Training as a Way of Moving a Constraint
How to pick the training that actually changes a rota, why most training budgets do not, and the arithmetic that makes the case in one line.
What two days of training did to one site's rota, measured over the following quarter
The constraint cost about £2,900 a quarter and £740 removed it. The last line was not expected and turned out to be the one the rota builder noticed first. This is one employer's own before-and-after, not a benchmark.
Training is the main lever for moving a skills constraint, and most training budgets are spent in ways that move nothing. The money goes to whoever asked, or to whatever is due for renewal, and the constraints that determine the rota are untouched at the end of the year.
The skill constraint in “Training as a Way of Moving a Constraint” should remain visible separately from simple headcount. For organisations assessing follow this supporting Monitask page in relation to ethical employee monitoring, work and attendance records can support the operational picture without replacing qualification records or professional judgement.
Choosing differently requires one input: a list of capabilities with the number of holders against each.
The assumptions behind “Training as a Way of Moving a Constraint” can also be checked against GetApp employee-scheduling directory. Record what applies locally and what requires qualified advice.
The rule for choosing
Spend on the capabilities with the fewest holders, in order, and stop when every capability on the critical list has three.
- List capabilities by number of holders, fewest first.
- Strike out any that cannot be trained inside a quarter; those are a recruitment question.
- Spend on the rest in order until every critical capability has three holders.
That rule is crude, it ignores individual development, and it will change more about a rota than any other use of the same money. It is best used as a first claim on the budget rather than as the whole of it.
The second input: pattern coverage
Holders are not useful if they are all on the same pattern. The question is not how many people hold a capability but how many of the shift types have one available.
A capability held by four people who all work days is a day capability. On the rota it behaves exactly like a constraint, and the training that fixes it is training somebody who works nights.
Why the budget rarely goes there
Because training requests come from individuals and constraints come from the rota, and nobody holds both lists at once.
The fix is a standing agenda item rather than a process: once a quarter, the rota's thinnest capabilities are read out in the same meeting where training is allocated. That single change reallocates most budgets within a year.
The backfill problem
Training costs twice: the course and the shift somebody is not working while attending. The second is usually larger and it is the reason training gets postponed in a short-staffed team.
Backfill should be in the number from the start. A case built on course fees alone collapses the first time somebody asks who is covering, and the honest total is still almost always smaller than the constraint.
Training the right person
Three criteria, in order: do they work a pattern that needs the capability, will they still be here in two years, and do they want it.
The third is last and is not optional. A reluctant holder is a holder who finds reasons not to be scheduled, and the constraint reappears wearing a different name.
The in-house route
Many rota constraints are removable by internal sign-off rather than external courses: a period of supervised practice, an assessment, a record.
Where that route exists it is faster, cheaper and better matched to the work. Where it does not exist, creating one for the two or three capabilities that constrain the rota is a project of about a week.
What to measure afterwards
The same numbers as before, one quarter later: shifts with only one qualified option, overtime generated by the gap, leave requests refused for cover reasons.
If none of them moved, the training was for a capability that was not constraining the rota. That is worth knowing, and it is the only way the next year's allocation gets better.
The effect nobody predicts
Rota build time. Removing a single-holder constraint usually takes twenty minutes a fortnight out of the build, because the hardest part of building a rota is fitting the shifts that only one person can work.
It is a small number and it is the one the person doing the work feels, which makes it the number most likely to produce another round of training.
The capabilities that cannot be trained quickly
Some constraints are held by qualifications that take months or years, and no budget decision this quarter will change them. Treating those the same way as a two-day ticket wastes the exercise.
Split the list in two at the start: capabilities that can be added inside a quarter, and capabilities that are a recruitment question. The first list is this year's plan. The second belongs in the conversation about the establishment, where it will take a year and should start now.
The order to spend in
Thinnest capability first, then the one whose single holder is closest to retirement or to the end of a fixed term, then the one that appears in the most shift types.
Those three criteria agree with each other surprisingly often. Where they disagree, the second wins: a constraint that is about to get worse on a known date is the only one with a deadline attached.
Recording each training decision
For each training decision: which constraint it was meant to move, the before figures, the cost including backfill, and the after figures one quarter later.
Four lines per decision. After two years that record is the most persuasive document in the department, because it is the only one that connects training spend to something anybody cares about.