Who Might Seek Out This Film? Build an Audience Hypothesis You Can Investigate
Who Might Seek Out This Film? Build an Audience Hypothesis You Can Investigate
Almost every early pitch document contains a version of this sentence: This film is for anyone who has ever felt like an outsider. Or for everyone who loves a good story. The sentence isn’t dishonest. It’s just uncheckable, and a claim nobody can check can’t help you decide anything — not what goes on the poster, not where to submit, not whether to keep going.
The usual next move is to reach for a demographic. Adults 25–44. Urban creative professionals. Art-school graduates. This looks more rigorous because it has numbers near it, but it’s the same problem wearing a different shirt. A category is not a reason.
What you need instead is small and specific enough to be wrong. Not a count of customers. Not a forecast. A hypothesis about why a person would choose this film — and a way to go ask whether that reason holds up.
Describe the attraction before naming the audience
Start with the film, not the people. Before you can say who might watch, you have to say what someone would be choosing.
The attraction is usually one of a few things: a subject they already care about, a conflict they recognize from their own life, a formal pleasure — the particular satisfaction of how this film moves — or a perspective they rarely get to see. It has to be supported by the film you actually have. If you catch yourself writing an attraction your footage doesn’t deliver, that isn’t an audience problem. It’s a film problem, and it’s better to find it now.
Take an invented example, which I’ll carry through the rest of this piece: a quiet observational documentary about night-shift sign painters. Ninety minutes, no narration, no interviews. The camera watches people mix paint, measure letters, redo a curve, and step back to look at what they’ve made. The work is repetitive. Some of the design choices are genuinely hard, and the film lets them stay hard.
The lazy version of the attraction is “for artists.” Even the slightly better “for people who like art” says almost nothing. What is this film offering? Not information about signs. Not the story of a sign-painting business. The offer is close attention to skilled manual work — the pleasure of watching competence at its own pace, and the small drama of decisions being made in real time.
Notice that the attraction says nothing about who appears on screen. This is the first trap. A film about sign painters is not automatically for sign painters. A film about nurses is not for nurses. Some of them may come, and some may be the best possible audience. But the people on screen are the subject, and the audience is whoever wants the experience the camera makes available. Those are different sets, and blurring them produces a paragraph that flatters the film’s participants without describing anyone who would buy a ticket.
There’s a related mix-up worth naming early, because it costs filmmakers a lot of false confidence. A pitch reader — a festival programmer, a funder, a distributor — who responds warmly to your description is not a member of your audience. They’re a professional doing a job, and their approval tells you the description was legible. That’s real information. It is not evidence about viewers.
Try it now: write one sentence naming the experience your film offers, then check it against what you have. If the sentence describes a film you didn’t make, either you have a different film or you have a description problem.
Turn a broad label into an investigable relationship
“Artists” is a category. It can’t be investigated, because it says nothing about behavior. You can’t ask artists whether they’re artists. You can ask what they chose to watch last week.
So convert the label into a relationship: a reason to choose this film, plus a description of where that interest is already being expressed. Those two together are what make a hypothesis investigable.
For the sign painter film, the relationship might sound like this: people who voluntarily watch process — restoration videos, time-lapse builds, long sequences of someone making a thing by hand — and who stay through the slow parts instead of skipping ahead.
Now you have something you can fail at. That sentence describes a behavior, and behaviors leave traces. Somewhere, people are already spending their evenings on this. They aren’t waiting to be discovered; they’re watching something else right now, and the useful question is what.
Seed&Spark’s guide to interviewing your audience before crowdfunding — dated May 19, 2020 — makes a related point within its limits. It recommends asking prospective audience members about their existing interests and discovery habits rather than guessing at those habits. That’s the move. What you’re looking for isn’t permission to market to a group but a sense of where the interest already lives. Keep the limits in view, too: a crowdfunding platform’s practitioner advice is a set of suggestions about asking, not research establishing who will show up, what anyone will pay, or how many of anything there are.
Now name, in advance, what would weaken your hypothesis. A hypothesis you can’t imagine losing isn’t one.
For this film, the obvious weakener is that people who care about the subject may expect instructional content. Painting tutorials are enormously popular, and for specific reasons: they teach technique, they solve problems, they end with something you could try. The observational documentary does none of that. If your attraction is “people who like painting,” you’re probably wrong — not because those people don’t exist, but because “liking painting” bundles several experiences together and your film offers only one of them. If your attraction is “people who like watching skilled work without being taught,” it’s narrower, better, and testable.
One more thing to settle before you ask anyone: whose response could genuinely challenge your idea? The answer is usually a small number of specific people — not a community. And a community is not a headcount. If a craft group has tens of thousands of members, that number is not an audience, not reach, and not yours to use. You don’t have a relationship with a group because it exists. You have a reason to talk with a handful of individuals, one at a time, with their consent.
Ask about choices, not whether the idea sounds good
Most filmmakers ask the wrong question, and they ask it in a way that guarantees a pleasant answer.
Would you watch a documentary about night-shift sign painters?
Almost everyone says yes. It sounds interesting. It costs them nothing to say so. You leave the conversation with encouragement, which feels like data and isn’t. You’ve learned that your description isn’t repellent. That’s the ceiling of what this question can tell you.
