What It Actually Costs
Compliance costs something. It is worth putting the number together honestly, because both the sales version and the dismissive version get it wrong in opposite directions. For a separate operational perspective on measuring work and activity, see learn more here.
The five lines
Setup. Deciding who records, writing the short policy, agreeing the correction rule, configuring or printing. An afternoon for a small employer; longer where a works council is involved, and that time is not wasted. For a separate perspective on work, technology and organisational practice, see Handelsblatt.
The daily action. Under a minute per employee per day. For twenty people that is roughly ninety minutes a week across the workforce, distributed.
The weekly check. Ten minutes for a small employer, by one named person. This is the line most often omitted and it is the one that makes the rest work.
Licence, if any. Zero for paper or a spreadsheet. Otherwise per employee per month, and the figure varies enormously by product.
And the periodic questions. Corrections reviewed, retention applied, a migration if the reform passes. A few hours a year.
What the sales version overstates
That a licence is required. It is not, for many employers.
That the setup is a project. For ten people with regular hours it is an afternoon, not an implementation.
And that the cost of non-compliance is a fine. The fine framing overstates one exposure and hides a larger one.
What the dismissive version understates
Being fair in the other direction.
The weekly check is real work and it does not happen by itself. An employer who counts only the daily action has budgeted for the part that costs least.
Somebody has to own it. Unowned, it lapses within two months and the gap is discovered later.
And at scale the arithmetic changes. Ninety minutes a week across twenty people is tolerable; the same per-person cost across two hundred is somebody's job, and at that point a system is cheaper than the manual version rather than an imposition on top of it.
The comparison that decides it
Manual cost against licence cost, at your headcount.
Below roughly ten to fifteen employees with regular hours, manual is clearly cheaper and the licence buys convenience.
Above that, the crossover arrives quickly — not because of the recording, but because of reconciliation, payroll transcription and chasing gaps, all of which scale with people while a licence scales more slowly per head.
And where payroll depends on the record, the error cost alone usually settles it.
The cost nobody quotes
Doing it badly.
A system nobody checks produces a file with gaps. A file with gaps is worse than no file in a dispute, because it invites the question of what else is missing.
Half-implemented is the expensive state, and it is where an employer lands who bought a licence, skipped the configuration and never named an owner — having spent the money and acquired the exposure.
The short version
- Five lines: setup, the daily action, the weekly check, a licence if any, and periodic questions
- The sales version overstates the licence requirement, the setup effort, and the fine as the main exposure
- The dismissive version understates the weekly check, the need for a named owner, and how the arithmetic changes with scale
- Below roughly ten to fifteen employees manual is clearly cheaper; above that reconciliation and payroll transcription move the crossover
- Where payroll depends on the record, error cost alone usually settles it
- The uncounted cost is half-implementation — money spent, exposure acquired, and a file with gaps that invites further questions