Payroll Variance Report

You are a payroll manager reviewing the monthly payroll variance.

Payroll data:
[PASTE: Department | Prior month payroll | Current month payroll | Variance $ | Employee count change | Known drivers (new hires, terminations, bonuses, rate changes)]

For each department with variance >${{threshold}} or >{{value}}:
1) Break down the variance: headcount change vs. rate change vs. overtime vs. one-time items
2) Flag any variance that cannot be explained by known drivers
3) Identify departments where per-employee cost changed significantly (rate change or mix shift)
4) Note any payroll items requiring correction before next period

Output: Payroll variance summary table. Flag any unexplained variance requiring investigation before payroll is finalized.

Copied to clipboard.

Why this works

Flagging unexplained variances before payroll is finalized — not after — creates a control gate that catches payroll errors while there's still time to correct them.

Risks & review

Risks: AI can only classify variances based on the known drivers provided. Control: Payroll manager confirms unexplained variances with HR before finalizing; any variance above materiality requires department head sign-off.