Practical guide
Build a two-person check for new-hire data entry
A preparer and checker routine that catches new-hire data errors before they spread into payroll and benefits.
Published

Table of contents
Treat new-hire data as a chain of dependent records
A new hire creates the same facts in several places. The offer letter holds the title and rate, the onboarding form holds the address and tax data, the payroll system holds the pay code, and the benefits system holds the election. An error in one record can repeat in every downstream system. A two-person check is a way to stop that repetition at the source.
The preparer enters the data from approved documents and marks each field with its source. The checker reviews the entry against the same documents before the record is released. This is not a second pair of eyes on a screen alone. It is a comparison between the entered value and the evidence that authorized it.
Choose the fields that carry the most downstream risk. Legal name, date of birth, address, tax filing status, pay rate, pay frequency, job code, manager, and start date usually belong on the list. Add any field that triggers a payment, a benefit, or a legal reporting obligation. Keep the list short enough to review every time.
Separate the preparer and the checker
Name one preparer and one checker for each batch. The same person should not prepare and approve the same record, even when the team is small. If a second person is unavailable, use a documented exception with a manager review. If the team has only one trained preparer, the manager becomes the checker for that batch. Do not let the exception become the normal route.
The preparer should complete the entry and attach the source documents before asking for review. The checker should receive a ready record, not a half-finished screen. This order prevents the checker from filling gaps and becoming a second preparer by accident. Set a fixed review window so the checker is not interrupted by new entries. It also makes the review time predictable.
The checker should have authority to return the record. A review that can only approve is not a control. Record the reason for every return and the correction made. Track returns by field so the pattern is visible. If the same field is returned repeatedly, fix the source form or the entry rule instead of adding more review.
Use a checklist that follows the record
Build a checklist with the field, the approved source, and the expected format. For example, the legal name comes from the signed offer or identity document, the pay rate from the approved offer, and the start date from the accepted letter. The checklist should tell the checker where to look, not just what to verify. Add one example of a correct entry for the field that causes the most returns.
Add a section for cross-system consistency. Confirm that the payroll job code matches the approved role, the benefits eligibility date matches the start date, and the manager in the system is the person who approved the hire. If a system lacks the field, record the gap instead of assuming it matches. These comparisons catch errors that a single-form review misses.
Keep the checklist with the batch record. The completed checklist shows who prepared, who checked, what was returned, and when the record was released. Store the checklist with the batch, not in a personal folder. Name the folder in the onboarding checklist so the next batch files it correctly. A future reviewer should be able to reconstruct the decision without asking the team to remember the day.
Sample the work, not only the exceptions
A check that reviews only returned records misses the errors that both people overlooked. Each week, sample a few approved records and repeat the comparison independently. Choose records across roles, locations, and start dates. Choose the sample before the week starts so the review is not limited to convenient records. If the sample finds an error, expand the sample for that week.
Track the error types found in the sample and in the normal review. A cluster in tax data may point to a confusing form. A cluster in pay rate may point to an offer template that shows an annual figure without the hourly equivalent. Use the pattern to repair the input, not just the record.
Share the pattern with the preparer and checker without turning it into a blame exercise. The goal is to make the next batch cleaner. Ask the preparer and checker to suggest one form change that would remove the repeated question. A short weekly note with one fix and one reminder is usually more effective than a long training session.
Close the loop after the first pay run
After the first pay run, compare the new hire's pay, deductions, and benefits eligibility with the approved record. Check that the employee appears in the right cost center and that the manager can see the record. Confirm the first payslip shows the expected deductions. This step catches errors that looked correct at entry but failed in the downstream system.
If a correction is needed, record the original value, the corrected value, the source, and the approver. Do not overwrite the history. Tell the employee which value changed and why the change was needed. Send the corrected record to payroll before the next pay run closes. The correction trail protects the employee and gives the team a reusable example for the next review.
Review the two-person check after the first month. Remove fields that no longer carry risk, add fields that caused rework, and confirm that the checker has enough time to do the comparison. Keep the final checklist version with the process notes. A control that is rushed becomes a signature. A control that is scoped and measured keeps new-hire data dependable.
Sources
References supplied with this guide.
| Source | Publisher | Context |
|---|---|---|
| U.S. Department of Labor, New Hire Reporting | dol.gov | Primary reference supplied with this article. |
| U.S. Citizenship and Immigration Services, Form I-9 | uscis.gov | Primary reference supplied with this article. |
| National Institute of Standards and Technology, Cybersecurity Framework 2.0 | nist.gov | Primary reference supplied with this article. |