What is time theft on a construction site, and what does it cost?

It is quiet, and it is structural

Not dramatic. Fifteen minutes at each end of the day. One worker clocking in for another who is not there. Hours allocated to no activity at all, which is the most convenient place for an hour to go.

The estimates circulate widely — roughly two per cent of gross payroll, with a large majority of businesses affected by buddy punching. We attribute those loosely on purpose: they come through secondary sources and vendor material, and the order of magnitude is what is usable, not the decimal. Even so, two per cent of payroll on a thousand-worker Gulf site is a six-figure annual number nobody budgeted and nobody will find.

The consequence that is not about money

This is the underrated part. Burned hours are the denominator of every productivity measurement you make. Overstated hours therefore make your crews look SLOWER than they were.

So your disruption claim is understated, because you cannot prove the productivity you actually achieved during the unimpacted period. And the norms you carry into your next tender are quietly wrong, in the direction that loses you the job — or wins you it at a price you cannot deliver.

Time theft is discussed as a payroll leak. Its larger cost is that it degrades the only objective record you have of how good your own crews are.

What actually helps

Punch data taken from the gate rather than re-keyed. Every hour allocated to a cost code, so unallocated drift becomes visible. And exceptions surfaced BEFORE the approval — identical punch pairs, overtime anomalies — because approval-by-habit is how ghost hours pass, and they pass with a genuine signature attached.

Biometric gates eliminate buddy punching outright, at the cost of a throughput problem you must plan for. A gate that cannot clear a 2,000-worker shift change will be propped open, and a propped-open gate records nothing.