Build a Film Press Kit Without Recycling the Development Pitch

The film is finished. A festival, a programmer, or an outlet has asked for press materials, and the most presentable document on your drive is the deck that raised the money. It has a good cover, tidy type, a synopsis, a cast slide, and a spread of images that feel like the film.

Several of those things are no longer true, and the cover hides it.

A press kit is not a pitch deck with the asks removed. It's a different document for a different reader. The deck argued for a film that didn't exist yet, in front of someone who might help make it. The kit describes the film that exists, in front of someone who has to present it to other people: a programmer writing a program note, a journalist filing eight hundred words, a publicist forwarding a link upward. Those readers need facts they can repeat without checking with you first. Every promise the finished film doesn't keep becomes a liability in their hands.

So the work isn't redesigning the cover. It's reading the old deck against the finished film and sorting what's still true from what was only ever a hope.

Name the recipient and the presentation task

Before you open the deck, write down what was actually asked for. That sounds obvious and it's the step people skip, because the deck is right there and the request is in an email three screens up.

The requests fall into two kinds, and they are not interchangeable.

Background material helps someone understand the film well enough to write or talk about it. A journalist working on a feature may need a production history, a director interview, technical detail, and the ending. Hiding the ending from a reporter who is about to write about the film is not protecting the film; it's making their job harder and guaranteeing an error.

Publication-ready selections are things that get printed or posted as they arrive, with your name on them. A program-guide synopsis, a credits list, stills with captions. Nobody edits these before they go out. If they're wrong, they stay wrong in public.

Most kits contain both, and the single most useful thing you can do is keep them in separate files with obvious names. A PDF that contains a program synopsis and an unmarked assessor synopsis full of plot resolution is a document that will eventually be forwarded to the wrong person. It always is.

Here's the one public source behind the EPK specifics in this piece: The Guerrilla Rep's guide to what goes in an electronic press kit, recorded in our notes as checked on 8 September 2026, names a synopsis, a trailer or clips, a full bio, a photo gallery or downloadable zip file, and a line about how people should reach the kit. That's one practitioner's list, the post carries a commercial offer at the end, and the displayed page doesn't show a publication year. It's a useful reminder of what recipients tend to want. It is not a requirements document, and you should not paste it into your process as a checklist. The recipient's own request is the specification.

That matters because the same fact needs different precision in different places. Your film's runtime might be 16 minutes 48 seconds. A printed program listing will run "17 min," because that's the space it has. A distributor's spec sheet wants 16:48. Both are correct; neither works everywhere. A kit that gives one number for every recipient is a kit that hasn't thought about who's reading.

Audit the development deck for expired promises

The deck is not a lie. It's expired. Everything in it was true or reasonably hoped-for on the day it was made, addressed to someone deciding whether to help. That's the generous reading, and it's also the accurate one — which makes the audit easier, because you're not hunting for dishonesty. You're checking dates.

Read the deck slide by slide against two things: the finished film and your current production records. Give every item one of three outcomes.

• Replace it with confirmed material.
• Remove it.
• Retain after checking.

That third one is the outcome people skip, and it's where you recover good material instead of starting from a blank page.

Here's an invented example, built so the pieces can be checked against each other. Ferry Weather is a fictional 16-minute-48-second narrative short, written and directed by Dana Reyes, produced by Priya Nandakumar. The deck in question is two years old.

Two things in that table are worth sitting with.

The technical line is one slide doing two jobs. "On location at a working harbor" survived the whole production and belongs in the kit. "Shot on 16mm" was a plan that changed. Same sentence, different outcomes, and the only way to tell them apart is to check the film. This is what expired material looks like: not wrong when written, wrong now.

The cast slide is a different species of problem. A wish-list name on a deck signals viability — this actor is interested, so this film is real enough to fund. A name in a credits list states a fact. On a slide, both look identical: a headshot and a name. But if the kit carries "Colm Devereux" because it was on the deck, you have published a false credit about a person who did not work on your film, and you have insulted the actor who did.

The look-book spread is the trap that catches people who think visually. It's the most beautiful page in the deck, and it has to go. Those frames belong to another film, and even if the finished film looks exactly like them, you have no right to distribute them as promotion for yours — and printing them tells a journalist something false about what they're about to watch. A mood board is a working document. It is not an asset.

And the outreach plan is a tense problem. "Twelve coastal towns" was a proposal. A press kit can only report. If those screenings happened, restate them in the past tense with numbers you can actually support. If they didn't, the plan comes out.

