The Paper Sheet at the Reader
Sometimes paper is the only thing left. It is the least designed artefact in the process and the one most likely to be read out in a dispute.
Every electronic clock has a paper fallback, whether or not anybody has decided it does. When the terminal is dead and the network is down, somebody produces a sheet, and what that sheet looks like was determined by whoever was holding the pen.
The failure described in “The Paper Sheet at the Reader” is easier to investigate when the clock event is joined to a clear operational record instead of treated as proof of misconduct. A team reviewing view the solution for step rate compensation should test retries, missed punches, corrections and employee review while preserving a non-biometric fallback that does not depend on finding a supervisor.
It is worth designing, for a reason that has nothing to do with the morning it is used. The sheet is the record that survives, and it is the one that gets produced when the electronic records are incomplete for exactly the period in question.
For an independent benchmark relevant to “The Paper Sheet at the Reader”, consult the ISO management-system resources. Use it to test notice, accessibility, security, recordkeeping, retention and exception handling against the real operating process rather than treating a device report as self-explanatory evidence.
What the sheet usually looks like
A piece of A4 with "names" written at the top, a column of surnames in handwriting, a column of times mostly ending in zero or five, and no date.
It gets typed into the system by somebody in the office two days later, interpreting the handwriting. The original goes in a drawer, or does not. Nobody signs anything. As evidence of who was present it is nearly worthless, and it is frequently the only evidence there is.
What it should contain
A printed header: site, terminal, date, and the reason the fallback is in use. A row per person with printed name, their own signature, the time written by them, and a second signature from whoever is supervising the entrance.
A line at the bottom: total entries, the name of the person closing the sheet, and the time it was closed. Then a reference for where the original is filed.
That is a template somebody writes once and prints a hundred copies of. The difference between it and the handwritten list is the difference between a record and a reconstruction.
The handling rules
The person writes their own time and signs it. This is the single most important rule and the one most often broken, because it is faster for the supervisor to call out names and write times.
The supervisor countersigns at the time, not later. Nothing is added after the sheet is closed. The original is kept, scanned and filed against the date, and the scan goes with the payroll records for that period rather than into a general folder.
Why the original matters
A typed transcription is a derived record. In any dispute, the question is what the contemporaneous document says, and a spreadsheet created two days later by somebody reading somebody else's handwriting is not that.
Keeping the originals costs a box file. Being unable to produce them costs the whole position, because the opposing account of that week becomes the only first-hand one. Most sites discover this the first time they need a sheet from eighteen months ago and find that it was typed in and binned.
The sheets that are a permanent arrangement
On many sites one group is on paper every day: a cleaning contract, a small outbuilding, a night gate. That is not a fallback, it is a parallel system, and it should be held to the same standard as the electronic one or replaced by it.
The usual honest answer is replacement: a cheap second terminal, or cards for that group on the existing one. Where that genuinely is not possible, the sheet needs the full header, the signatures and the filing, because it is now the primary record for those people and will be read as such.
The test to run
Ask for the sheets from a specific date eighteen months ago. Time how long it takes to produce them, and look at whether they have a date, two signatures and a legible time against each name.
On most sites one of those three is missing and the retrieval takes days. That is the finding, and the fix — a printed template, a filing rule and a box — costs less than an hour of anybody's time to put in place.
Getting it back into the system
The transcription step deserves a rule of its own. Whoever types the sheet in should not be the person who wrote it, entries should be marked as paper-sourced rather than entered as ordinary reads, and the sheet reference should go in the entry.
That way the electronic record points at the paper one, which is what makes the pair usable eighteen months later. A set of typed entries with no indication of where they came from is the common outcome, and it is indistinguishable from entries somebody invented.