KPIs that nobody uses: why measurement frameworks fail and what to do instead
Most organizations don't have too few KPIs. They have too many. The numbers keep growing, the decisions don't. What's missing is a clear link between each indicator and who makes which decision based on it.
Signs your KPIs aren't working
- The monthly report runs to dozens of pages. Meetings read out numbers but don't decide anything based on them.
- Two departments have two different numbers for the same thing, and the meeting is spent arguing about which one is right.
- For half of the indicators, nobody knows who is responsible.
- Targets haven't changed in years, or change without explanation.
- Leadership sees detail that doesn't help them, while teams see aggregates they can't act on.
Why measurement frameworks fail
They measure what is easy to measure. Indicators come from whatever data the system happens to produce, not from the questions leadership needs answered.
There is no link to a decision. An indicator nobody acts on is just information. Information without a decision takes up meeting time and changes nothing.
There is no owner. When everyone is responsible for a number, nobody is. Deviations get explained rather than fixed.
They sit at the wrong level. A line supervisor needs different numbers from a managing director. One set of KPIs for the whole company serves nobody well.
The set is never pruned. Indicators get added but never removed. After a few years there are so many that the important ones get lost.
Five questions for every indicator
Before an indicator goes into a report, answer five questions:
- What decision will someone make based on it, and who?
- Who owns it, and can they actually influence it?
- How often, where and in what form will the people acting on it see it?
- What value is good, and why that value?
- What happens when it deviates?
If you can't answer the first question, the indicator doesn't belong in the report. It can stay in the data for analysis, but in a meeting it only takes up space.
From goals to teams
A well-designed set of indicators starts from the company's goals. For each goal, you identify a few key result areas, usually three to seven, that the outcome truly depends on. Only then do you choose indicators, separately for leadership, for departments and for operational teams.
A useful test: can every team tell from its own numbers whether it had a good week? If yes, the indicators are right. If not, the team either sees numbers it can't influence, or sees none at all.
Rhythm matters as much as selection. Operational indicators belong in a short weekly meeting, summary indicators in the monthly leadership meeting. Once a quarter, review the whole set and drop what nobody uses.
A real example
A fast-growing technology company was running more than fifteen parallel initiatives: new markets, product development and team onboarding. Nobody had a view of the full portfolio, ownership was scattered, and priorities went to whoever pushed hardest. We set up a lightweight project management office with a portfolio overview, an ownership matrix and a weekly review rhythm. Within six weeks, leadership had full visibility, and three critical initiatives had a clear owner and deadline.
Where to start
Take your current report and ask the first question above for every indicator. Most companies drop between a third and a half of their KPIs after this exercise, and start using the rest. If you want support, a Quick Scan maps what is measured today, what is missing, and what should drive decisions.