The Decision Test: Retiring the Metrics That Earn No Place in the Pack
The quiet cost of a full dashboard
Most UK management packs are not short of numbers. They are short of consequences. A metric can be accurate, timely and beautifully presented, and still fail the only test that matters: does it change what someone does?
When a dashboard accumulates without discipline, the cost is rarely dramatic. It appears as preparation time that no one questions, as slides that are skipped in operating reviews, and as a familiar silence when a figure moves and no one responds. The reporting pack becomes a record of activity rather than a instrument of management.
The answer is not to strip out measurement. It is to apply a decision test to every KPI and to treat orphaned metrics as a operational housekeeping issue. Fewer numbers, better arguments and a pack that earns its place are the goal.
Start with the decision, not the number
A useful audit begins by working backwards. For each metric in the current pack, ask a simple question: what decision does this inform?
The answer should be specific. Not 'performance management' or 'oversight', but a named decision that someone actually makes. A margin measure might inform a pricing decision. A backlog measure might inform a resourcing decision. A client concentration figure might inform a bid or account-planning decision. If the answer is vague, the metric is probably an orphan.
This is not an exercise in suspicion. Many metrics were introduced for good reasons and have simply outlived them. The audit gives the organisation permission to notice that.
A three-line entry for every metric
A short register is enough. For each KPI, record three things: the decision it supports, the person who owns that decision, and the date on which the metric will next be reviewed.
The decision owner matters more than the metric owner. The person who prepares a figure is not necessarily the person who acts on it. In many UK SMEs, finance prepares the number, operations reads it, and no one is formally accountable for the decision it should trigger. That is how a metric becomes orphaned even while it remains visible.
The review date is equally important. Metrics should have a shelf life. A measure introduced during a particular programme or market condition should not remain indefinitely because no one has asked whether it still applies. An expiry date creates a natural moment to retire or redesign.
Distinguish orphans from insurance
Some metrics do not prompt a regular decision but are still legitimate. These are insurance metrics: safety indicators, compliance measures, cash buffers or other guardrails that exist to trigger an exception. They earn their place not by informing a weekly choice but by defining a threshold at which someone must act.
That distinction matters. An insurance metric still needs a named owner and a clear threshold. If it has neither, it is not insurance; it is decoration. The audit should separate active decision metrics, insurance metrics and orphans. Only the third group should be a candidate for retirement.
The owner problem
A metric without a decision owner is a number without a home. This is especially common in organisations where reporting has grown around functions rather than decisions. The finance team may own the production of a figure, but the commercial team owns the response. If that split is not made explicit, the metric floats between meetings.
Naming the decision owner is not about assigning blame. It is about making the route from evidence to action visible. Once the owner is named, the next questions become easier: is the metric timely enough for their decision, at the right level of detail, and in a form they can use?
Review, redesign, retire
When a metric fails the decision test, there are three options. The first is retirement. If no decision depends on it, archive it with a short note explaining why. The second is redesign. If a decision exists but the metric does not inform it well, change the frequency, granularity or presentation. A monthly average may need to become a weekly leading indicator. A group-level figure may need to be broken down by site or client segment.
The third option is retention with a clearer contract. Some metrics are genuinely useful but have been allowed to drift. Give them a decision owner, a threshold and a review date, and they become part of the management system rather than background noise.
A leaner pack for the 2027 planning round
As UK organisations enter the 2027 planning round, there is a practical opportunity to reset the reporting pack before new measures are added. The annual planning cycle tends to add metrics; it rarely removes them. A short audit before the new year can prevent another layer of inherited reporting.
The output should be a one-page decision log: the metric, the decision it supports, the owner and the review date. Anything that cannot be entered on that page is a candidate for the archive. The pack that remains will be shorter, but it will be easier to defend and easier to use.
Running the audit in ninety minutes
A first pass need not be elaborate. Gather the last three management packs. List every metric that appears. For each one, ask the decision question, name the owner and note the last occasion on which the figure changed a decision. Mark each as active, insurance or orphan. Then agree the next action: retain, redesign or retire.
The discipline is in the follow-through. Retirement should be visible, not silent. If a metric is removed, say so and explain why. That gives the pack a clearer logic and signals that reporting is subject to the same scrutiny as any other operational process.
The aim is not a minimalist dashboard for its own sake. It is better decision quality. A metric that cannot find a decision is an orphan. Give it a home, or let it go.
