Two Plausible Audiences, Two Different Films in the Pitch: Which Hypothesis Holds?

<!-- ARTICLE BODY START -->

Compare the experiences the two audience hypotheses promise, not the apparent size or enthusiasm of the groups. Put the same accurate film description in front of the comparison. Ask what people expect to watch, why that experience interests them, and which expectations the actual project cannot meet.

If one hypothesis works only when the pitch implies scenes you do not have, it does not support the film you are making. You need to narrow the invitation or acknowledge a creative decision. Better targeting cannot supply the missing film.

The useful result is a defensible choice about the pitch: emphasize one reason to watch, retain different compatible reasons, change the project explicitly, or leave the question open. This is not a way to turn informal conversations into a market forecast.

Write the conflict before comparing the groups

Consider The Last Repair, an invented observational documentary about two siblings closing their late father's workshop. Mara wants the tools to remain in use. Eli wants to preserve the room. The current cut follows sorting, negotiation, and disagreement. It neither teaches viewers to perform repairs nor resolves the siblings' relationship.

The team has two plausible audience hypotheses:

A: skilled work. People interested in making and repairing things may choose the film for its attention to tools, working knowledge, and what happens to a workshop when its owner is gone.

B: inherited family obligations. People drawn to stories about what families leave each other may choose it for the negotiation between siblings who attach different meanings to the same possessions.

Neither hypothesis is absurd. Both refer to material in the film. But each can slide into a different promise. A becomes “watch the craft of restoring treasured objects.” B becomes “watch a family repair its broken relationship.” The first implies practical processes the cut does not show. The second implies an outcome it does not deliver.

Those are conflicts between expectations and material, not necessarily conflicts between kinds of people. A repair enthusiast might value unresolved family observation. Someone interested in inheritance might also watch restoration videos. Do not mistake labels that help you find possible participants for mutually exclusive reasons to watch.

Write each hypothesis as a conditional claim: this person might choose this film in this situation because of this experience. Then name the part of that claim you are least certain about. Does an interest in tools extend to watching arguments over who gets to keep them? Does an interest in family stories depend on reconciliation? Those questions can distinguish the hypotheses more usefully than “Which audience loves the idea?”

A historical example shows why theme lists are starting points rather than settled findings. Sundance's 306 Hollywood case study describes the filmmakers beginning with groups connected to the film's themes and adjusting outreach as information arrived.[^case] It does not report a controlled, same-description comparison of competing audience expectations. The exercise below is an original editorial proposal, not a method demonstrated by that release.

Establish what is not allowed to change

Before seeking reactions, record the project's present facts. Keep aspirations separate from material that exists. For The Last Repair, the working constraints are:

This is not a demand that the film never change. It defines what you are comparing now. If the team decides to add a substantial process sequence or pursue another ending, revise the project record and acknowledge that you are evaluating a changed proposition.

The distinction also applies to the material shown. Suppose one group receives a montage of tools accompanied by brisk music and the other receives a scene of the siblings arguing. Different reactions could concern the different samples, not simply the different audience interests. You have changed two things at once.

You may legitimately want to compare those messages. But label that as a message comparison and record what each version reveals, conceals, or implies. Do not report its result as though two groups encountered an unchanged film invitation. Neither version may imply a repair tutorial or family resolution that the film cannot sustain.

For a first comparison of expectations, use one shared description and the same surrounding context. “Neutral” here means avoiding a preferred interpretation in the wording; it does not mean that any description is free of choices or influence.

Give both hypotheses the same description

Here is the original description card for the invented film:

The Last Repair — an observational feature documentary

Two siblings spend the final weeks before their late father's workshop closes deciding what to keep, give away, and put back into use. As they sort its contents, their conversations reveal different memories of the man and different ideas about what they owe the place. The film stays with their practical decisions and disagreements inside the workshop.

The card supplies actions, form, and subject without calling the experience inspiring, instructional, healing, or universally relatable. It is not a perfect final sales paragraph. Its purpose is to let you examine what people infer from the same bounded account.

Use the same proposed viewing context as well. For example, ask participants to consider a feature-length film watched in one sitting, not a short repair clip for A and a post-screening family discussion for B. State that this is a proposed context. Interest conditional on an event you cannot offer should remain conditional.

At the project stage described here, an expectations conversation is different from a rough-cut screening. It tells you what someone imagines from the invitation. It cannot establish how the finished film feels. If you use a sample instead, identify that sample and its limitations; if you show a cut, separate the response to it from the initial expectations.

Begin with the same core questions for everyone:

• What would you expect to spend most of this film watching?
• Which part of that description, if any, gives you a reason to choose it?
• What would need to be present for that reason to hold? What would make the film a different proposition from the one you imagined?

