Figures that describe something an organisation actually did are onthe published use cases, sourced to a EUROCONTROL guideline and quoted with the unit size and the roster cycle they came from. That is the difference this site keeps: a model explains a mechanism, a case study records an outcome, and they never share a page or a visual treatment.
Evidence
Models, kept separate from results
These are ways of thinking about rostering economics. They are not outcomes anyone achieved, and they are not on this site to look like evidence. They are here because the reasoning is useful and because the previous version of this website presented material like this in the same format as real case studies, which was a mistake worth correcting in public.
The cost of a manual swap process
modelled- The shape
- A mid-sized operation handling several thousand swap requests a year, each one taking a supervisor some minutes of form-filling and roster editing.
- What it shows
- Multiply handling time by volume and you get a staffing cost that nobody has on a budget line, because it is spread across people who were hired to do something else.
- What it does not show
- The arithmetic is simple and the inputs are yours. A published version of this shape, with a real customer's volume, appears in the EUROCONTROL case studies.
Overtime as a distribution problem
modelled- The shape
- A unit that is short of people and covers the gap with overtime, unevenly, so a minority of staff absorb most of it.
- What it shows
- Two things are worth separating: the total overtime, which is a staffing question, and its distribution, which is a rostering question. Redesigning the pattern can move the second without changing the first.
- What it does not show
- How much is available to recover depends entirely on how uneven the starting point is. A well-distributed roster has little left to gain here.
What a qualification expiry costs
modelled- The shape
- An endorsement lapses unnoticed. The person stays on the roster for a position they are no longer current for until someone catches it.
- What it shows
- The cost is not the administrative fix. It is the coverage gap created when the correction lands, usually at short notice, and the overtime used to fill it.
- What it does not show
- This is a risk model, not a saving. It describes an event that may not occur in a given year.
Where the real numbers are
Run the model on your own numbers.
Bring your swap volume, your overtime distribution or your qualification calendar. Forty-five minutes is usually enough to find out whether there is anything there worth pursuing.