Labour costing reconciliation

The situation

Labour hours were recorded in two places. The finance system held what was costed against each project. The rostering system held what people were actually scheduled to work. Nobody could compare them, because the two exports shared no common shape and reconciling them by hand across an entire team was not realistic work for anyone to do monthly.

The gap between the two, once measured, was approximately $3 million. That is not $3M lost. It is $3M of labour expense that could not be confidently attributed to the work it belonged to, which makes budgeting for the next project an educated guess.

What we built

A reconciliation tool that ingests both raw exports, cleanses them, and lays them side by side per person, per day.

• Both source extracts load untouched. No manual re-keying, no reformatting step for anyone to forget.
• Name, date and cost-centre matching is handled in the tool, including the inconsistencies that exist between any two systems that were never designed to talk.
• The output is a single grid: rostered hours against costed hours, with the variance surfaced and sorted so the largest problems are the first thing you see.

The point was never the spreadsheet. The point was making the discrepancy small enough to act on, one row at a time, instead of a number too large to start on.

The result

The discrepancy was identified and worked down. Labour costing completion for the team reached a management-reported 96% for the financial year, described internally as an exceptional result.

The lasting benefit is not the one-off cleanup. It is that labour expense is now attributable to the work that incurred it, which makes forward budgeting a calculation rather than an estimate.