Finance closed the accounts in six days this month. The target is five, so the box is red, and in the management meeting you ask what went wrong. The finance manager has an answer ready: someone was on leave, and two branches sent their figures late. Next month it's five days, the box is green, and nobody mentions it again.
Many founders run their monthly reviews this way. Each red number gets a question, each question gets an explanation, and the meeting moves on. The trouble is that the explanation may have nothing to do with the result. Sometimes the number moved because something happened. Often it moved about as much as it usually does.
Same average, different problems
Take two finance teams, measured on how many days they take to close the monthly accounts. (This is an illustrative example, not a client.)
Team A, over six months: 5, 5, 6, 5, 4, 5. Team B, over the same six months: 2, 9, 3, 8, 2, 6.
Both average five days. If your scorecard shows only the average, or only this month's figure, the two teams look the same.
They aren't. You can plan around Team A. Next month will probably be five days, give or take one. Team B could close in two days or nine, and even the finance manager can't tell you which. Something in how that team works is unsettled: the data arrives at different times, one person holds a step nobody else can do, or approvals depend on who's around.
Now say your target is four days. Team A has a different problem from Team B. It's predictable, and it predictably misses. Asking Team A to explain each month's miss gets you twelve explanations a year and no change, because the way month-end is set up produces five days.
A red box doesn't tell you why
Most scorecards use colours. Green if the target was met, amber if it was narrowly missed, red if it was missed by more. The colour tells you whether the result was acceptable. What happens next is the problem: the colour starts being read as if it also explained the result.
The idea that most variation is ordinary is old. In the 1920s Walter Shewhart, working at Western Electric in the United States, separated the variation a process produces as a matter of course from variation that has a specific cause you can find. W. Edwards Deming later carried that distinction into management teaching. The cost of confusing the two runs both ways. Treat ordinary movement as a crisis and you waste time chasing causes that aren't there, and you may blame people for the system. Treat a real change as ordinary and you miss it until it's expensive.
There's a second cost. When the founder asks "why?" about every red number, managers learn that "I don't know yet" isn't an acceptable answer. So they bring a story. Someone was on leave, a supplier was late, the market was quiet. Those things may all have happened. Whether they caused the number to move is a different question, and without history there's nothing to test the story against. If response times have bounced between three and five hours all year, a month at 4.5 hours during someone's leave tells you very little about the leave.
Four situations, four responses
Look at the pattern and the target together. Put simply, a department measure can be in one of four situations, and each calls for something different.
Stable and meeting the target. Leave it alone. Asking for explanations of small monthly movements here takes management time and gives nothing back.
Stable and missing the target. This is a process problem. Imagine a finance team that takes seven or eight days every month, with no drama, while the business needs management accounts in four. Telling them to try harder won't move it. Something in how month-end works has to change. Branch figures may come in late, M-Pesa and bank reconciliations may be done by hand, or one approval may sit with someone who's often travelling. Fixing the target means fixing those.
Unstable but meeting the target on average. Be careful here. Say proposals are meant to go out within three days and the average is 2.7. The last eight took 1, 5, 1, 6, 2, 1, 4 and 2 days. The average looks fine, but a customer can't tell whether their proposal will come tomorrow or next week. The rest of the business can't depend on a department like this, even when the monthly number is green.
Unstable and missing. Investigate. Find out what's different about the good months and the bad ones. This is where looking into individual months is worth the time.
In many businesses all four get the same response: the manager is asked to explain the result. That's the habit to change.
What to ask in the monthly meeting
Change the first question. When a number is red, ask whether it's outside the range this measure normally moves in. You can only answer that if the scorecard shows history, so put the last twelve months beside each figure, or a simple line chart in the same sheet you already use.
A rough working rule is enough to start. If this month sits inside the range of the past year and there's no run of several months heading the same way, treat it as ordinary. Then the useful conversation is about the system: "We've been between 88% and 94% all year. The target is 97%. What would have to change for 97% to be possible?" If a result falls well outside the usual range, or several months in a row move in one direction, that's worth looking into straight away.
None of this means waiting while the business loses money. A large customer leaving, or sales falling sharply in one month, needs action that week. The point is to save your attention for changes like that, and to stop spending it on movements that sit inside the normal range.
Keep the target and the pattern side by side. The target tells you whether performance is good enough for the business. The pattern tells you what the department is currently set up to produce. The gap between the two is where management work belongs.
Why this is hard to keep going
A founder can follow this argument in one reading. Keeping it going is harder. It needs scorecards that carry history, kept up to date every month. It needs managers who come to the meeting having already looked at their own pattern, and who take ownership of fixing a process when the numbers show it can't meet the target. And it needs the founder to stop asking "why?" about every red box, which is a hard habit to break when you're the one who has always held the business together.
In founder-led businesses we often see a scorecard that's been built once and then left. The numbers get filled in, the colours change, and the meeting still runs on explanations. Talentos works with founders to build a way of managing performance where expectations are clear, managers review their own results, and the monthly meeting is used to decide what to change. If you want to see how performance is actually being managed across your business, the Performance Picture is a ten-working-day assessment. We hear from leadership and staff, look at your records, and show you where things break down and what to fix first. The free three-minute Quick Picture is a quicker place to start.
For more on what a founder needs to see each month, read our piece on founder performance dashboards. If the problem is that departments are all green while the business struggles, see why teams can hit their KPIs and the business still struggles.