Ask about behavior instead. Recent, specific, already-completed behavior. What’s the last thing you watched? Why that one and not something else that night? What did you start and abandon? Where did you hear about it — a friend, a channel you follow, a festival, a stray recommendation? What did you watch right before it? These questions have answers, because they’re about things that already happened. People are good at reporting their own recent choices and bad at predicting their future ones, including their own.
Only afterward do you present the film. Use the accurate description — the one you’d put in a program note, not the one you’d put in a sales email. Then ask what experience the description suggests to them. Not would you watch it. Ask what they think it would be like to sit through. What do they expect to happen? What does the description lead them to expect they’d get out of it? That last question isn’t a forecast of whether they’d attend; it’s a report on what your description is communicating.
The work happens when you compare two things: the attraction you believe the film offers, and the experience a listener hears in your description. When those match, you’ve got something. When they don’t, you’ve got something more valuable.
An invented illustration of the reasoning — constructed here to show how a response can be misread, not a report of any interview that took place. Say a person tells you they watch painting tutorials most evenings and love them. It would be easy to hear that as confirmation: here is someone interested in your film’s subject, and they watch this kind of thing all the time. But look at what tutorial-watching actually is. It’s instruction. A problem at the start, a solution at the end. The steps transfer; you could do it yourself. The viewer’s attention is pointed at the outcome.
Your film asks for a different kind of attention. Duration. Repetition. A design decision that stays unresolved because the painter is still deciding. Nobody is teaching. The pleasure on offer is closer to watching someone work than to learning how. The same subject — paint, hands, craft — is serving two entirely different uses of an evening.
So enthusiasm for tutorials does not establish appetite for this film. It might be a hint: it tells you the subject draws attention, and it gives you a place to ask the next question. What makes you stop a tutorial halfway through? Have you ever watched a long process video all the way to the end, and what kept you there? Those are good questions. They are not the same as an answer.
Record what people actually said, in their words, before you interpret it. The gap between the two is where most audience paragraphs go wrong.
Write the hypothesis with its evidence and next question
Now write it down, in a shape that keeps three different kinds of statement apart: what someone reported, what you make of it, and what you still don’t know.
Here’s the sign painter version, kept visibly provisional:
Hypothesis: People who choose to watch skilled manual work unfold at its own pace — the same attention that holds people through restoration and process footage — may seek out this film for its close, unhurried observation of design decisions being made rather than explained.
What I actually heard: In a few conversations with people I know, several described watching long process videos and said they skip tutorials that feel like instruction. One described stopping a video when it turned into a list of steps. (Three people who know me is not a sample, and I’m not treating it as one.)
What would weaken this: If people drawn to the subject turn out to want technique they can apply, or if process-video watchers mainly want a satisfying finished object at the end, the film’s unresolved, repetitive form will lose them.
Next question: Ask people who watch process footage what makes them leave a video partway through — and whether they’ve ever rewatched one.
That paragraph is smaller than the one you wanted to write. It’s also the one that survives contact with a person who has three other things to watch.
Two rules for revising it. First, when expectations conflict with the film, narrow the attraction — don’t rewrite the film’s description to satisfy the response. If everyone hears “tutorial,” the fix is not to call your film a tutorial. Second, treat every sentence as potentially false. The hypothesis is doing its job when it tells you what to go ask next, and it has finished its job the moment you start defending it.
And the specific failure to avoid, because it’s the most tempting one: turning a handful of friendly reactions into expected attendance. A few expressions of interest are not demand. A group’s membership is not a headcount of willing ticket buyers. Neither is a reason to use market-size language, and both are perfectly good reasons to keep asking.
The paragraph you can defend
An audience hypothesis connects a particular film to a reason to watch. That’s all it is, and it’s enough to work with.
Notice where the sign painter example landed. Not at “people who like art,” and not at a number. It landed at a relationship — between a specific formal pleasure and a specific viewing habit — plus one question capable of breaking it. That’s a small result. It’s also the kind you can carry into the next conversation, and the one after that, revising as you go, without claiming to know more than you do.
You can put a sentence like that in a pitch. You just have to owe it something: the next question, asked honestly, of someone who might say no.
Frequently asked questions
Why isn't a demographic label like 'adults 25–44' enough for an audience hypothesis?
It looks more rigorous because it has numbers near it, but it is still a category rather than a reason. A category can't be investigated, because it says nothing about behavior. A usable hypothesis needs a small, specific reason a person would choose the film and a way to ask whether that reason holds up.
What should be described first: the film's attraction or the audience?
The film's attraction. Start with what someone would be choosing, supported by the footage you actually have. If you write an attraction the film doesn't deliver, that's a film problem, not an audience problem.
How do you turn a broad label like 'artists' into something testable?
Convert it into a relationship: a reason to choose the film plus where that interest is already expressed. For example, people who voluntarily watch process videos and stay through slow parts. That describes behavior, and behaviors leave traces you can ask about.
What kinds of questions should you ask instead of 'Would you watch this?'
Ask about recent, specific, already-completed behavior: what the last thing watched was, why that one, what was abandoned, where it was heard about, what came before. People are good at reporting recent choices and bad at predicting future ones.
What is a common mistake when interpreting friendly reactions?
Turning a handful of friendly reactions into expected attendance. A few expressions of interest are not demand, and a group's membership is not a headcount of willing ticket buyers. The useful move is to keep asking.