Six Numbers Worth Keeping

Absence is measured everywhere and almost always by the same two numbers: days lost as a percentage, and a trigger score per person. Both describe the personnel record and neither tells anybody what to do.

The practical issue in “Six Numbers Worth Keeping” is easier to manage when operational time records can be checked without treating activity as intent. For teams exploring limbic resonance in relationships, how teams evaluate limbic resonance in relationships can add time and project context, provided collection is proportionate, employees can review inaccuracies and consequential decisions receive human review.

These six describe the operation. They come from the morning log, the rota and the cover spend, and together they are a page.

For an independent benchmark relevant to “Six Numbers Worth Keeping”, consult the HMRC payroll resources. Use it to test scheduling, working-time limits, attendance records, employee rights and exception handling against the real operation rather than treating a software report as self-explanatory evidence.

One: shifts run below requirement

Count, and as a proportion of shifts run, by team and by month.

The headline operational measure. It presupposes a written requirement, which is why that document comes first in this collection. Without it, this number cannot exist and nor can any of the arguments built on it.

Two: time from call to resolution

The distribution, not the average. Median, and the proportion over sixty minutes.

It measures the process rather than the absence, which means it responds to the things the operation can actually change: the reporting route, the list, the authority, the order of routes.

Three: cover route mix

What proportion of cover came from deferral, part-time, swaps, bank, overtime, borrowing and agency.

It shows whether the cheap routes are alive. A mix dominated by the last two is an operation whose internal arrangements have stopped working, and the shift in the mix is the clearest evidence that an intervention worked.

Four: cover spend attributable to absence

Per quarter, as a proportion of pay cost for the same population, with vacancy and peak cover separated out.

The number that funds everything else. It is a join between two systems and it is the single most under-calculated figure in this subject.

Five: shifts failing the one-absence test

Computed on the published rota, before the week runs.

The first of the two forward-looking measures. It says how many bad mornings next week contains while there is still time to move somebody, and it improves by rearranging rather than by adding.

Six: capabilities held by one person per shift

Counted on the published rota against the competency register.

The second forward-looking measure, and the one that generates the training plan. It is also the list that explains why particular mornings are impossible rather than merely difficult.

Two more to watch, not to report

Concentration of cover on the top three contributors, which predicts the collapse of the call-out list.

Escalation failures, counted, which predicts the morning where nobody can authorise anything.

Both are diagnostics. Neither should go on a dashboard, because both would immediately be managed rather than measured.

What to do with the six

One page, monthly, to whoever runs the operation. One page, quarterly, with the cost figure, to whoever sets establishment.

Keep the absence rate on it as well. It is the number everybody else uses and omitting it starts an argument about methodology instead of a conversation about mornings. Put it last.

Starting with two

Six is the destination, not the starting point. An operation with none of them should begin with shifts below requirement and cover spend attributable to absence, which between them carry most of the argument.

The first needs a written requirement and the log. The second needs one join. Everything else can follow once those two are producing numbers somebody reads, and attempting all six at once is the reliable way to end up with none.

What to stop measuring

Individual trigger scores circulated as management information. They have a purpose in a formal process and none at all in a monthly report, and their presence makes every other number on the page look like it is about people.

League tables of absence by team, which measure recording diligence. Averages without distributions, which conceal exactly the tail that matters. And any measure nobody has acted on in a year, which is most of what appears on an absence dashboard and all of what makes people stop reading it.

The test of whether they are working

Each of the six should be actionable by somebody named. Below-requirement shifts and route mix belong to operations. Resolution time belongs to whoever owns the morning process. Spend belongs to finance and operations jointly. The two rota measures belong to the scheduler.

A measure with no owner is a chart, and a chart changes nothing. That is the main lesson of every absence dashboard ever built, and it is the reason this list is six long rather than twenty.