Ask a producer mid-shoot how the production is doing financially and the honest answer is never the budget total. The budget was locked weeks ago and has been quietly wrong since roughly the second day of prep. What they actually need is a single number: what will this cost when everything is invoiced and delivered?
That number is the estimated final cost, and producing it every week is the entire purpose of the cost report.
The columns, and what each one is for
A cost report is a table with one row per budget line and a fixed set of columns. Naming varies between territories and house standards, but the structure is remarkably consistent:
| Column | What it means |
|---|---|
| Budget | The locked, approved figure for this line |
| Actual to date | Money actually paid out and posted to the ledger |
| Commitments | Purchase orders and contracts signed but not yet invoiced |
| Estimate to complete | What is still to be spent from here to delivery |
| Estimated final cost | Actual + committed + estimate to complete |
| Variance | Budget minus estimated final cost |
The arithmetic is trivial. The judgement is not, and it lives almost entirely in one column.
Actuals are the easy part
Actual to date is the least interesting number in the report, though it takes the most administrative effort to get right. It is history. It tells you what has already left the account, and there is nothing to be done about it.
Its only real difficulty is coding. An invoice that lands against the wrong budget line makes two lines wrong at once, and by the time anyone notices, the pattern has usually repeated for a month. This is why the chart of accounts has to be the same in the budget and in the ledger. Where it is not, somebody spends every week mapping one to the other by hand, and the report arrives on Thursday describing last Friday.
Commitments are what most people miss
A purchase order signed on Monday for equipment arriving in three weeks is not an actual. Nothing has been paid. But the money is gone in every sense that matters: you are contractually obliged, and you cannot spend it twice.
Productions that track only actuals against budget consistently believe they are in better shape than they are, right up until a wave of invoices lands. The commitment column exists precisely to close that gap, and it is the first thing that goes missing when cost reporting is done in a spreadsheet maintained separately from the purchase order process.
If your PO system and your cost report are different documents, your commitment column is a manual transcription, and manual transcriptions decay.
The estimate to complete is where the real work is
Everything else in the report is arithmetic on facts. The estimate to complete is a forecast, and it is the reason a good production accountant is worth what they cost.
It has to be revised every week against what has actually changed:
- Schedule slippage, and what an extra day costs across every department
- Overtime patterns that have become the norm rather than the exception
- Weather days used against weather days allowed
- Scope changes agreed with the director or a head of department but not yet costed
- Rate changes, currency movement on foreign spend, supplier price increases
- Departments consistently running under, which is information too
The failure mode is mechanical rolling forward: taking last week's estimate to complete, subtracting what was spent, and calling it this week's. That produces a report that always agrees with itself and never tells you anything. If the ETC only ever moves by the amount that was spent, nobody is forecasting, they are subtracting.
Estimated final cost is the number that matters
EFC is what the production will cost. Not what was budgeted, not what has been spent. When a financier, a broadcaster or a completion guarantor asks how it is going, the EFC is the answer, and the variance against budget is the follow-up.
The point of producing it weekly is that it is actionable while there is still time to act. An EFC that is 6% over budget in week two of a nine week shoot is a problem with solutions. The same number discovered in the final week is a phone call about who is paying for it.
This is the entire economic argument for cost reporting. Not compliance, not reporting obligations, but the fact that variance found early is cheap and variance found late is not.
Variance without commentary is noise
Every cost report should arrive with a written variance commentary from the line producer or production accountant. Not a description of the numbers, which the reader can see, but the reasons behind them:
- Why a line moved
- Whether the cause is one-off or structural
- What has been done about it
- Any transfers between accounts, and the reasoning
- Anything expected to move next week
A report without commentary produces a predictable conversation in which somebody reads a number out loud and somebody else asks what happened, and the meeting becomes the commentary. Writing it down first is faster, and it creates a record of decisions that is extremely useful three months later when nobody remembers why the art department got an extra allocation.
Transfers, and the discipline around them
Departments run over and under. Moving an underspend in one account to cover an overspend in another is normal practice, and it is also the mechanism by which a budget silently stops resembling reality.
Two rules keep it honest:
- Every transfer is recorded, with a reason, so the original budget stays visible underneath. A budget that has been quietly rewritten to match the spend is not a budget any more, it is a description.
- Transfers do not touch the contingency. Contingency is drawn against deliberately and separately, with a decision behind it. Absorbing overspend by quietly eating contingency is how a production arrives at week seven with no reserve and no record of where it went.
Weekly, and on time
Cost reports are conventionally weekly during production, and the day matters more than people expect. A report covering the week to Friday that arrives the following Thursday describes a situation that is already six days stale, on a production where a single day changes the picture.
Most of the delay is not analysis. It is assembly: pulling actuals out of the accounting system, chasing purchase orders, re-keying numbers between documents, reconciling a chart of accounts that does not quite match. On productions where the budget, the commitments and the actuals live in the same system, the report is largely a by-product of work already done, and the accountant's time goes into the estimate to complete, which is the part that actually requires a human.
The short version
- Actual to date is history, commitments are money already gone, ETC is the only forecast
- EFC = actual + committed + ETC, and it is the number people are really asking for
- Variance without written commentary trains everyone to ignore the report
- Record transfers, and keep contingency separate from them
- Late reports are not reports, they are minutes
Related reading: how to build a production budget and how much contingency a production needs.