Month-end job costing is an autopsy. It tells you precisely how the patient died.
A weekly committed-cost view is less precise and far more useful: it's wrong by a few percent while there is still time to do something about it.
Weekly vs. monthly, side by side
| Weekly committed view | Month-end actuals | |
|---|---|---|
| Accuracy | ~95% | 100% |
| Lag | 2–3 days | 18–30 days |
| Can you still act? | Yes | Rarely |
| Inputs | Labor + committed POs | Everything, reconciled |
3 wk
Earliest useful signal on a 16-week job
95%
Accuracy from labor + commitments alone
2 days
Target lag on the weekly view
The cadence
- 01Monday: labor posts from the prior week's approved time.
- 02Tuesday: committed POs and subcontract values refresh.
- 03Wednesday: PM reviews variance over 5% with a named action.
- 04Month-end: reconcile for accuracy, not for decisions.
Frequently asked
- How often should job costing be reviewed?
- Weekly, using committed costs. Monthly actuals are for accuracy and close, not for controlling an active job.
- Isn't a 95% accurate report risky?
- Only if you treat it as final. Used as a control signal, a 95% accurate report available in two days is worth far more than a perfect report available in four weeks.
