How to Set KPIs Your Team Actually Tracks and Cares About
By Brendan Levin. 20+ years scaling businesses, from teams of 8 to 600 and budgets from startup to $100m.
Key takeaways
- KPIs fail when they measure what is easy to count rather than what the person can actually move.
- Every KPI needs exactly one owner and a number that defines good before the week starts.
- Leading indicators predict the outcome. Lagging indicators confirm it after the fact.
- Embed KPIs in a standing weekly meeting so they are reviewed without the founder pushing.
Why most KPIs get ignored
The most common KPI design mistake is measuring the output nobody controls. Revenue is a lagging number. Deals closed is a lagging number. The person responsible for sales cannot reach into the pipeline and pull a deal closed this week. What they can do is book calls, send proposals, and follow up on outstanding quotes. Those are the numbers worth tracking.
The second mistake is too many of them. A team member who tracks twelve KPIs effectively tracks zero. When everything is important, nothing is actionable. Three to five numbers per functional area is the ceiling. If you have to choose between two metrics that seem equally important, ask which one the person can move with their own decisions this week. That is your metric.
The third mistake is no owner. A KPI owned by the team is owned by nobody. One name, one number. If two people share accountability for a result, the result is unowned.
Leading vs. lagging: the only distinction that matters
A lagging indicator confirms what already happened. A leading indicator predicts what is going to happen and is influenced by action taken now. Most founders track lagging indicators almost exclusively, then wonder why the weekly meeting feels like a post-mortem rather than a planning session.
Here is a concrete example. A professional services business cares about revenue. That is the lagging metric. The leading indicators might be: proposals sent this week, average days to follow-up on an open proposal, and number of scoping calls booked. The person running business development can move all three with their own effort this week. If those three numbers are healthy, revenue follows. If revenue drops but you have no leading indicators, you find out six weeks too late to do anything about it.
| Lagging indicator | Leading indicator that predicts it | Who owns it |
|---|---|---|
| Revenue this month | Proposals sent this week | Head of Sales / BD |
| Client retention rate | Check-in calls completed per account | Account Manager |
| Project delivery on time | Tasks flagged at-risk before due date | Project Lead |
| New leads this quarter | Outbound contacts made this week | Marketing or Sales |
| Team utilisation rate | Capacity flagged more than 5 days ahead | Operations Lead |
How to design a KPI your team member actually owns
Start with the outcome you care about, then ask: what action, taken this week by this person, is the strongest predictor of that outcome? That action becomes the KPI. Then define what good looks like before the week starts. Not 'more proposals' but 'five proposals sent.' A number without a target is a data point, not a KPI.
Next, check the control test. Can the person hit or miss this number based on their own decisions, without needing approval from you? If the answer is no, the KPI is not theirs to own. This is where decision rights and KPI design intersect. If your team member has to ask for sign-off before they can act, the metric is yours, not theirs.
Finally, keep the number visible between meetings. A shared document, a simple spreadsheet updated on Mondays, or a board that lives in the team's main channel. The format does not matter. Visibility does. When a number is updated in public, people update it. When it lives in a folder nobody opens, it dies.
Embedding KPIs in the weekly rhythm so they are checked without being pushed
A KPI that only gets reviewed when the founder asks for it is not a team KPI. It is a founder task. The fix is structural, not motivational. Put the KPIs on the standing weekly meeting agenda. Not as an appendix. As the first agenda item, before status updates and before problems. Numbers first, then conversation.
The meeting structure that works: each owner reads their three to five numbers, states whether they are on track or off track, and names one thing they are doing differently this week if a number is off. That is it. No lengthy reports, no slides. This takes about fifteen minutes for a team of four to eight people. A well-structured weekly meeting is what turns KPIs from a document into a discipline.
When a number is consistently off track and the same explanation keeps appearing, that is useful information. It usually means either the target is wrong, the person does not have the tools or authority to hit it, or something upstream is broken. Any of those is worth your attention. None of them are visible if the KPI only gets checked when you remember to ask.
The founder's role once KPIs are set
Your job after KPI design is to review, not to report. If you are chasing people for their numbers, you have a rhythm problem, not a KPI problem. The operating system around the KPIs is what determines whether they get used.
When a number comes in below target, resist the urge to fix it. Ask the owner what they think is driving it and what they plan to do. That conversation builds the judgment your team needs to own results. When you solve it for them, you pull the ownership back to yourself, which is exactly the pattern most founder-led businesses are trying to break. Accountability built on questions holds longer than accountability built on pressure.
Good KPI design is also where direction and measurement connect. If your team's KPIs do not trace back to the one outcome the business is organising around this year, you have a set of individual scorecards with no shared purpose. The leading indicators your team tracks should make the annual goal feel close enough to act on this week.
KPIs mean more when each one sits with an owner and a regular check-in. The Momentum Engine builds that ownership and rhythm with you over 12 weeks.
See the Momentum EngineCommon questions
How many KPIs should each team member have?
Three to five is the working range. Fewer than three and you probably have not found the real leading indicator. More than five and the person cannot hold them all in focus. When in doubt, cut the KPI that measures what is easiest to count and keep the one that most directly predicts the outcome you care about.
What is the difference between a KPI and a task?
A task is something you do. A KPI is a number that tells you whether the doing is working. 'Send five proposals this week' is on the border: it is an action with a countable output, which makes it a useful leading indicator. 'Update the CRM' is a task with no predictive power. When in doubt, ask what outcome the number predicts. If there is no clear answer, it is probably a task, not a KPI.
My team tracks the numbers but nothing changes. What is wrong?
Tracking without consequence is a ritual. The numbers need to connect to a conversation about what is different this week. In the weekly meeting, when a number is off, the owner should name one specific action they are taking to move it. If the number is off week after week with no change in action, you have an accountability gap, not a KPI gap. The structure is working. The conversation around it needs to sharpen.
Should KPIs be tied to bonuses or pay?
That is a separate design question from whether KPIs get tracked. Many businesses that tie KPIs to pay find the numbers get gamed rather than genuinely moved. The more reliable path is KPIs that are visible, owned, and discussed weekly, with compensation structured separately around results over a longer period. Get the tracking habit working before you attach money to it.
How do I set KPIs for a function I do not fully understand?
Start by asking the person in the role: what action, done well this week, most predicts a good result for this function? They usually know. Then ask what would tell you, from the outside, that the function is working. The gap between those two answers is often where the real leading indicator lives. You do not need to be an expert in the function to ask those two questions.
How often should KPIs be updated or changed?
Review the KPI set at the start of each quarter. Some leading indicators stop being relevant as the business shifts. Others reveal that you were measuring a proxy rather than the real driver. A KPI that has been hit consistently for eight weeks is worth interrogating: is it still the right number, or has it become too easy? The goal is a number that creates useful tension, not one that gets ticked off by default.