The Numbers Worth Looking At
Four measures that tell a small business something useful, and the ones that mostly rank people.
Legal
Attendance data can generate a great many numbers. Very few of them change a decision in a business of twenty people.
The four
Total absence days, by month. The overall picture, and whether it is moving.
Absence by reason, so sickness, holiday and other are visible separately.
Coverage failures: how many times a shift ran short or someone had to be called in.
Leave balances at the autumn checkpoint, which predicts the December problem.
Four numbers, once a month, on one page.
What they tell you
Rising sickness in one team is a working-conditions question before it is a people question. Look at the shift pattern, the workload and the physical environment first.
Coverage failures rising means the minimum staffing is too tight or the leave limit is too loose.
A cluster of absences around one period is usually a workload or scheduling signal.
High untaken leave means people feel they cannot take it, which is a coverage problem wearing a different hat.
What not to produce
A ranking of people by absence days.
Individual absence scores, which have their own note and their own legal risk.
Comparisons between people on data that reflects health and caring responsibilities as much as anything else.
A chart on a wall. Public attendance displays humiliate and change nothing.
Sickness benchmarks
Published national averages exist and are of limited use: they vary by sector, by workforce age, by whether people can work while unwell.
Compare against your own previous year, which is the comparison that can act on.
And treat a low figure carefully. A business with almost no recorded sickness may have healthy staff, or may have people coming in ill because they cannot afford not to.
The pattern worth watching
Absence concentrated in one team, one shift or one task.
That is a working-conditions finding, and it is the most valuable thing this data produces.
Investigate the conditions before the individuals, consistently, and say publicly that this is the policy — because otherwise people conceal the reasons and the data stops describing anything.
Who sees it
Aggregate numbers: whoever runs the business.
Individual records: the person themselves, and their manager for coverage purposes.
Sickness detail: one named person.
Not the team, however small and however much everyone already knows.
Conditions before individuals
A policy worth stating publicly.
Where absence concentrates — a team, a shift, a task — look at the conditions first.
Say that this is what you do.
Because the alternative teaches people to conceal reasons, and then the data stops describing anything real.
It is also where the genuine improvements come from: a shift pattern, a physical task, a workload, all fixable, and none visible if people stop being candid.
Connect policy to configuration
The practical choices behind this note can be compared with employee productivity reports. Keep the written purpose in control of the setup, enable only data needed for that purpose and review the result with affected employees.
Independent reference
For an external point of reference, see the Office for National Statistics. Its published labour-market statistics are a useful external benchmark, provided the definitions are compared before drawing conclusions.