Rebuild the factual core first, then decide what design survives. The typography doesn't make a claim, so it can stay. The title treatment should match the film's actual title card, and the facts should be checked before anything gets a new coat of paint — otherwise you're polishing sentences you're about to delete.

Write for someone presenting the film accurately

A synopsis is not a summary. It's a performance: a short one that makes someone want to watch, written under a length limit, for a specific page. The length limit is usually a layout constraint rather than a suggestion, so read it as one.

The first decision is the spoiler boundary, and you make it deliberately, per audience.

Ferry Weather turns on a small thing: Vera has a bag behind the counter and a ticket for the 4:40 crossing, and she doesn't get on the boat. The program-guide synopsis must not say that. The assessor synopsis, written for a programmer deciding whether to watch, probably should, because it's the moment that shows what the film is doing.

The public version, roughly forty-eight words:

Vera has run the ticket booth at a small harbor ferry for eleven years. On the last day of the service she works a shift that is mostly waiting, while the crew she has known for a decade prepares a crossing that will not repeat. A 17-minute drama.

A longer public version can add texture without adding the turn — the low cloud, the three tickets sold before noon, the lines Ilyas keeps coiling. The bag under the counter can appear; a detail that pays off is fair. What can't appear is the sentence that tells the reader she stays, because then there's no film left to watch.

The assessor version is where you spend the ending. Name what happens, name the last shot, and say why it's the right ending for this film. A programmer reading forty short synopses in an afternoon is not looking for mystery; they're looking for evidence that the film knows what it's doing.

Then keep them apart. Two files, labeled, with the public one named as public and the assessor one named as background. Do not put them in the same PDF and do not trust that whoever receives it will read the filenames carefully.

One more rule, and it's the one that keeps the audit from undoing itself: describe the film's actual form and premise. If it's a 17-minute drama, don't carry over language from a deck that called it "a hybrid documentary experiment" while you were still deciding what it was. Early descriptions are aspirational. A synopsis is descriptive.

Credits come from the film's own credit roll, in its order and its wording, because the roll is the document of record. If someone's title changed between the deck and the deliverable, the deliverable wins. And keep the director's statement distinct from neutral production facts: a statement is where intent lives, in first person, and it can be quiet and unglamorous. What it can't do is smuggle in a production fact the credits contradict, or promise what the film will do to an audience.

Treat stills as supplied assets, not atmosphere

Everything in the previous section has a visual counterpart, and it fails in the same way.

For Ferry Weather, the usable images are three frame grabs from the finished master — Vera at the booth window; Ilyas on the foredeck coiling a line; the empty dock after 4:40 — plus one on-set photograph by Bea Kowalczyk. The festival asked for two to four stills. Four is a good answer.

Every one of those needs five things attached to it, and the fifth is the one people forget.

• Filename, matching the file in the folder exactly.
• Caption identifying who and what is visible.
• Credit line — "Photo: Bea Kowalczyk" for the on-set image; a frame from the film carries the film's own credit convention.
• Source — frame from the master, or on-set photograph. These are different things and recipients may treat them differently.
• Authorized use, as far as you've actually agreed it. If the photographer's agreement requires a credit line, that's not a suggestion.

A gallery of pretty images with no captions leaves the recipient guessing, and guessing produces errors with your film's name on them. Worse, it produces the specific error that's already lurking: a caption reading "Colm Devereux and Yusuf Adeyemi on set," carried over from the deck's cast slide, sitting in the same packet as a credits list that only names Adeyemi. Two documents in one delivery, contradicting each other. The only way to catch that is to read the text and the asset inventory against each other, name by name.

Then check the inventory against the actual folder. In this case the earlier draft of the asset list had five entries; one of them was the look-book frame, now removed; the list wasn't updated, so it names a file that doesn't exist and the numbering has a gap. Small, embarrassing, and completely avoidable.

Watch for the resolution mismatch too. "High-resolution" is a real requirement, and an upscaled phone photo of a monitor is not a high-resolution still. Frame grabs pulled from the master usually are.

And keep the categories straight: an actual film still, a development reference, and an image made only for the pitch are three different objects. A location scout photograph is not a still. A concept frame is not a still. Neither becomes one because it's the best-looking thing in the folder.

Assemble one current packet with a reachable owner

Now build the thing. Here's an invented request, of the shape a real festival or outlet might send:

For each accepted short: a 50-word and a 150-word synopsis for the printed program and website; final runtime to the second; exhibition format and aspect ratio; a full credits list; a director's statement of up to 200 words; two to four high-resolution stills with captions and credit lines; key art if available; and a press contact who can respond within two business days.

And here's the packet that answers it, item by item, with the reason each one belongs.

