← All posts
Framework Guide   May 29, 2026 · 8 min read

The difference between a KPI and a metric — and why most companies get it wrong

Generated illustration for the post >-

A metric is any number you can track. Website visits, employee NPS, average deal size, support ticket volume, time-to-first-response, percentage of meetings starting on time. All metrics. Most companies have hundreds, sometimes thousands, scattered across BI tools, spreadsheets, and the slide somebody included in a quarterly update three years ago and nobody has had the heart to delete.

A KPI — a key performance indicator — is something narrower. It is a metric you have deliberately chosen because it serves as a leading or lagging indicator of strategic progress. The word "key" is doing almost all of the work in that acronym, and almost nobody respects it. Not all metrics are keys. Most metrics are not keys. But most companies treat the words as synonyms, slap "KPI" on every chart in the dashboard, and then wonder why the leadership conversation about performance never converges.

The three tests for KPI status

A metric earns KPI status if it passes three honest tests. Skip any of them and what you have is a metric pretending to be something more important than it is.

Test 1: Is it directly connected to a strategic objective? A KPI without a strategic objective attached to it is not a performance indicator — it is a tracking habit, often inherited from whoever set up the dashboard two years ago. Revenue growth is a metric. Revenue growth from new enterprise customers in the DACH region, connected to the explicit objective of expanding the enterprise segment in German-speaking markets, is a KPI. The difference is not pedantic. The first version cannot tell you whether your strategy is working. The second can.

Test 2: Does it change when strategy execution improves? Some metrics move regardless of whether you are executing your strategy well — they respond to market conditions, seasonality, or general activity. Some metrics move specifically because you are executing well. KPIs need to be in the second category. They have to be sensitive to strategic performance, not just to general business motion. If a metric goes up whether you focus on it or not, it is telling you about the weather. That is useful, but it is not a KPI.

Test 3: Does someone own it, by name? A metric without an owner is just a number in a dashboard, refreshed nightly, watched by nobody in particular. A KPI needs a single accountable person who is responsible for understanding it, influencing it, and explaining it when it moves in the wrong direction. If the owner is "the team" or "marketing and ops, jointly," the metric does not have an owner — it has the appearance of one. The first time it moves badly, the meeting will reveal which of the joint owners is actually accountable, which is usually: neither.

Why the confusion is expensive

When everything is a KPI, leadership attention gets diluted across dozens of numbers, sometimes hundreds. The genuinely important signals — the five or ten metrics that are actually connected to strategic bets — get buried in the noise of operational tracking. The dashboard becomes wallpaper. The MBR becomes a tour of charts rather than a decision-making session, because no single chart carries enough weight to force a decision.

The real cost of KPI inflation is not the time spent looking at too many numbers. It is the decisions that do not get made because the signal that would have triggered the decision is buried under twenty-seven other signals of equivalent declared importance. When everything is critical, nothing is. The leadership team loses the muscle of treating any single number as truly diagnostic, because every number gets the same level of visual emphasis in the same dashboard.

There is a second-order cost: the credibility of measurement itself starts to decay. People stop trusting the dashboard because too many of the numbers on it are stale, contested, or obviously not predictive of anything. Once that trust is gone, even the genuinely important metrics get treated with the same skeptical shrug. The signal-to-noise ratio of the entire measurement function collapses.

What good KPI discipline looks like

A company with good KPI discipline has five to ten KPIs at the company level. Not fifty. Each one is directly connected to a specific strategic objective in language anyone in the company could explain. Each one has a named owner who shows up in the MBR ready to explain movement, not just report it. Each one is reviewed in the context of the goal it is meant to move, not as a free-standing number on a slide.

Everything else is a metric — useful to track at the operational level, not necessarily useful to review at the leadership level. The discipline is not just choosing the right five or ten. It is having the political courage to demote everything else. Demoting a metric somebody has been presenting for two years is genuinely uncomfortable. That discomfort is the work. A company that cannot demote metrics is a company that cannot focus.

How to do the demotion

Run a one-hour exercise with the leadership team. List every metric that currently appears on a leadership dashboard or in a leadership review. For each one, apply the three tests. If it does not pass all three, it gets demoted to the operational layer. No exceptions for sentiment, history, or "but the board asks about it."

The demoted metrics do not disappear. They keep existing in the operational systems where they belong, watched by the people who actually act on them. They just stop demanding leadership attention every month. The board, when it asks, gets the answer with context. The dashboard at the leadership level becomes legible again.

The Vindaris view

In Vindaris, KPIs sit adjacent to the goals they serve and the work meant to move them. The connection is not a label — it is structural. When a KPI moves, you can trace it back, in the same view, to the work that caused the movement and the owner whose decisions drove it. When a KPI is amber, leadership knows in seconds whether the cause is execution drift, capacity shortfall, or external conditions — without a meeting, without a chief-of-staff investigation.

A KPI that lives in a dashboard far from the goal it serves and the work that moves it will always be retrospective. Bring them together, and the K in KPI starts meaning something again.