The plays

Start with something specific. Not “arthouse audiences” or “fans of the genre” but: women aged 30 to 50 who follow the director’s previous work, coming in pairs in the first ten days, hearing about it from a friend rather than an advert.

Then name the single thing the release depends on, in a form you can check. “This works if the Ukrainian community in four cities turns out in week one.” “This works if the ending gets argued about rather than admired.” Then name what would show you, early, whether it is happening: community organisations engaging and group bookings for the first, argument in the comments rather than praise for the second.

Three or four indicators are enough, chosen because they speak to that assumption. Decide now what a good and a bad reading of each looks like. It prevents both common failures: reacting to whichever signal is loudest in week two, and missing the one that mattered because nobody had said what to watch.

Most comping collapses into a single number, what did the similar film do. Separate it out:

• Positioning comp. How to frame the film. Can reach back years.
• Commercial comp. What it might realistically do. Only useful if recent; the market has moved.
• Mechanism comp. How did a comparable title actually travel? A debate, a fandom, an event, a community, a piece of casting?
• Audience comp. Which specific group turned out, and can you reach them directly?

The last two are where the work is. Reading the conversation around a comparable title tells you more about how to run your campaign than its opening weekend ever will: what was said on Letterboxd, which subreddits picked it up, which clips travelled and why. It costs nothing but attention.

Diaspora organisations, subject-matter groups, fan communities, local societies. The conventional move is to approach them once the release is booked; the more useful move is to approach them before. And ask for a booking, not a view, “would twenty of your members come on a Saturday?” gets a truthful answer because it involves a commitment, where “does this look interesting?” gets politeness.

One distributor saw a Brazilian community turn out in numbers at a Swiss festival weeks ahead of release. Another built a release around Georgian audiences in Germany. In every case the community existed long before the film did. Nothing prevented anyone asking them first.

Audiences on each platform expect something different and recognise a repurposed asset instantly. One distributor now holds a standing budget line purely for short-form built for the platform rather than adapted for it, and treats it as a fixed cost of release rather than an optional extra.

One documentary distributor launched artwork for a film about two well-known figures and discovered in the comments that the two fan bases actively disliked each other. The assumption of a single combined audience was wrong. Because they read the comments closely and early, they repositioned toward two audiences instead of one.

Most things get fixed early out of habit rather than necessity: positioning signed off in one meeting, creative produced in a single batch, media booked in one flight, the team moved to the next title at week six. Very little of that is forced by the release calendar. Before each of those, ask whether the decision genuinely has to be made now.

Then decide what would change your mind. Agree in advance the signal, the threshold or the date that would make you spend more, spend less or shift the message, because once a campaign is live, nobody has time to reopen the strategy.

Release platforms are often built when showtimes are ready, a few weeks out, which wastes the entire period in which interest is forming. A platform live from announcement collects audience data, shows where traffic is coming from, and gives you somewhere to engage people who are already looking. It also has time to build search visibility, so that when someone asks where a film is playing, or asks an assistant whether it is worth seeing, the answer points somewhere you control.

How it is built matters as much as when. Across 2025, purpose-built campaign platforms sent 16.2% of visitors through to a cinema. Standard release pages carrying the same information sent 4.2%. The difference is structure: a clear path to showtimes, fewer competing calls to action at the moment somebody is deciding, and a page that changes when tickets go on sale.

Someone lands on your platform, looks up showtimes, picks a city, then clicks through to a ticketing page. Those steps are the visible half of the journey to a sale, and most distributors never set them up to be counted. Do it before launch, watch them daily through the release window, and keep the numbers afterwards. For context, across 2025 a standard release page moved about one visitor in twenty-four through to a cinema, and a purpose-built campaign platform about one in six. After three or four releases you have something better than either figure: your own benchmark for what normal looks like on your slate.

Every link, every post, every partner placement, properly tagged. It is unglamorous and widely skipped, and it is the difference between counting the people who came close to booking and knowing which of your work brought them there.

Sentiment tells you how people feel. Requests tell you what they intend. “Why isn’t this playing in Sheffield” is not a complaint, it is a person telling you they would go.

One distributor collected exactly those comments and took them to exhibitors as proof of regional demand. They won additional screenings. Thirty people naming the same city is an argument exhibitors recognise.

Not two creative routes but as many as you can produce, live at the same time, with a standing commitment to shift budget toward whatever is working every few days. Pre-launch testing tells you what people said they liked. Live variants tell you what actually moves someone toward a showtime. Most teams run some version of this. The discipline is not the testing, it is acting on the result fast enough for it to matter.

A distributor releasing a Turkish-Georgian film found the Georgian community in Germany responding far more strongly than modelled, and moved budget toward them mid-campaign. The signal was visible. Acting on it required having something left to move.

One distributor uses social listening to find the misconception forming around a title, in one case, confusion over whether the film was a comedy, and addresses it directly in community management. Fixing a misread is cheaper than building fresh awareness.

One distributor sent teams into their ten largest cinemas and surveyed six hundred attendees on what brought them in. Social dominated; billboards barely registered. Close to the cheapest research available, and rarely done.

Compare what looked promising early against what actually happened. Across a few releases this builds the thing no external research can supply: a calibrated sense of which signals your team can trust.