Film Sales Estimates: How the Numbers Are Built and How to Read Them | American Film Association

Film Finance Guides

Film Sales Estimates: How the Numbers Are Built and How to Read Them

A sales estimate is built from your cast's last three films in five years, priced territory by territory. How the build works, what the cast tiers are worth, and how to test an estimate before it touches your plan.

A sales estimate is a build, not a guess. In Austin Spicer's teaching at the American Film Association, an estimate is constructed from evidence: the last three films your lead actors made in the past five years, what those films earned territory by territory, averaged into a value for each market your film could sell into. Built that way, an estimate is the most useful planning number a producer has. Built any other way, it is a hope with a dollar sign on it. This guide is for producers assembling a package, and for anyone who has been handed an estimate sheet and needs to know whether the numbers on it were built or invented. It covers the build itself, the cast tiers that drive the pricing, the classroom math Spicer uses to teach it, and the tests an estimate has to pass before it earns a place in your financing plan. What an estimate is not, and how it differs from a signed minimum guarantee, is covered in the companion guide, Pre-Sales, Honestly.

The build: three films, five years, territory by territory

The method Spicer teaches his accelerator students is mechanical on purpose. Take the actor you are considering. Find the last three films that actor made in the past five years. Pull what those films did in foreign markets and what they did domestically. Average the record, and you have a working read on what that actor is worth, market by market, for a film like yours.

The territory part is not decoration. Independent films sell across 38 foreign territories, plus domestic, and an actor's value is not the same in all of them. A name that moves Germany may do nothing in Japan. The estimate has to be built the way the sales will actually happen, one territory at a time, because that is how buyers will price the film when it reaches them.

Underneath the whole build sits one driver. A buyer's commitment is driven almost entirely by cast and genre, and in Spicer's words it is "almost like 99% of the time just cast." Genre sets the lane and the ceiling. Cast sets the number. This is why the estimate cannot be built before the cast conversation, and why changing the lead changes every line on the sheet.

The data for the build is public and checkable. Spicer sends his students to Box Office Mojo and TheNumbers for territory-level results on comparable titles and cast records. An estimate that cannot be traced back to that kind of data was not built. It was asserted.

The cast tiers are the pricing engine

Because cast drives the number, Spicer prices cast in tiers, by verifiable sales value rather than fame.

Tier A cast moves numbers internationally. A Tier A name pre-sells foreign territories and unlocks minimum guarantees at meaningful levels, because buyers in many markets will pay for that name on a poster.

Tier B cast has market-specific value: strong in a genre, a faith audience, or a set of territories, quiet elsewhere. Spicer prices a Tier B actor's pre-sale value in a wide band: "They might be sitting at $500,000, right? They might be at a million." Across a whole film, his teaching puts total international minimum guarantees for a contained genre film with Tier B cast at roughly $200,000 to $800,000. That range is the honest shape of Tier B. It is real money, and it does not carry a mid-tier budget by itself.

Tier C cast is talented and carries no established sales value. Per-territory sales run at levels like $5,000 or $10,000. Spicer is clear that Tier C packages can be legitimate and can perform, when they are priced honestly. The failure is dressing Tier C value in Tier A numbers.

Fame is not an input. Social following does not convert into territory value, because buyers price from sales records, not audience counts. Spicer teaches the point with a streaming-era name his students all recognize, an actor with enormous visibility and, on the sales data, almost zero pre-sales value, because there are no territory results to price. He pairs it with a warning from the top of the market: a $50 million film built around one of the most followed actors alive still lost $30 million, and that was including foreign sales. The estimate built on followers is built on nothing a buyer will pay for.

The classroom math

Spicer teaches the estimate with one piece of arithmetic until it becomes reflex. A buyer in a single territory commits $50,000 as a minimum guarantee for the right cast in the right genre. Twenty territories at that level is $1 million in signed guarantees. Against a $3 million budget, that is a third of the budget pre-sold.

Run the same arithmetic through the tiers and the estimate sheet explains itself. Twenty territories at Tier C levels of $5,000 to $10,000 is $100,000 to $200,000, a useful layer and a small one. The Tier B band of $200,000 to $800,000 sits between the two. The total on any estimate is just the per-territory values, added up. Which means every estimate can be audited the same way: line by line, territory by territory, asking what cast record supports that line.

Estimates are not only foreign. On the domestic side, Spicer teaches that a US streaming acquisition for a contained independent film runs roughly $300,000 to $1.5 million depending on the cast. A complete estimate accounts for both: the territory build abroad, and a realistic domestic read at home.