These are proposed prompts, not a validated questionnaire. “Nothing here interests me” is an available answer. Do not make participants manufacture a reason because they agreed to help.

The GOV.UK Service Manual advises using open, neutral interview questions and follow-ups that clarify specific accounts, within a planned but flexible discussion guide.[^interviews] That guidance concerns service research. Its relevance here is the discipline of asking and listening, not proof that this film exercise measures demand.

After the initial answer, use follow-ups appropriate to the expectation the person actually describes. For A, ask what “watching skilled work” means in this instance. Would they expect demonstrations, sustained observation of someone working, discussion of tools, or decisions about a workshop's future? Invite them to describe the distinction before offering those possibilities yourself.

For B, ask what they expect to happen between the siblings and how important that outcome is to their interest. An expectation of reconciliation and a desire to observe an unresolved negotiation are not the same reason to watch. Do not ask “Wouldn't the unresolved ending be more honest?” That supplies the filmmaker's preferred judgment.

Both lines of questioning remain available to any participant. Following someone's account is different from telling a person recruited through a repair group that they must speak for “the repair audience.”

State what would count against each hypothesis

Do this before reading the reactions. Otherwise, any response can be recruited as confirmation: praise proves interest, confusion proves curiosity, and rejection proves that the work is challenging.

For A, the central risk is dependence on repair instruction or visible restoration. If that is what makes the proposition attractive, and it is absent, the broad craft label has concealed a mismatch. Narrowing A to an interest in the value and inheritance of working tools might remain plausible, but that narrower claim needs its own support.

For B, the risk is dependence on a relationship being resolved. Someone may want a family story specifically because they expect estranged people to come together. Interest under that expectation does not establish interest in the actual unresolved film.

The same logic applies to context. A person might consider the film only as part of a facilitated discussion. That condition matters if your proposed invitation is to an unaccompanied viewing. Do not remove the condition when reporting their interest.

To explore these dependencies in the fictional exercise, introduce the same clarification after the initial questions:

The current cut does not include repair demonstrations, and it does not resolve the siblings' disagreement. It follows the sorting and negotiation themselves.

Then ask: “What, if anything, does that change about the experience you expected or your reason to watch?”

Record the initial response and the response to clarification separately. They were made with different information. A revised answer does not erase the first expectation; the difference may identify language in the pitch that needs work. Equally, do not describe the clarification as independent confirmation. It is an explicit intervention in the conversation.

Ask for permission before recording or retaining personal accounts, explain the purpose, and allow questions to be skipped. Family obligations can be discussed without requiring someone to disclose a death or dispute in their own life. Recruit for a relevant perspective, not for access to a painful story.

Compare the conditions attached to interest

Keep the reason and the mismatch together. A page of positive adjectives strips away exactly the information the comparison needs.

Use a record with these fields for each actual conversation: how the participant was reached; prior knowledge of the project; the version of the description and any other material shown; expected scenes; reason to watch; dependencies revealed by follow-up; changes after clarification; and what the account does not settle. Record overlapping interests rather than assigning everyone to only one box.

For the invented film, the following comparison sheet contains hypothetical branches—not interview results:

The last branch is easy to mishandle. A friend can sincerely want the film to succeed without being interested for either reason under comparison. Preserve that support for what it is. Do not relabel it as evidence about the experience promised to strangers.

Look at contradictions within a hypothesis as well as differences between hypotheses. If two people reached through the same group imagine radically different films, the group name has not resolved the audience question. It may be hiding the distinction you need.

Keep recruitment and prior familiarity visible when accounts differ. If one set of participants already knows the filmmaker and the other sees the project cold, the difference is not cleanly attributable to subject interest. You can retain the observations while withholding that explanation. A second round with more comparable conditions may be useful; it should not be described as a statistical correction merely because you repeat the exercise.

Do not manufacture a market from the comparison

An accessible group of respondents is not automatically representative of everyone with the same interests. Statistics Canada describes the selection and coverage problems in non-probability sampling and the assumptions required to infer population characteristics from it.[^sampling] For the exploratory conversations proposed here, keep conclusions attached to the people, material, and circumstances actually encountered. Do not publish a margin of error, population estimate, or claim of statistical confidence.

The problem is not solved by adding more enthusiastic acquaintances. Nor does splitting a small informal sample into two named groups make it a comparative market study. This article supplies no sample-size rule or validated prediction model. A study intended to support broader inference needs suitable research expertise and a design for that purpose.

Overlap creates a separate error. A repair-group member can also be interested in family inheritance. Adding the sizes of two communities would count shared people twice unless overlap were established and handled; even a deduplicated membership count would still not be a count of people who want this film. Membership, exposure, interest, and attendance are different claims. None follows automatically from the previous one.

