If you run a large, distributed operation and you have just come back from summer leave, there is one number worth twenty minutes of your time this week. Not the one in the board pack. The one underneath it.
Summer is an operational stress test
For most large employers, the summer months are the closest thing to a controlled stress test their operation gets.
Cover runs thin. Annual leave, sickness and emergency childcare all overlap in the same weeks. Agency staff and already-stretched colleagues absorb the difference. Rotas that work comfortably in February stop working in August.
Some sites come through that intact. Others do not. And the difference between them has just been recorded in your absence data.
That is a genuinely useful thing, because it is very hard to see under normal conditions. A site that is quietly fragile looks much like a site that is robust when both are fully staffed. Put them both under strain and they separate.
The conversation that prompted this
I was talking to a prospective client about absence recently. I asked what their current run rate was, and they gave me the number — the one that goes to the board.
Then I asked a different question. Where is it worst?
That is when the conversation became useful. Because they knew. Not precisely, and not in a form they could put in front of anyone, but they knew. Certain sites. Certain shifts. Certain teams where the rate ran at a multiple of the headline figure.
Almost every operational leader I speak to has that same instinct, and almost none of them have it in a form they can act on.
Two numbers doing two different jobs
The headline absence rate tells you what is at stake.
Converted to money it is never a small number. At four or five per cent, a large employer is looking at millions a year, and for the very largest, tens of millions — before agency cover, overtime and lost output are counted. In the cost models we build with prospective clients, the direct figure routinely doubles once those are included. That number is real and it deserves attention.
What the headline rate cannot tell you is two things: how bad the problem actually is in the worst-affected parts of the business, and where you would start if you decided to do something about it.
Averages are designed to obscure exactly that. A single organisation-wide percentage is a weighted blend of everything from your best-run site to your worst, and the arithmetic of that blend systematically flatters the worst. A minority of locations running at two or three times the average will move the overall figure by very little while accounting for a wildly disproportionate share of the cost, the disruption and the risk.
Hot spots are not random
The most important thing about hot spots is that they are not distributed evenly, and they are not distributed by chance.
They cluster. The same depot. The same night shift. The same team under the same manager. The same weeks each year.
And they persist. Not for a quarter, but for years — because nobody is looking at that level of granularity, and the aggregate smooths them out of existence long before the number reaches anyone with the authority to act.
A hot spot is rarely just an absence problem
This is the part that gets missed, and it is the reason this matters to Operations rather than only to HR.
Go and look at the same locations that show elevated absence, and you will usually find the rest of it there too. Productivity below the network average. Customer satisfaction scores worse. Process compliance patchy at best. Turnover higher. Agency and overtime spend concentrated in the same teams.
These are not four separate problems that happen to share a postcode. They are one problem, surfacing in four different reports, owned by four different people, none of whom is routinely looking at the other three.
Absence simply happens to be the one that gets counted every week. That makes it the earliest and cheapest place to see a problem that is not fundamentally about absence at all.
It also explains something that frustrates a lot of leadership teams: why a site can absorb repeated intervention — new rotas, new incentives, a new manager, a wellbeing initiative — and stay stubbornly where it was. Each intervention treats one symptom. None of them addresses the thing generating all four.
Why the board pack cannot show you this
The number reported to the board every week hides all of it, and it is not the data’s fault. It is the aggregation.
The same underlying data, cut by site, by shift, by team and by manager, shows you precisely where to look. Most organisations already hold it. Very few present it that way, because the reporting was designed to answer “how are we doing?” rather than “where is this happening, and who owns it?”
Use both numbers properly
The headline tells you the size of the prize. The hot spots tell you where change starts.
Ten thousand people at four and a half per cent feels enormous and structural, and leaders react to it accordingly — either demoralised, because where would you even begin, or complacent, because it looks like the cost of doing business. Four sites and two shift patterns feels like something a competent operational team could take on this quarter.
It is the same problem. The second framing is the one that gets acted on.
And it is worth saying plainly: most organisations assume HR is managing this detail on their behalf. HR cannot. Not because the function is failing, but because the accountability for how a specific site is run has never sat there.
What to actually ask
Four questions, in this order, using the data you already have.
Where was absence worst this summer? Not organisation-wide. By site, by shift, by team.
Why there? Look for what those locations have in common — shift pattern, rota design, physical demands, tenure, recruitment pressure.
Who manages that area? In our experience this is the single most predictive variable, and the one organisations are most reluctant to examine.
What are the other performance metrics doing in the same place? Productivity, customer measures, compliance, turnover, agency spend. If they move together, you are looking at one problem.
It responds to treatment
This is not an abstract diagnostic exercise. Targeted at the right places, it moves.
A comparable large multi-site operator reduced average days lost by 37% over six years, worth approximately £4m a year — not by adding HR headcount, but by giving managers a guided process and giving leadership visibility of whether it was being followed. A large acute NHS trust running a targeted programme in one condition area saw returns to work 55% faster and eleven days off the average case duration, delivering over £1m of savings in six months.
Neither began with an organisation-wide programme. Both began by knowing where to start.
One last thing
The answer to “why there?” is rarely, in our experience, mostly about health.
It is almost always about how those teams are managed. Plenty of people will want to argue with that. Our data supports it consistently, across sectors, year after year.
If your summer data has just told you something you did not previously know, that is worth a conversation.