The two synopses go to the program and the website, which means the spoiler boundary applies to both, and both are written from the finished film rather than adapted from the deck. Source: you, watching the finished cut.

Runtime: 16 min 48 s. That's the master. A program listing may round it. A spec sheet gets the seconds.

Exhibition format: DCP; 2.39:1, color, English dialogue, stereo. The request asked for exhibition format and aspect ratio, so the line names the thing that gets projected, not the medium it was made on. "Digital" answers neither question. If the DCP doesn't exist yet, say what does — the ProRes 422 HQ master, with its codec, aspect ratio and audio — and when the DCP follows. Acquisition belongs on its own line: not "16mm," not "shot on film."

Credits, copied from the roll: written and directed by Dana Reyes; produced by Priya Nandakumar; director of photography Tomas Brandt; editor June Okafor; original score Ruth Lindqvist; sound design and mix Ade Salako; cast Marisol Enriquez, Peter Nwosu, Yusuf Adeyemi.

A director's statement under 200 words, first person, about what was actually made.

Four stills with filenames, captions, credit lines, sources, and any use conditions the production has agreed to.

No key art. The request said "if available." The film doesn't have a poster, so the answer is silence, not a mock-up. Conditional requests are answered by omission.

One press contact: the producer, at an address that reaches a human, plus the promise to respond within two business days — but only if the producer has actually agreed to that. A response-time promise nobody owns is the same class of error as an unbuilt outreach plan.

A date and a version line, stating that this packet supersedes the development deck and any earlier draft. Date it. Undated press kits circulate for years.

Two more checks close it out.

The public material and the background material should be visibly separate. The assessor synopsis that names the ending belongs in a background file. The program synopsis doesn't. If a journalist asks for the version with the ending, send it — deliberately, to them, as background — rather than keeping one file that serves both jobs badly.

And if the whole thing lives behind a link, describe what the link does rather than how safe it is. "Anyone with this URL can view and download these files; the link can be rotated if it travels further than intended" is accurate and useful. "Secure access" is a claim you can't support, and a recipient who believes it may treat the folder as private when it isn't.

Decide who owns corrections, too. If a caption turns out to be wrong after three programmers have the packet, one person should be able to say what the correct version is, and how recipients hear about it.

Keep the deck. It's the record of an argument you won, and there's no reason to delete it. Just don't send it.

The deck was about a film that could exist. The kit is about a film that does. Somewhere in the second document, a programmer will find a sentence they can print without checking with you first, and that's the whole job — a press kit succeeds by being accurate enough to disappear into someone else's work.

Frequently asked questions

Why is a development pitch deck not a press kit?

The deck argued for a film that did not exist yet, in front of someone who might help make it. A press kit describes the film that exists, for a programmer, journalist, or publicist who needs facts they can repeat without checking first. Expired promises from the deck become liabilities in the kit.

How should a development deck be audited for a press kit?

Read it slide by slide against the finished film and current production records, giving each item one outcome: replace with confirmed material, remove, or retain after checking. In the example, a technical line splits—one half survives, one is false—an old synopsis ending was cut, an attached actor never joined, a look-book spread has no publicity rights, and an outreach plan was proposed rather than reported.

What is the difference between background material and publication-ready selections?

Background material helps someone understand the film well enough to write or talk about it, and may include the ending. Publication-ready selections are printed or posted as they arrive, with nobody editing them before they go out. Keep them in separate files with obvious names; a program synopsis and an assessor synopsis full of plot resolution should not share one PDF.

How should stills be prepared and checked?

Each still needs a filename matching the folder, a caption identifying who and what is visible, a credit line, a source stating whether it is a frame from the master or an on-set photograph, and any authorized-use condition. Check the text and asset inventory against each other name by name, and check the inventory against the actual folder. A look-book frame, concept frame, or location scout photograph is not a film still, and an upscaled phone photo is not a high-resolution still.

How should conditional requests, contact details, and versioning be handled?

If key art is requested only if available and no poster exists, answer by omission rather than a mock-up. Name a press contact who can respond and promise a response time only if that person has agreed to it. Date the packet and state that it supersedes the development deck and earlier drafts. If the kit lives behind a link, describe what the link does, not that it is secure.

Further reading

• What Must You Deliver With a Finished Film? — Delivery requests and press packets overlap: this sorts a written list by what each item enables, which helps when the same stills and synopsis go to a distributor.
• Introduce a Film to a Sales Agent Without Sending Only the Creative Deck — A sales agent is another recipient with different needs; this shows why a creative deck alone is not enough and what factual layer to add.