.

A decision rule for the dashboard

For each number on a dashboard, ask three questions.

1. Does it have a target and an owner? If yes, it is a KPI, and it belongs in the top section.
2. If it does not, does it help explain a KPI when the KPI moves? If yes, it is a supporting metric, and belongs in a drill-down view.
3. If neither, why is it there? Remove it unless a regulation or a contract requires it.

Review the dashboard once a year with this rule. See how to choose KPIs that drive decisions.

Related terms

Common mistakes

• Calling everything a KPI.
• KPIs without targets, owners or review.
• Vanity metrics as KPIs. Followers, page views and downloads that do not link to revenue or profit.
• KPIs the owner cannot influence.
• Dropping supporting metrics.
• Using the same KPIs at every level.

Metrics that are often mistaken for KPIs

Some numbers are popular and easy to produce, which makes them tempting candidates. Social media followers, website visits, downloads, tickets opened and hours logged are all metrics. They become KPIs only when a team can show how they link to a goal and a decision. A company with 50,000 email subscribers has a large list, and the question for the marketing team is what the list contributes: revenue per subscriber, the rate of purchase per campaign, the cost to acquire a subscriber. Those are closer to KPIs because they connect to money. The size of the list is context. Ask of every popular number: if this doubled and nothing else changed, would the business be better off?

A hierarchy across levels

A good set of KPIs forms a hierarchy. At the top, the company has a few KPIs such as operating margin, cash conversion cycle and customer retention. Below, each department has KPIs that drive those: procurement may track purchase price variance and supplier on-time delivery, operations may track inventory turns and service level, and sales may track discount rate and win rate. Below those are supporting metrics that teams use to diagnose problems. Each level should be traceable to the one above, so that a department can see how its work contributes. If a department KPI does not connect to any company KPI, either it is unnecessary or the company list is missing something.

A service desk example

A support team tracks tickets opened (1,200), tickets closed (1,150), average handle time (14 minutes), first-contact resolution (70 percent) and customer satisfaction (4.4 out of 5). All are metrics. The KPIs depend on the goal. If the goal is to retain customers by solving problems quickly, first-contact resolution and satisfaction are the KPIs. Handle time is a supporting metric. If handle time becomes a KPI on its own, agents may rush calls and close tickets before the problem is solved, which will push repeat contacts up and satisfaction down. Pairing a speed measure with a quality measure prevents this. The same logic applies in a warehouse (picks per hour with error rate) and in sales (calls made with conversion).

How to promote and demote

Metrics move in both directions. Promote a metric to a KPI when strategy changes: a company that decides to focus on retention may promote repeat purchase rate. Demote a KPI to a supporting metric when it has been stable at target for a long time and no longer needs executive attention, or when the goal has changed. Review the KPI list when the planning cycle starts, and record the reasons for changes. A KPI list that never changes is probably out of date.

FAQs

What is the difference between a KPI and a metric?

A metric is any measure. A KPI is a metric chosen because it tracks progress toward a key goal and is managed with a target, an owner and a review.

Can a metric be a KPI for one team and not for another?

Yes. Conversion rate may be a KPI for the marketing team and a supporting metric for the executive team.

How many KPIs should I have?

Few enough to focus attention, often between five and ten for a leadership team and three to five per department.

Are vanity metrics KPIs?

Not usually. A measure like page views is a KPI only if it connects to a goal and is managed with a target, which is rarely true.

Should I delete metrics that are not KPIs?

No. Keep them as supporting metrics for diagnosis, in drill-down views or appendices.