You can nevertheless learn something valuable without estimating a market. An account that identifies an absent repair demonstration gives you a concrete expectation to examine. You do not need to know its population frequency to remove a promise you already know is false. You do need stronger evidence before declaring that this expectation defines an entire audience.

That distinction keeps a modest exercise useful without enlarging its authority.

Make the pitch decision the evidence allows

Suppose actual future conversations produced one of the patterns below. Each outcome would change the pitch differently; none is a result of research conducted for this article.

Prioritize a narrower, supportable invitation. If interest connected to family negotiation survives the factual clarification while craft interest repeatedly depends on absent demonstrations, you might lead with inherited obligations. Keep the workshop specific, but stop implying instruction. The decision concerns the strongest supported reason in the observed material, not the intrinsic value of one audience or the size of its market.

Retain two compatible entry points. If people interested in working objects and people interested in family obligations both describe an interest in the same sorting and negotiation, the apparent rivalry may dissolve. One invitation can begin with what happens to a person's tools; another with what siblings owe a parent's memory. Both must lead to the same film. Preserve the unresolved relationship and observational form in each.

Acknowledge a development decision. If neither reason survives an accurate description, inspect what the team itself wants the project to be. You may decide the film needs a different structure or more attention to a process already in the material. That is creative development. Do not quietly promise the change in the deck before making or establishing it, and do not assume the film must be remade to satisfy these particular respondents.

Leave the comparison unresolved. If accounts are mixed, material differs, or prior familiarity explains too much of the apparent contrast, both hypotheses may remain open. You can still remove inaccurate claims now. Being unable to choose a primary invitation does not require retaining false ones.

Record the decision with the evidence that could reopen it. For The Last Repair, a useful pre-research note would be:

Current boundary: Neither repair instruction nor reconciliation may be promised. Question to resolve: Does interest in the workshop's objects survive without demonstrations, and does interest in family obligation survive without closure? Permitted change: Emphasize a supported reason to watch the existing sorting and negotiation. Still unknown: Which reason is better supported, how widely it occurs, and how viewers experience the cut.

After real conversations, replace the unanswered comparison with the actual bounded finding and retain the material behind it. Do not replace “still unknown” with a confident audience label simply because the pitch needs a heading.

The film can have more than one way in. It cannot honestly become a different film each time someone opens the deck.

<!-- ARTICLE BODY END -->

Sources

[^case]: Jess Fuselier, “Distribution Case Study: 306 Hollywood”, Sundance Institute, “Identifying the Audience.” Historical 2018–2019 context; inspected September 19, 2026. A release account, not a same-stimulus audience experiment. [^interviews]: GOV.UK Service Manual, “Using in-depth interviews”, published February 21, 2017. See “Design the interview” and “Do the interview.” Inspected September 19, 2026. Service-research guidance; transferred here only as bounded interviewing guidance. [^sampling]: Statistics Canada, “3.2.3 Non-probability sampling”, Statistics: Power from Data!, modified September 2, 2021. See the introduction, convenience/volunteer sampling, and disadvantages. Inspected September 19, 2026. General sampling limitations, not evidence about a film audience.

Frequently asked questions

What should be compared when a pitch has two plausible audience hypotheses?

Compare the experiences the hypotheses promise, not which group seems larger or more enthusiastic. Put the same accurate film description in front of the comparison and ask what people expect to watch, why that experience interests them, and which expectations the actual project cannot meet.

When does a hypothesis stop supporting the film you are making?

When it works only if the pitch implies scenes or outcomes you do not have, such as repair demonstrations or a family reconciliation. In that case you need to narrow the invitation, acknowledge a creative decision, or change the project explicitly; better targeting cannot supply the missing film.

How should an audience hypothesis be phrased before you seek reactions?

Write it as a conditional claim: this person might choose this film in this situation because of this experience. Then name the part of the claim you are least certain about, such as whether an interest in tools extends to arguments over who keeps them.

What must be fixed before comparing reactions to the two hypotheses?

Record the project's present facts and keep aspirations separate from existing material. Use one shared description and the same surrounding context; if you show different samples or messages, label that as a message comparison rather than reporting it as one unchanged film invitation.

Can informal conversations with accessible groups estimate a market for the film?

No. An accessible group is not automatically representative, and adding more enthusiastic acquaintances or splitting a small informal sample into two named groups does not make it a comparative market study. Keep conclusions attached to the people, material, and circumstances actually encountered; you can remove a false promise without estimating its population frequency.

Further reading

• The Deck Has Become More Conventional Than the Film — If the comparison pushes the team toward changing what the film resolves, read this first: it separates clarifying an existing ending from promising a new one.