The matrix check

Before an actor's numbers go into an estimate, Spicer has his students score the candidate in the Cast Value Matrix. The matrix grades four things: international sales value, domestic recognition, prestige and festival value, and budget-tier alignment. The point of the exercise is separation. An actor can score well on recognition and badly on international sales value, and it is the second score that prices the estimate.

The matrix also catches the reverse error. A genre actor with a loyal market and no mainstream fame can score modestly on recognition and strongly on the sales criteria that matter for a contained film. Producers who price cast by celebrity miss both directions: they overpay for names buyers will not price, and they overlook names buyers quietly will.

Where an estimate belongs in the plan

In Week 4 of the accelerator, every student rebuilds their equity gap under a realistic pre-sale scenario, using estimates they built themselves. The gap formula is the one Spicer teaches all year: budget, minus the tax credit, minus soft money, minus pre-sales, leaves the equity gap, and that gap typically runs at 50 to 70 percent of the budget. The estimate is what lets a producer see the gap before an investor does, test cast choices against it, and learn whether the film's numbers can work at all.

Then comes the rule he attaches to that homework, and it is the rule that governs every estimate in this guide. The scenario is "for estimation purposes only. This is not something to put on your deck as this is an actual pre-sale. You don't want to confuse the investors by putting that."

Estimates are internal instruments. They guide cast decisions, budget decisions, and the producer's own judgment about whether a package holds together. The moment an estimate is presented to an investor as if it were a sale, it has changed from a planning tool into a misrepresentation, and due diligence treats it that way. Inside the plan, label every layer for what it is: confirmed, targeted, or hoped for. The estimate lives in the second and third categories, never the first.

Reading an estimate in a falling market

Two pressures push estimate numbers up, and a producer reading a sheet should know both.

The first is the market itself, moving the other way. Spicer teaches that pre-sale values have fallen over the past five years. The buyers are still there, but they pay less for the same cast than they did, and an estimate built on older results inherits the older market. Guest speaker Ron Fogelman, a former studio international licensing executive speaking in an AFA masterclass, put the same point harder: the international sales model that films were once financed on "is compromised; sales-agent estimates can no longer be relied on" the way they were when distributor advances could carry a production. An estimate is a forecast made inside that weaker market, and it should be read as one.

The second pressure is the person presenting the sheet. An agent who wants your film has reasons to price it warmly, and the companion guide to this one covers how to question an agent's numbers. One of Spicer's own rules works on any sheet, whoever built it: pay at most half of what a sales agent says an actor will bring. If an actor costs $200,000 and the sheet credits them with $300,000 of value, the deal fails his test before it reaches an investor.

A last word on generated numbers. In class, students practice the build with rough figures to learn the method. A real package is different. Spicer forbids AI-generated figures for budgets and cast sales values in material an investor will see, because a number nobody can source is an automatic no in diligence. The estimate in your plan has to trace to sales data and to agents who sell your genre. If it traces to a prompt, it does not belong in the plan.

Frequently asked questions

How do I build an estimate for my own film?

Use the method: for each lead actor, take their last three films in the past five years, pull the foreign and domestic results from Box Office Mojo or TheNumbers, average the record, and price your film territory by territory from it. Then rebuild your equity gap with the total and see whether the package works. Treat the result as an internal planning scenario, and test it against agents who sell your genre before you rely on it.

Should sales estimates go in my pitch deck?

Not as pre-sales. Austin Spicer's rule for his own students is that the estimate scenario stays off the deck, because presenting an estimate as a sale confuses investors at best. Estimates belong in your internal plan, clearly labeled as estimates. What goes in front of investors is signed paper, or an honestly labeled scenario at the low end.

Why is the estimate total so much lower than my budget?

Because the estimate prices what buyers will pay for your cast and genre in the current market, and your budget prices what the film costs to make. The two numbers answer different questions, and the distance between them is the equity gap your financing plan has to close with incentives, soft money, and equity. An estimate that conveniently equals the budget was built backwards from the answer someone wanted.

How current does the cast data need to be?

The build uses the last three films inside the past five years. Older results describe a market and a career stage that may no longer exist, and buyers will not price from them. If an actor has no qualifying record inside that window, the honest estimate treats their sales value as unproven, whatever their reputation.

Do streaming deals count in an estimate?

Yes, on the domestic side. Spicer teaches a US streaming acquisition range of roughly $300,000 to $1.5 million for a contained independent film, depending on cast. A complete estimate carries both the foreign territory build and a domestic read, each labeled as an estimate until a buyer signs.

The American Film Association teaches film financing structure through its programs and guides. AFA does not provide film financing or capital, and it does not represent films for sale. What it teaches is how to build a number from evidence, so the plan resting on that number can survive the questions investors ask.