Attendance & payroll

Month-end attendance: reconcile punches, daily results and monthly totals

Attendance & payroll · 3 min read · Updated 6 October 2026

A plausible monthly total can hide incorrect days. Follow the data from raw punches through assigned shifts and approved requests to daily results and the monthly summary. This proposed workflow gives HR and shift managers a shared review process.

1. Fix the scope before comparing numbers

Record the company, departments, attendance period, date range and employee population. If the business keeps two attendance datasets, select the same dataset in every report. Review mid-period starters and leavers against their effective dates rather than treating a full month as the default denominator. Save an export and its timestamp before recalculation. Use it to explain changes after importing punches or approving requests. Comparing yesterday’s export with today’s recalculated results without recording that difference creates misleading exceptions.

2. Check punches and assigned shifts

• Confirm that the employee’s card mapping was valid on the punch date. Identical codes from different sites need source verification.
• Check the latest available log date for each device, as well as overall record counts.
• Flag duplicates, single punches and next-morning exits for overnight shifts.
• Review shift changes, breaks and punch collection windows against the agreed rules.
• missing exit does not establish a full-day absence. Put the day into an exception list with its punches and assigned shift. Obtain supporting confirmation before adding a missing time point.

3. Compare requests with actual attendance

An employee with a full set of punches and a half-day leave request needs review. Counting eight worked hours and four leave hours against one eight-hour shift creates overlap. HR must establish whether the request is outdated or the employee worked during the requested absence. For overtime, compare requested times, approval status, actual punches and payable hours side by side. Differences may reflect a post-shift break, partial attendance or a manual change. A monthly overtime total cannot establish the original time points for a particular day.

4. Use checks with explicit assumptions

Illustrative example: an eight-hour normal shift includes two hours of credited leave, 0.25 hours of lateness and no early departure. Where these intervals do not overlap under the agreed rules, worked time is 5.75 hours. The four components sum to eight. This is a data check, not a universal formula for every day type. Compare daily and monthly totals for the same employee and period. Investigate duplicate rows, position-change splits, date coverage and data versions before changing a total. Use the actual standard hours for shifts of different lengths.

5. Prepare the period-close handover

• Every exception has an owner, decision and supporting evidence.
• Relevant requests are processed; unresolved requests remain listed separately.
• Manual changes record the reason, editor and before/after values.
• Employees or managers review attendance within their own scope.
• Save the final version and agree who may reopen the period and trigger recalculation.
• rn recurring exceptions into corrective work on card mappings, shift settings or approval timing. Adjusting only a monthly total leaves the underlying problem ready to recur.