Article
Anatomy of a missed clock-in: what one forgotten punch costs, per employee, per month.
A company with 400 employees burns around €32,000 a year on missed clock-ins. It is not fraud and it is not absenteeism: it is people reconstructing from memory hours that nobody can prove any more. And it never shows up as a budget line, because it leaves in eight and ten-minute pieces, spread across dozens of people over the month.
The figure comes from the calculation this article builds from the bottom up, missed punch by missed punch, with data measured in a store in production. Start with one.
At 07:58 on a Tuesday, Marta comes in through the service door with a phone in her hand and her head on Saturday’s rota. She walks past the terminal, does not clock in, and goes to change into her uniform. Nothing special happened. It is precisely because it is not special that it costs so much money.
What happens when someone forgets to clock in
On the old system, clock-ins were only checked at the month-end close. Marta works the full shift and the record is left with a hole: entry missing, exit recorded at 16:04. To the old software, Marta either did not come in, or worked zero hours, or worked eight, depending on who opens the file.
A hole in the record does not fix itself. It gets fixed with people interrupting other people:
- The hole only shows up at the month-end close. The missed punch is already weeks old.
- Accounting asks the store manager; the manager asks Marta, who tries to remember a Tuesday from three weeks ago.
- The manager records the correction on paper, in Excel or in an email.
- Accounting validates it against the rota and fixes the record by hand.
- If the close has already run, the correction reopens the close. And sometimes the payslip.
Five steps for a two-second lapse.
Who touches the error, and how many minutes it takes
| Who | What they do | Minutes |
|---|---|---|
| Employee | Reconstruct the entry time from memory | 3 |
| Store manager | Ask, reconstruct with Marta, record | 8 |
| Accounting | Spot it at the close, cross-check the rota, fix it | 10 |
| Total per error | 21 |
The missed punches are counted in the database. The minutes for each step are our estimate, made by watching the real operation.
The frequency is measured too: in our store, with 36 employees, there were 21 missed clock-ins in 15 days of September 2026. That is 42 a month, or 1.2 per person. That is the rate we apply to the company of 400 employees, which gives 467 missed punches a month. A linear projection, not a measurement.
What the corrections cost
The €11/h is the blend of the three profiles that touch the error: employee, store manager and accounting.
There is a cross-check, measured in the same store: before the system, the close cost 1.5 to 2 days of accounting per month, 12 to 16 hours, and the bulk of that time was exactly this. The calculation above, done missed punch by missed punch, lands at 14.7 hours. Two routes, the same range.
The €11/h is deliberately conservative. The 10 most expensive minutes of every error are carried by accounting, and an accountant’s hour costs considerably more than a store’s blended average. With daily validation, the 12 to 16 hours of close became 3 to 4, so you recover 9 to 12 hours of accounting per month. At €20/h, which is what an external accountant easily charges, that is €180 to €240 a month on that slice alone.
The calculation also does not include the close reopened after processing, the payslip issued with the wrong hours, or the manager doing this on a Saturday. We left them out because we cannot measure them with any rigour. Count them on top of the number, not inside it.
The convenient lapse
Not every hole is born innocent. There is the lapse that arrives 10 or 15 minutes late and would rather leave no trace. There is the card handed to a colleague, who swipes at 07:58 while the owner is still parking. To the old system, none of these is distinguishable from a genuine lapse, and the correction, made weeks later against the rota, tends to fill in the scheduled time. Lateness turns into paid time, leaving no trail.
How many of the 42 were convenience? We do not know, and that is the defining feature of the old system: it leaves no way to find out. So this calculation is written as a hypothesis, and with a number that is an average: ten minutes hidden in each missed punch. Some people always arrive exactly on time and their lapse hides not one minute. Others hide far more than ten. The average is what can be said without inventing, because the old system does not let you tell them apart.
The 10 minutes are an average, and this is where the calculation assumes the most. Think it is only half that in your operation? Halve this line: the numbers change, the mechanism does not.
A camera-verified record closes both doors at once. The card handed to a colleague stops working, because the system records who is there, not who is holding the card. And forgetting stops paying off, because the hole shows up the same day with the real time beside it. The same verification protects the other side just as hard: whoever arrives on time gets their hours proven, and no manager can argue minutes with them over a missing punch.
The hidden cost: disputes and trust
The most expensive minute is not in the table. It is in the conversation where Marta says she came in at 07:58 and the corrected record says 08:15.
It is almost never bad faith on anyone’s part. It is memory against paper, weeks later. But each of those conversations spends something that does not come back: the employee’s confidence that their hours are being counted properly. When an employment dispute does happen, the working time record is one of the first things a court asks for. A record full of manual after-the-fact corrections defends the employer badly and defends the worker badly. It defends nobody.
What changes when the exception is handled the same day
The answer is not to clock in better. People forget: that is a given of the problem, not a character flaw. The answer is to shorten the distance between the error and the correction.
When the system validates the records at the end of each day, Marta’s hole shows up to the manager on Wednesday morning, with context one tap away. Marta still remembers Tuesday. The manager approves it in under a minute, on the phone. Accounting never sees the error at all.
And in most cases the manager is not even needed: 86% of missed punches are filled by the camera on its own, with no human hands. The ratio, counted case by case, is in the case study.
At €11/h, that is €160 a month in the store and €1,771 a month in the company of 400, on the corrections slice alone.
The final bill
Everything added up, per year, with nothing hidden in brackets.
The first line is measured. The second is the hypothesis of 10 minutes, on average, hidden in each missed punch: in some people it is zero, in others it is far more.
Marta’s lapse still happens. It just stops spreading: it dies on the day it was born. That is the difference between managing attendance and doing attendance archaeology.
If you want to run this calculation with your own numbers (your error rate, your hourly cost, your stores), it is a 20-minute meeting. Bring a recent month-end close: we do the maths live.
Guilherme Gameiro
Engineer and Co-Founder of Notchel.