A New Metric Is Worth Adding to the Dashboard or Not: Decision Tree

Why this matters

Every metric someone suggests sounds reasonable in the moment. "We should track how many callbacks come from each service type." "Let's watch average response time on quotes." Say yes to enough of these, one at a time, and a dashboard that once fit in a glance turns into forty tiles nobody reads anymore, and the two or three numbers that actually run the business get buried in the noise. Saying no to a reasonable-sounding metric feels harder than it should, because there is rarely an obvious reason to refuse. This tree gives you one.

Start here: what problem would this number actually solve

Before evaluating anything else, ask the person proposing it, or yourself, one question: what decision would we make differently if we had this number that we cannot make today without it? A surprising number of proposed metrics fail this question immediately, because the honest answer is "it would be interesting to know" rather than "it would change what we do."

If there is no clear decision attached, this metric does not belong on a permanent dashboard. It may still be worth a one-time look or a temporary tracking period if a specific hunch prompted it, see related: Turning a Gut Feeling Into a Number Worth Tracking, but a permanent tile earns its place by driving a recurring decision, not by being interesting.

If there is a clear decision attached, move to the next step.

Step 1: Check whether an existing metric already covers this

A new metric proposal often turns out to be a slightly different angle on something you are already tracking. Before adding a tile, check whether the decision the new number would drive is already served, even imperfectly, by a number already on the screen.

• If an existing metric already drives the same decision, even if the new one is slightly more precise, the marginal accuracy rarely justifies a new permanent tile. Note the refinement as a possible future improvement to the existing metric instead of a new one alongside it.
• If nothing currently on the dashboard drives this particular decision, this is a genuine gap, and the new metric has cleared a real bar. Move to the next step.

Step 2: Check whether the underlying data is already reliable enough to trust

A metric can pass both tests above and still be premature if the data behind it is not yet clean. Adding a tile for a number you privately do not trust does more harm than not tracking it at all, because it trains people to treat the whole dashboard as suspect. See related: A Metric Was Wrong Because of Bad Data Entry.

• If the data behind the proposed metric is already captured consistently and accurately by the team today, it is ready to track now.
• If the data requires a new habit, a new field, or a process change before it can be captured reliably, put the process change in place first, run it for a trial period, and confirm the numbers look sane before promoting it to a permanent tile. A metric added before the underlying data is trustworthy is worse than no metric.

Step 3: Decide whether it earns a permanent slot or a temporary watch

Not every worthwhile metric belongs on the daily glance forever. Sort it by how it will actually get used:

• Permanent dashboard tile: something you would want to see at every regular glance, indefinitely, because it reflects an ongoing health check on the business (see related: The Dashboard Numbers That Matter for how a permanent tile should be sized and placed).
• Temporary watch item: something tied to a specific initiative or concern with a natural end date. If you just changed how estimates get followed up, watching the follow-up rate closely for the next quarter is smart. Once the initiative has proven out or failed, that metric should come off the regular glance, whether or not it stays available in a report you can still pull on demand.

Naming which of these two a new metric is, at the moment you add it, prevents dashboard bloat. A temporary watch item added without an end date in mind tends to become permanent by default, simply because nobody remembers to remove it.

Step 4: If you add it, remove something else

The healthiest discipline for keeping a dashboard useful is a one-in-one-out rule. If a new metric genuinely earns a permanent slot, look at what is currently on the screen and ask whether anything there has stopped driving a decision, gone stale, or become a "nobody looks at this anymore" tile. Retiring a tile at the same time you add one keeps the total count stable and forces an honest audit of what is actually still useful, rather than letting the dashboard only ever grow.

The recap

Start with the decision test: what would we do differently with this number that we cannot do today. Check whether an existing metric already covers the same ground. Confirm the underlying data is clean enough to trust before promoting it. Decide up front whether it is a permanent tile or a temporary watch with a natural end date, and if it earns a permanent spot, retire something else to make room. A dashboard that only grows stops being a dashboard and becomes a wall of numbers nobody actually reads.

References

• U.S. Small Business Administration (SBA), small business performance measurement practices
• Trade-standard practice for KPI dashboard design in field-service operations
• See related: The Dashboard Numbers That Matter; Turning a Gut Feeling Into a Number Worth Tracking