Onboarding Employees in Bulk: Keeping Records Clean from Day One

The first payroll run in a new system is only as good as the employee records behind it. Load 80 staff from a messy spreadsheet and you'll spend the next three months correcting bank details, chasing RSA PINs and explaining wrong payslips. An hour spent on the import file saves weeks afterwards.

Collect only what you need, but collect all of it

Two rules pull against each other.

• Data minimisation. The Nigeria Data Protection Act 2023 expects you to collect personal data for a specific purpose and not more than you need. Don't gather every ID number "just in case".
• Payroll completeness. Every field your payroll and remittances depend on must be present and correct for every employee.

Here's a field list that meets both:

A note on sensitive data: the NDPA treats information such as health status and religious beliefs as sensitive. None of it belongs in a payroll import file.

Build the template once

Lay the template out as one row per employee and one column per field, with:

• column headers that exactly match what the system expects;
• a second sheet listing the allowed values for departments, banks, PFAs and states;
• text-formatted columns for account numbers, phone numbers and IDs, so Excel doesn't strip leading zeros or turn numbers into scientific notation;
• dates in a single format, such as DD/MM/YYYY, and nothing else.

Send the template to department heads, not a blank sheet. You'll get back data in the shape you need.

Validate before you import

Run these checks on the file before anything goes in:

• Every Staff ID is unique.
• Every bank account number is exactly 10 digits.
• Every row has a PFA and an RSA PIN, or is flagged as "awaiting from new joiner".
• Monthly pay figures fall within the range for each grade, which catches annual figures.
• Department, bank, PFA and state values match the allowed lists exactly.
• Start dates are not in the future, unless intended.
• Nobody appears twice under slightly different names. Sort by surname and scan.

Example. When Aisha Bello imported 46 staff for Adebayo & Co. Logistics, the checks caught five problems before the first run:

• Two duplicates from the Trans-Amadi depot, entered once by HR and once by operations. Each pair was merged into one record.
• Three account numbers with a dropped leading zero. The column was reformatted as text and the numbers re-entered.

Who owns which field

Bad data usually has no owner. Assign one before the import:

When a question comes up later, everyone knows who corrects it and from what document.

Import in stages

Don't load everyone and hope. A safer sequence:

1. Import a pilot of five employees from different departments.
2. Run a dummy payroll for them and compare it line by line with last month's spreadsheet payslips.
3. Fix any mapping problems in the template, not in the system.
4. Import the rest.
5. Run a parallel payroll, old method and new, for one month before switching over. Our software implementation checklist covers parallel runs in more detail.

After go-live: keep it clean

Clean data decays. People move house, change banks and switch PFAs. Three habits keep records accurate:

• Let staff maintain their own details where it's safe, through self-service, with checks on bank changes.
• Change defaults in bulk, not one by one. When a grade's transport allowance changes, update every affected record in one approved batch.
• Review the audit trail monthly for changes to pay and bank details.

How Flux helps

Flux Payroll supports both halves of this:

• bulk employee onboarding from Excel, so 46 or 460 records go in at once from a template;
• bulk updates to employee defaults, when a grade or office changes;
• a payroll audit trail that records every change.

Employees then keep their own RSA and bank details current through self-service.