Every attendance system has the same quiet failure mode. Someone forgets to clock in. Someone clocks out four hours late because their laptop went to sleep. Someone’s badge doesn’t read on the back door on Tuesday. The data is wrong, everyone knows it’s wrong, and now there is a small bureaucratic argument about how to fix it.
The new attendance regularization workflow is the answer to that argument. Employees raise the correction. Managers approve or reject in one click. The original record is preserved. The audit trail is intact.
The missed-punch problem
Ask any operations lead what their attendance complaints look like and the list is short:
- “I forgot to clock in this morning.”
- “The system logged me out at 2 pm but I worked until 6.”
- “My phone died, the kiosk was broken, the office was on fire — pick one.”
Each of these is a five-minute conversation. The trouble is that there are forty of them a month across a thirty-person team, and every single one currently routes through the same person: an HR coordinator with admin access, who ends up editing the underlying record by hand because nobody else can.
That arrangement has three problems. First, it does not scale — the coordinator becomes a bottleneck as the team grows. Second, it loses information; once the record has been edited, there is no native way to show what it used to be, who changed it, when, or why. Third, it pushes a record-keeping risk onto a single role. If an inspector or auditor asks for the change history six months later, the answer is “let me see what I can pull from the database backup,” which is not the answer anyone wants to give.
A regularization workflow solves all three at once. It moves the request to the person who knows what happened (the employee), the approval to the person accountable for it (the manager), and the audit trail into the system itself.
How regularization works
The flow has three actors and four states.
Employee submits. From their attendance page, an employee opens any day in the past — usually within a configurable window such as the current pay period — and raises a regularization request. They see the original record (clock-in, clock-out, total hours), they propose the corrected times, they pick a reason from a structured list (forgot to clock in, system error, off-site work, kiosk unavailable, other), and they add a free-text note if more context is needed. The submission is timestamped and immutable from that point on.
Manager reviews. The manager gets an email and a Slack notification with the request. They see the original record, the requested change, the reason, and the note, side by side. They approve, reject with a comment, or ask for more detail. The decision is timestamped against their account.
Record updates with provenance. On approval, the attendance record updates to the requested values, but the original values are not overwritten — they are retained on the change-log entry attached to the day. The employee gets a notification confirming the outcome. On rejection, the original record stands and the rejection note is recorded.
Audit trail persists. Every regularization — approved or rejected — leaves a row in the day’s audit history with the before, the after, the reason, the requester, the approver, and both timestamps. The history is visible to the employee, the manager, and HR admins. It does not age out.
The window for raising requests is configurable per account. Typical settings are seven days, the current pay period, or the current month — long enough to catch genuine issues, short enough to stop disputes from drifting in months later when nobody can remember the context.
Why the audit trail matters more than the workflow
The workflow is the visible feature. The audit trail is the one that earns its keep when something goes wrong.
Under the UK ACAS guidance on working-time records, employers are expected to keep “adequate records” of hours worked, particularly where the Working Time Regulations are engaged. “Adequate” is not defined to the minute, but the consistent expectation across employment tribunals is that records can be reconstructed if challenged. A spreadsheet that an HR coordinator silently edits each Friday does not pass that bar. A system that retains the original entry, the change, the reason, and the approver does.
In Australia, the Fair Work Act and the Fair Work Regulations require employers to keep employee records for seven years, including hours worked, leave taken, and any agreed adjustments. The Fair Work Ombudsman’s record-keeping audits look for exactly this: not just current data, but evidence that changes to historical data are documented and authorised. Regularization gives you that evidence by default. There is no separate “audit log we should probably switch on” — the log is the workflow.
The same logic holds for any internal review. If a payroll dispute lands on someone’s desk in November about a Tuesday in May, “the record was changed because the employee said the system logged them out early, the manager approved it, here is the original value” is a complete answer. “I think Sarah might have edited that one” is not.
How this compares to the alternatives
The two common alternatives in this category are admin-only edits and ticket-and-spreadsheet hybrids.
Most mainstream HR platforms — BambooHR, Personio, HiBob, Employment Hero — let admins edit attendance records directly. Some surface a change history; most do not surface it to the employee whose record was changed. The friction is real: every fix routes through HR, employees have no visibility into their own corrections, and the record-keeping story depends on whoever has admin access being disciplined about notes. It works for small teams; it strains at fifty people; it breaks at a hundred and fifty.
The hybrid approach — Slack messages, an internal form, an HR coordinator who edits the record — is more common than anyone admits. It is fast for the people in the loop and opaque to everyone else, including, eventually, an auditor.
A first-class regularization workflow puts the same data in everyone’s hands at the same time, and stores the trail without anyone having to remember to file it.
A note on what we deliberately did not build
We did not add automatic regularization. There is no rule that says “if an employee was in Slack between 9 am and 5 pm, accept their proposed times.” That kind of automation looks clever in a demo and produces fraud cases the first time someone games it.
We did not build a separate manager dashboard for regularization. Pending requests appear on the same approvals queue managers already use for leave. One queue, one habit, one thing to learn.
We did not gate it behind a higher tier. Attendance is included on every Leave Balance plan, and regularization is part of attendance. There is no add-on.
Try it
Regularization is live for every Leave Balance account on the attendance dashboard. If you are running attendance through spreadsheets, admin-only edits, or a Slack channel called #timesheet-fixes, the trial includes everything in this post — workflow, notifications, audit trail, calendar view — for fourteen days, no credit card.
leave emails? Track your employee's leave with Leave Balance
