The Thing You Are Spending
A clock-in system depends on people using it as intended. What the workarounds look like, and what they cost you.
The boundary · Analysis
Every attendance deployment rests on something not in the specification: people tapping the button honestly rather than working around it.
What working around it looks like
Punching from the car park where the signal is better.
Asking a colleague to punch, which is the problem the system was bought to prevent.
Punching late, at the workstation, and appearing late.
Not submitting corrections, and absorbing the unpaid minutes.
Leaving the app permission denied and using the fallback every day.
Each is a rational response to a specific failure, and each degrades the record the system exists to produce.
What spends the cooperation
A refusal treated as lateness.
Corrections that take days.
A radius nobody will explain.
Discovering the app collects more than was said.
Deducting pay for a technical failure, which is the fastest of all.
One instance of the last, in a small workforce, is known by the end of the shift.
What builds it
Publishing the refusal rate, which says plainly that failures are the system's problem.
Fixing the worst site visibly.
A fallback that takes seconds.
Paying people when the system fails, without being asked.
Showing people how to check the permission themselves.
The asymmetry
Cooperation is slow to build and fast to lose.
And its loss is invisible in the system, which keeps reporting punches while the punches stop describing arrivals.
Which is why the data quality checks matter beyond their technical purpose: a refusal rate that stays flat while fallback usage climbs is telling you something about behaviour rather than about coverage.
The honest tension
Some people will game any attendance system, and pretending otherwise helps nobody.
But the response to one person is not a tightening for everyone, which is the commonest mistake here.
Deal with the individual through the ordinary process and leave the arrangement that works for the other forty alone.
A tightening punishes the compliant and is adapted to within days by whoever it was aimed at.
The measure
Do people submit corrections, or absorb the minutes?
A workforce that reports a misfiring site is cooperating.
One that quietly works unpaid has told you what it thinks of the system, and no configuration recovers that.
Watch corrections, not just refusals
The measure that reveals what people are actually doing.
A flat refusal rate with falling corrections can mean the system improved.
Or it can mean people stopped submitting and started absorbing unpaid minutes.
The two look identical in the data.
Ask a few people. At that point their answer is more reliable than the number.
Check the difficult case
Use employee screen monitoring software to frame one representative test. The useful evidence is the record created when an employee corrects an entry and an administrator exports it.