The three questions every financier asks
Every conversation about financing a first feature reduces, eventually, to the same three questions. Can this director hold this tone for ninety minutes? Does the film’s central device actually work on an audience? Can this team deliver on a schedule and a budget?
A script answers none of them. A look-book answers none of them. A trailer cut from a script and some stills answers none of them and quietly damages your credibility. Fourteen minutes of finished film answers all three, and it does so in front of the person asking rather than in a document they will skim.
What the short is not
It is not a trailer. There is no montage, no title card promising a feature, and no line of dialogue that exists to sell a bigger version. It is a self-contained film with a beginning, a middle and an end.
It is not a scene from the feature. This is the mistake we see most often. Lifting scene 34 out of a screenplay and shooting it produces something that plays as an excerpt, because it is one. It has no shape and audiences feel the absence immediately.
It is not a calling card. Calling-card shorts are built to demonstrate range — a bit of action, a bit of comedy, a crane shot. This short is deliberately narrow, and demonstrates one thing very precisely.
How you actually get fourteen minutes out of 108 pages
The method that worked, after two that did not: find the smallest complete unit that contains the film’s central device, and build a new film around it.
For us that unit is the basement sequence in the second movement — the night the explosions come close, the family moves underground, and Margarita sees the boy accompanied by a younger sister. It contains the voice, the vision, the marriage under pressure, and the film’s exact tonal problem.
Around that unit we wrote roughly six new minutes that exist nowhere in the feature: an opening that establishes the marriage in three scenes, and an ending that resolves the night without resolving the film. Those six minutes are original work, and writing them was harder than the adaptation.
The test we applied
Could someone who will never see the feature watch this and feel they had seen a whole film? If the answer required any explanation at all, the cut was wrong.
Shooting a short on feature discipline
Six days, three locations, eleven crew. That schedule is deliberately tight in the way a feature schedule is tight rather than the way a short schedule usually is — which is to say, it is the same rate of pages per day we have budgeted for the feature.
This is the part of the exercise that has nothing to do with the audience. If we cannot achieve our page rate on six days with eleven people and one lens, we will not achieve it on thirty days. Better to discover that now, on a fraction of the money, than in week two of principal photography.
We are keeping honest records: setups per day, pages per day, how many takes, where time actually went. Those numbers go into the feature budget as evidence rather than as an estimate, and they will go to our financing partners whether they flatter us or not.
What we are prepared to learn
The genuine value of this exercise is the possibility of a bad result, so it is worth naming in advance what a bad result would look like.
If the camera grammar does not read — if audiences do not intuit that movement means the voice is present — then the feature’s structure needs rebuilding, and we will find that out for the price of six days rather than thirty.
If the voice tips into sentiment, the problem is in the writing of the boy, not the performance, and that is a script problem we can still solve.
If audiences want to know whether the child is real, and are frustrated rather than held by the ambiguity, then the film’s central premise does not work at feature length and we need to know that before anyone commits a feature budget.
We would rather be wrong in March 2026 than in production in 2027. That sentence is the entire argument for making a short first.
The financing logic
Crowdfunding a short is not the same as crowdfunding a feature, and conflating them is how campaigns fail. The short has a small, specific, verifiable budget with a defined deliverable and a delivery date. A backer can understand exactly what their money buys.
The feature will be financed conventionally — co-production, public funds, private equity, sales estimates. The short exists in part to make that conversation possible, because it converts every claim in the pitch deck into something a financier can watch.
There is a version of this project where we spend two more years developing on paper and go to market with a stronger script and no film. We have watched that path from close enough to know how it usually ends.
The two methods that failed first
Before arriving at the approach described above, we spent four months on two others. Both are the obvious ideas, both are widely attempted, and both produce something that does not work.
The compression. Take all 108 pages and reduce them proportionally — every scene shorter, nothing cut. What you get is a film where every beat is present and none has room to land. Read on the page it looks complete; watched, it is a synopsis with actors in it.
The trailer-essay. Fragments from across the feature, cut for tone, held together by atmosphere. This has one fatal property: it can only be judged by someone who already knows the story. Shown to people who did not, it produced polite bafflement and no useful information at all.
Both failures point the same way. A short must be a film, with its own complete shape. Anything that depends on the feature for its meaning cannot test whether the feature works.
What we are measuring, precisely
“See if it works” is not a test. Before shooting we wrote down what we will measure and how, so that we cannot retrospectively decide the result was good.
Comprehension. Screened to two groups who have not read the script. Afterwards, a written question: describe what happened, in your own words. We are looking for whether the camera grammar was absorbed — specifically, whether anyone mentions noticing when the film becomes restless.
Tonal placement. One question: what kind of film is this? We are looking for the absence of the words ghost story and supernatural. If they appear, the boy is being read as a haunting and the feature has a problem.
Where attention goes. Recorded during the screening rather than asked afterwards. Which sections lose the room. Self-reporting is unreliable; watching an audience is not.
Schedule performance. Setups per day, pages per day, takes per setup, hours of overtime. These numbers go into the feature budget as evidence.
The budget, in shape if not in figures
We are not publishing line items, but the proportions are worth setting out because they are unusual and the reasoning may be useful to someone.
Cast and crew take the largest share by a wide margin. Locations and art department come next, and the art department line is mostly clearing rooms rather than filling them — a rare instance where the design concept genuinely reduces cost.
Camera, grip and lighting are conspicuously small, because a one-lens film that refuses artificial fill does not need a truck. That saving is redirected, deliberately, into the sound and post line, which is proportionally larger than a short of this size would normally carry.
Nothing in the crowdfunding budget pays a producer fee. Contingency is six per cent, which is low, and we know it is low.
If the short succeeds
The plan after delivery is deliberately unglamorous. Festival submissions in the first cycle. A private link to backers before any of that. Then the feature financing conversation, restarted with a film attached.
What we are not planning: expanding the short into the feature. They are different films that share a sequence, and the feature will be shot from its own script with its own schedule. A short that becomes act two of a feature is a compromise made in an edit suite, and audiences can feel the seam.
And if it fails — if the grammar does not read, or the tone tips — the plan is to say so here, in this Journal, with the screening data. That is the deal we have made with anyone backing this. It is easy to commit to in March.
The festival question
We will submit the short to festivals, and we are aware that a proof-of-concept short with a feature behind it occupies an awkward position in most programming conversations.
Our approach is to be straightforward about what it is. Programmers are not naive, and a short that pretends not to be attached to a feature is transparent and slightly insulting. It is a complete film that also exists for a reason, and both of those things can be said in a submission note.