Choosing a Timekeeping System: The Questions That Actually Matter
- Section
- Timekeeping & the Law
- Written
- 2026-08-06
- Last checked
- 2026-08-06
Timekeeping selection usually runs on a feature comparison — mobile app, geofencing, integrations, dashboards. Almost none of that determines whether the system will protect you in a wage and hour dispute, which is the only thing a timekeeping system is finally for.
Monitask discusses time tracking tools in this resource.
System requirements should be checked against guidance from the U.S. Department of Labor Wage and Hour Division.
Here are the questions that do.
The compliance questions
Does it retain the raw punch data, or only processed totals?
The single most important question, and it is rarely asked. If the system rounds, aggregates, or applies rules and then discards the underlying entries, you have destroyed the evidence you would need to demonstrate that the processing was neutral. See rounding time entries.
Ask specifically: can we export every raw punch, with timestamps, for any historical period.
Is there a complete audit trail on every edit?
Who changed what, when, from what value to what value, and why. Records that can be altered without a trail are worth substantially less in a dispute — and a system where a manager can silently change an employee's hours is a liability in itself. See correcting time records.
Does it handle the rules of every state you employ in?
Daily overtime in California. Meal and rest premiums. Different rounding restrictions. Seventh-day rules. Split shift premiums. Minor work restrictions.
A system configured for federal rules only will underpay in several states, silently and at scale. Ask the vendor to demonstrate the configuration for your specific states, not to confirm that it is "fully compliant."
Can we turn rounding off entirely?
Some systems assume rounding. If you employ in California, Washington or Oregon, you need it off, and you need it off per state rather than globally.
Can employees see their own records?
Required in some states, good practice everywhere, and the cheapest error-detection mechanism available.
What happens to the data when we leave?
Retention obligations follow the records, not the software. If you switch vendors, you must still produce three years of history. Ask what the export contains, in what format, and whether it includes the raw data and the audit trail — not just the processed totals.
Get this answer before you sign, because it is the question a vendor has least incentive to answer well.
The practical questions
How long does it take an employee to record twenty minutes of work?
If the answer is more than about fifteen seconds on a phone, people will not record short increments, and unrecorded compensable time is your largest exposure. See off-the-clock work.
This is the question that most determines whether your records are complete, and it never appears on a feature comparison.
Does it work on a phone, properly?
Not a responsive web page that technically loads. A usable interface for someone standing in a corridor.
Is it accessible?
Keyboard operable, screen reader compatible, sufficient contrast, usable at 200% zoom. If an employee cannot use the system, you have an accommodation obligation and a compliance gap at the same time. Ask for a demonstration with a screen reader rather than a document claiming conformance.
How are exceptions handled?
Missed punches, missed breaks, corrections. If the exception process is cumbersome, exceptions will not be reported — and an exception rate near zero is a sign that the process is broken, not that nothing goes wrong.
Does it integrate with payroll without manual re-entry?
Manual transfer is where errors enter.
The questions about what it also does
Many timekeeping products now bundle monitoring — screenshots, activity levels, application tracking. Treat that as a separate decision, not a bonus feature.
Ask what is collected by default, and whether monitoring can be fully disabled rather than merely hidden.
Bundling matters because:
- Monitoring triggers notice obligations in several states, and biometric statutes where the clock uses fingerprints or facial recognition. See employee monitoring: what is lawful
- The trust cost of monitoring lands on the timekeeping rollout, which is the thing you actually needed. See rolling out time tracking without destroying trust
- Data collected because a feature was switched on by default is data you will have to explain
Biometric clocks specifically: convenient, effective against buddy-punching, and the source of some of the largest privacy settlements in US employment. Get legal advice before deployment, not after.
Selection process
Write the requirements from your states first. Not from a vendor's feature list.
Test with real edge cases. A missed punch, an interrupted meal period, a shift crossing midnight, an employee working in two states in one week, a correction three weeks after the fact. Vendors demo the happy path.
Involve the people who will use it. A supervisor approving forty timesheets a week will tell you in five minutes what a procurement evaluation will miss.
Run a parallel period. One pay cycle in both systems, comparing results. Discrepancies at this stage are cheap.
Read the contract terms on data. Ownership, export, retention after termination, and what happens if you dispute an invoice.
The one-question version
If you can only ask one thing:
"Show me how I would produce, for one employee, three years ago, every raw punch, every edit, who made each edit and why."
A system that can answer that will survive an investigation. A system that cannot is a payroll convenience, and you should know which one you are buying.