KPIs that earn a spot on the wall

The typical SMB dashboard has somewhere between 20 and 80 metrics on it. Almost none of them change anyone's behavior.

A KPI that doesn't earn its spot on the wall is noise. Here's the five-question filter we use with clients to decide which numbers deserve a dashboard, a weekly review, and someone's name next to them.

1. Does a change in this number change a decision?

If the number moved 20% in either direction, would anyone do anything different next week? If not, it's an observation, not a KPI. Fine to track, but don't put it on the wall.

The things that earn the wall are the ones where a 10% change would trigger an actual conversation: "Why is this down? Who's working on it? What are we doing differently?"

2. Can you name the person who owns it?

Every KPI on the wall should have one person — not a team, one person — whose job is to move it. "Operations owns this" is not ownership. "Maya owns this" is ownership.

KPIs without named owners get discussed at quarterly reviews and then forgotten. KPIs with named owners get moved.

3. Is it leading or lagging?

You want both, but in different places. Leading indicators (pipeline coverage, response time, intake completeness) tell you what's coming. Lagging indicators (revenue, churn, margin) tell you what happened.

Your weekly review should be mostly leading indicators — the stuff you can still influence. Your quarterly review should be mostly lagging — the stuff that tells you whether the leading indicators were the right ones.

If your weekly wall is all lagging indicators, you're driving by looking in the mirror.

4. Can it be measured without a human in the loop?

A KPI that requires someone to manually compile it each week is a KPI that will stop being measured in six weeks. Pick numbers that come out of systems automatically, or fix the plumbing until they do.

This is also the KPI test that surfaces your data gaps. If you can't auto-measure something you all agree matters, that's your next instrumentation project.

5. Is it hard to game?

Every KPI creates an incentive. Some KPIs create good incentives; some create bad ones. "Tickets closed per day" sounds good until people start closing tickets that shouldn't be closed.

Before you put a number on the wall, ask: what's the dumbest way someone could move this number without moving the underlying business? If there's an obvious dumb path, either pick a different metric or pair it with a counterbalancing one (quality score, CSAT, re-open rate).

The 5×5 rule

For an SMB, aim for something like 5 KPIs on the weekly wall and another 5 on the quarterly one. That's it.

Not 50. Not 500. Ten numbers that pass all five questions, with named owners, automatic collection, and clear leading/lagging split.

Everything else belongs in a report that's read when a specific question arises — not on the wall. A dashboard with 50 metrics doesn't tell anyone what matters. A dashboard with 5 does.

The quiet win

The real benefit of this exercise isn't the dashboard. It's the conversation that happens while picking the five. You'll find that your team doesn't agree on what matters. You'll discover that two people are both trying to own the same number. You'll surface metrics that you thought you were measuring and aren't.

That conversation is worth having once a year, even if the numbers never change.