If your dashboard has 47 metrics, you have zero. That isn't a paradox. It's what happens when a leadership team confuses measurement with attention.
A scorecard with 47 numbers is a place data goes to die. Nobody scans 47 rows on a Tuesday morning and walks away knowing whether the business is healthy. They glance, they feel vaguely behind, and they move on. The number that was about to break gets the same half-second as the number that's been green for two years.
Too few numbers and you're flying blind. Too many and nobody looks. The discipline is in the middle, and it's narrower than most teams want it to be.
Why the Range Is 5 to 15
Trinity Cadence puts the working range at 5 to 15 numbers on a weekly Scorecard. Not because there's something mystical about those bounds, but because of what each end protects against.
Below five, you can't triangulate. A single revenue number tells you what happened but not why, and not what's coming. You need a small constellation — a leading indicator, a lagging result, a quality measure, a capacity measure — before the picture has enough dimensions to predict anything.
Above fifteen, attention collapses. The human reading the Scorecard in a weekly Pulse has a finite budget of focus, and every number past fifteen spends that budget without adding signal. The team stops reading and starts skimming. Skimming is how a number goes red for six weeks before anyone says it out loud.
A metric you don't act on isn't a measurement. It's decoration. And decoration is expensive when it crowds out the numbers that matter.
The Test for Whether a Number Earns a Slot
Most teams pick metrics by asking "what can we measure?" The wrong question. The right question is: "what would change a decision this week if it moved?"
Run every candidate number through three filters before it gets a slot on the Scorecard:
- Predictive, not just descriptive. Does this number tell you something about next month, or only confirm last month? A Scorecard heavy on lagging metrics is a rear-view mirror.
- Owned by one human. If no single person can move the number, it doesn't belong on a leadership Scorecard. Shared accountability is no accountability.
- Actionable inside the week. If the number goes red, is there a play to run before the next Pulse? If not, it's context, not a Scorecard line.
A number that survives all three earns its place. A number that fails any one of them is clutter, no matter how interesting it is.
Leading Versus Lagging — Get the Mix Right
The most common scorecard failure isn't too many numbers. It's the wrong ratio. Teams stack their Scorecard with results — revenue, churn, closed deals — and wonder why the meeting feels like an autopsy.
Results are lagging. By the time they move, the cause is weeks old. A Scorecard built only on results can tell you that you missed, but never in time to do anything about it.
The fix is to anchor the Scorecard on leading indicators — the activities and early signals that predict the results. Demos booked predicts pipeline. Pipeline predicts revenue. Onboarding completion predicts retention. When you measure the leading number, you get a week or a month of warning instead of a quarter of regret. Aim for at least half your Scorecard to be genuinely leading.
The discipline here is harder than it sounds, because leading indicators are uncomfortable. A lagging number tells you what already happened — there's nothing to do but accept it. A leading number that's trending wrong is an accusation: it says someone could act, this week, and isn't. Teams unconsciously prefer lagging metrics because they feel like weather and not like responsibility. Resist that. The whole point of a Scorecard is to put the responsibility on the table while there's still time to act on it.
How AI Sharpens the Scorecard
Picking the right 5 to 15 is a human judgment call. Keeping them honest is where the Human + Machine Equation earns its keep.
In Trinity Cadence, AI does the unglamorous work that lets a small Scorecard stay sharp:
- Threshold drift. It flags when a number that's been quietly trending the wrong way is about to cross from green to red — before the Pulse, not during it.
- Correlation pruning. Over time it surfaces which of your numbers actually moved together, so you can cut the redundant ones and keep the count low without losing signal.
- Pulse prep. It assembles the week's Scorecard read and the one or two lines that need discussion, so the meeting opens on the exceptions, not the whole grid.
The machine never decides what matters. It just makes sure that the handful of numbers you chose stay legible, and that nothing red hides in plain sight.
Start by Cutting
If your current Scorecard is bloated, don't add. Cut. Print it, and for every line ask the three filters: predictive, owned, actionable this week. Strike everything that fails. Most teams are surprised how few survive — and even more surprised how much sharper the Pulse gets once the survivors have room to breathe.
The goal was never to measure everything. It was to know, in a single glance on a Tuesday, whether the business is on track. Five to fifteen numbers, chosen with discipline, will tell you that. Forty-seven never will.