Too Many Metrics and Nobody Looks at Any of Them: Decision Tree

Why this matters

Somewhere along the way, the report grew from five numbers to thirty. Every one of them seemed worth tracking at the time. Now the review meeting takes an hour, half the team has stopped opening the report at all, and the two numbers that actually matter are buried between eight that nobody has acted on in a year. Metric sprawl does not fail loudly, it fails by attrition: people quietly stop looking, and you lose the whole practice, not just the bloated part of it. This is how you find the actual cause and prune it back to something people will use.

Start here: confirm this is really what's happening

Before you cut anything, verify the symptom. Ask two or three people who are supposed to be using the report the same question: "what number in here would make you change something today?" If they hesitate, cannot name one, or name a different number than you expected, you have confirmed metric overload rather than assumed it. This also tells you which numbers are still alive to the team, which you will need in the next step.

Step 1: Find out how you got here

Metric sprawl almost always has one of a few specific causes, and the cause determines the fix.

• If every new question ("what about X?") got answered by adding a tile instead of answering the question and moving on, you have been building a report by accretion. The fix is a pruning pass, not a redesign.
• If the report tries to serve several audiences at once (the owner, the crew, a lender, a partner) with one shared view, you have an audience-mixing problem. The fix is splitting into role-specific views, not just trimming the list.
• If someone keeps a metric alive because removing it feels like admitting the thing it measures does not matter, you have an attachment problem, not a data problem. The fix is a direct conversation about whether the underlying question still matters, separate from the metric itself.

Name which of these is actually happening before you touch the list. Cutting numbers without understanding why they piled up just means the pile grows back the same way.

Step 2: Apply the "would we act on this" test to every metric

Go through the current list one at a time and ask a single question of each: if this number moved sharply next week, what would we actually do differently? Sort into three piles.

• A clear action comes to mind immediately. Keep it. This is an operating metric.
• You would "look into it" or "keep an eye on it," nothing more specific. This is a soft keep. Flag it for step 3.
• Nothing changes either way. Cut it. Nobody is acting on it, and nobody has been for a while, or you would have named the action instantly.

Be honest about the middle pile. "Keep an eye on it" is often a polite way of saying a metric survives on inertia, not usefulness.

Step 3: For the soft keeps, decide watch vs cut

The middle pile needs one more question: is this a genuine early-warning indicator for something you do act on, or is it just interesting?

• If it is an early-warning signal for a metric you do act on (an open-estimate count that precedes a close-rate dip, for example), keep it but demote it to a secondary check, not a headline number in the main review.
• If it is interesting but not predictive of anything you already act on, cut it from the recurring report. It can live in an occasional deep-dive if someone wants to study it, but it does not belong in the weekly or monthly rhythm.

Step 4: Rebuild around a hard cap

Once you have pruned, cap the primary review list at a number small enough that the whole group can hold it in their heads, typically five to ten depending on how the business is structured. A hard cap forces continued discipline: the next time someone wants to add a metric, something else has to leave to make room. That trade-off is the mechanism that prevents the sprawl from simply regrowing.

Step 5: Split by audience if that was the real cause

If step 1 identified audience-mixing as the cause, pruning the master list is not enough on its own, because different roles genuinely need different numbers.

• Owner view: the handful of numbers that answer "is the business healthy and what do I do about it."
• Crew-facing view: the numbers that reflect their own work directly and that they can influence day to day.
• Any outside party (lender, partner): a small, separate summary built for that specific relationship, refreshed only as often as that relationship actually needs it.

One shared thirty-tile report trying to serve all three groups is worse than three short ones each serving one group well.

Step 6: Re-check in a month

Metric sprawl rebuilds itself quietly if nobody watches for it. A month after the prune, ask the same question from the "start here" section again: what number would make you act. If people can answer instantly and the list has not crept back past your cap, the prune held. If it has crept back, you skipped step 1, someone is re-adding metrics without cutting anything to make room.

References

• General practice on operational dashboard design and decision-relevant reporting
• SBA guidance on performance measurement for small businesses
• See related: The Difference Between a Vanity Metric and an Operating Metric; The Review Meeting That Turns Numbers Into Decisions