Most budgets that go wrong were wrong on day one. Not because someone typed the wrong rate, but because the budget was written before anyone knew what was actually being shot. The numbers were plausible, the topsheet added up, and none of it described the production that eventually turned up on set.
So before any of the mechanics below, the rule that matters most: you cannot budget what you have not scheduled. A budget is a pricing exercise laid over a plan. Without the plan, you are guessing with decimals.
Start with the breakdown and the schedule
The order is always the same:
- Break down the script or format. Every scene, every location, every cast member, every vehicle, stunt, visual effect and special requirement.
- Build a schedule. How many shoot days, in what order, in which locations, with which units.
- Then price it. The schedule tells you how many days you are paying for crew, kit, catering, transport and locations. Almost every below the line number falls out of it.
If you skip step two, you end up guessing at the one variable that drives the largest share of the cost. A single extra shoot week on a mid-size drama moves crew, equipment, catering, transport, per diems and location costs all at once. That is why an unscheduled budget is not a budget, it is an estimate of an estimate.
For unscripted and factual work the logic is the same, the breakdown just looks different: number of episodes, shoot days per episode, contributor days, studio days, rig and de-rig, post weeks.
The shape of a budget: the topsheet
A production budget is a hierarchy with three levels:
- The topsheet, one page, one line per account category, the totals that get shown to financiers and broadcasters.
- The accounts, each category broken into its component lines.
- The detail, the individual calculations behind each line (rate, quantity, units, period).
Everything above is a summary of everything below. If a financier asks why the transport number moved, you should be able to walk from the topsheet down to the day rate on a driver and back up again without leaving the document.
The topsheet is conventionally split into four blocks:
| Block | What it covers | Typical share |
|---|---|---|
| Above the line | Story and rights, producers, director, principal cast | 25 to 35% |
| Below the line | Crew, equipment, locations, transport, everything physical | 40 to 50% |
| Post production | Editorial, VFX, sound, music, grade, deliverables | 10 to 20% |
| Other | Insurance, legal, completion bond, financing costs, contingency | 5 to 10% |
Treat those percentages as sanity checks, not targets. A VFX-heavy series and a single-location documentary will sit nowhere near each other, and both can be correct. The value of the ranges is that a budget sitting far outside them is worth a second look before it goes out.
Use a real chart of accounts
Every budget needs a fixed numbering scheme, and it needs to be the same scheme your accounting system uses. This is the single most under-rated decision in the whole process.
A chart of accounts does three things:
- It makes budgets comparable across productions, so last year's drama tells you something useful about this year's.
- It lets actual costs land against the right budget line automatically, instead of being matched by hand every week.
- It makes the cost report possible at all. If invoices cannot be coded to a budget line, nobody can tell you where you stand.
Whether you follow a broadcaster's scheme, a national film fund's, or your own house standard matters far less than following one consistently. What kills you is a chart of accounts that changes shape between the budget, the ledger and the cost report, because then somebody spends every Friday reconciling three documents that should have been one.
Above the line
Above the line covers the creative deals that are usually negotiated before the production has a fixed shape: rights and script, producer and executive producer fees, the director, and principal cast.
Two things make ATL different from the rest of the budget:
- It is mostly fixed early. These are contracts, not estimates. Once signed, they do not flex with the schedule.
- It is where the deals hide. Deferrals, back-end participation, overages if the shoot runs long, escalators tied to episode counts. Budget the cash cost, and note the rest somewhere visible. A deferred fee is not free, it is a cost with a later date.
Below the line
This is the physical production, and it is where a schedule turns into money. Broadly:
- Crew, by department, at daily or weekly rates, for prep, shoot and wrap
- Cast beyond principals: supporting, day players, background, stunts
- Equipment: camera, lighting, grip, sound, on rental periods that are rarely the same as your shoot days
- Locations: fees, permits, unit bases, security, making good
- Art department: construction, set dressing, props, graphics
- Costume, hair and makeup
- Transport: vehicles, drivers, fuel, travel and accommodation
- Production office: facilities, comms, catering, health and safety, welfare
The mistakes here are almost always about periods, not rates. A camera package quoted on a four day week, prep days that nobody counted, a rental that runs from collection to return rather than from first shoot day to last. The rate is easy to check. The period is where the money leaks.
Fringes are not a rounding error
Fringes are the employer-side costs sitting on top of gross wages: payroll taxes, social security, pension and holiday pay contributions, union and guild payments, and any agreed benefits.
Depending on territory and the agreements in play, fringes commonly add somewhere between 20 and 40% to labour cost. On a production where crew is half the budget, getting the fringe assumption wrong by five points is a six figure error on a mid-size drama.
Two practical rules:
- Calculate fringes as a percentage on the line, not as a lump at the bottom. When the schedule changes and crew days move, the fringes have to move with them automatically. A single fringe line at the end of the budget will quietly go stale.
- Different labour categories carry different fringe rates. Local crew, foreign crew, self-employed contractors and loan-out companies are not interchangeable. One blended rate across all of them is a convenience that eventually costs you.
Contingency, insurance and the rest
The final block holds the costs that exist because the production exists rather than because of anything you shoot: insurance, legal fees, audit, the completion bond if the financing requires one, financing costs and interest, and contingency.
Contingency is the one people argue about. It sits somewhere in the 5 to 10% range for most productions, and the number matters less than the discipline around it. Contingency covers what could not reasonably have been foreseen. It is not a slush fund for things you chose not to price properly.
We wrote about this separately in how much contingency a production needs, including the question of who is allowed to spend it.
Decide how detailed to be, then be consistent
A common failure mode is a budget that is forensic in one department and a single round number in another. It usually reflects who was available to consult rather than where the risk actually sits.
A workable rule: detail follows risk and size. A line that is large, volatile, or contractually complicated deserves calculation. A line that is small and stable can be a considered estimate. What you should not do is mix a bottom-up build in one department with a top-down guess in another and then present the total as if it carries one level of confidence.
Build versions, not one file
A production budget is never approved on the first pass. It goes out, comes back with a target, and has to be rebuilt around a different shoot length, a different location, a different episode count.
The teams who handle this well keep scenario versions side by side, so the conversation is "here is the same production at 34 days and at 30, and here is exactly what falls away" rather than "give us a week and we will rebuild it". The teams who handle it badly end up with budget_v7_FINAL_rev3_JB.xlsx and no reliable way to explain what changed between versions.
Version control is not administrative tidiness. It is what lets you answer a financier in an hour instead of a week.
What happens the moment the budget is locked
A locked budget immediately starts being wrong, and that is normal. From that point the useful question is no longer "what did we plan to spend" but "what will this now finish at". That is the job of the cost report, which compares the locked budget against what has been spent, what has been committed, and what is still to come.
The budget only earns its keep if it is built so that the cost report can be produced from it. Same account codes, same structure, same level of detail. When the budget and the cost report are two separate documents maintained by two different people, the reconciliation between them becomes somebody's entire week, every week.
That is the argument for keeping the budget, the commitments and the actuals in one connected system rather than in a chain of exports: not that spreadsheets cannot model a budget (they model them very well), but that a budget disconnected from what is actually being spent stops telling you anything by about week three of the shoot.
The short version
- Schedule first, price second
- One chart of accounts, used by the budget, the ledger and the cost report
- Fringes on the line, not at the bottom
- Check periods harder than you check rates
- Detail in proportion to risk
- Keep versions, so you can explain what changed and why