How to Choose Fonts and Build a Readable Film Pitch Deck
How to Choose Fonts and Build a Readable Film Pitch Deck
The title page works. "The Last Crossing" is set in something with weather in it, and it does exactly what you wanted it to do. Then you turn to the synopsis, and the same face that made four words feel like a film turns ninety words into a slog. Somewhere near the bottom of the page, a line in pale grey says the location isn't confirmed. It's the most consequential sentence on the spread and you have to hunt for it.
Neither of those is really a font problem. They're the same problem: the type has no jobs.
So the order that works is this. Collect the text your deck actually has to carry, at the length it actually runs. Decide what a reader meets first and what has to stay findable later. Assign size, weight, spacing and position to those roles before you get attached to anything. Then test the arrangement on the page that's hardest to read. Choosing the font comes third or fourth, not first.
Assemble the text that will stress the system
A font menu shows you an alphabet. Your deck has paragraphs, and paragraphs are where faces fall apart.
Before you style anything, write out the five kinds of text that do most of the work in a film deck:
- The title
- A synopsis passage, at full length
- A character name
- A reference caption tied to an image
- A factual qualification
Those five aren't the only things in a deck, but between them they stress everything: sustained reading, an entry point, an identifier, small type sitting on a photograph, and a limit on what you're claiming.
Use the real copy. Placeholder text hides every interesting failure — the line length where a paragraph starts to feel like a wall, the long surname that won't fit a column, the numbers, the parentheses, the em dash. If your synopsis is going to run as one block, write the block. If your protagonist has an unusually long name, use that name, not "Jane Doe."
The qualification deserves particular care because it's the one people write in shorthand and then set in twelve-point grey. If the line reads Location proposed; access not confirmed, that's what you test. Not "TBC." The shorthand hasn't been tested for the one thing that matters: whether a reader in a hurry can tell that the location is a proposal.
And do the editorial work before the typographic work. If the synopsis is two hundred words and the page holds a hundred and twenty, the answer is not a smaller size. Something has to come out. Shrinking everything to fit is a decision about the type rather than a decision about the content, and it's the wrong one in both directions — the page gets worse to look at and harder to read. Resolve what belongs on the page first. Then style what's left.
Build roles before you choose favorites
Start with the reading text, not the title. Pick one size for the synopsis that you can read at arm's length on the screen you'll present on. Then fix two interacting settings: how long the lines run, and how much space sits between them. Those two aren't separate knobs. A long measure wants more leading; a short one can take less. Matthew Butterick's Practical Typography treats body text and its spacing and length as a system on the "Typography in ten minutes" page — worth reading, though his measurements come from print and don't transfer to a screen as requirements.
Once the synopsis reads, ask whether you need a second face at all. Plenty of decks are better with one family. If the reading face has a usable bold, a decent italic and reasonable capitals, the title may only need to be larger and better placed. A display face earns its place when it brings something the reading face can't — a texture that belongs to this particular film, a shape the title wants. Otherwise you're paying for variety you don't need.
Here's the part most people skip: hierarchy is not only size. Butterick's "Headings" page makes the case that headings work as structural signposts, and that distinct roles don't require dramatic differences in size. He's right, and the practical implication is bigger than it sounds. You also have weight, case, letter-spacing, alignment, a rule, an indent, and — most underused — space. A character name separated from the synopsis by a line of air and set at the same size as the body, in bold, will read as a name and not as a section heading. Make it large and all-caps and you've just built a heading that happens to be a person.
Case is a signal you can spend only once. If your section headings are all-caps, setting the character name in caps too puts them in competition for the same meaning. Somebody has to lose that fight, and it'll be whichever one the reader needed.
Judge with the actual characters. "The Last Crossing" and "Alexandra Bell" are different problems — one has a long run of lowercase, the other has an awkward stack of capitals and a doubled letter. A face that does beautifully with the title might set the name lumpily at the size you need. You can't know from the preview strip. You know by setting the real words.
One boundary worth keeping: the deck's title treatment is not the film's title design. The deck is a document read on a laptop, maybe projected, maybe passed around as a file. Titles in a film exist in motion, for a fixed duration, at a scale you don't control, in a theater or on a phone. They can share a feeling. They don't have to share a typeface, and neither one should dictate the other. Design them separately.
Finally, keep the set small. Five roles, maybe seven once you count a section heading and a page number. Every additional distinct treatment is one more thing the reader has to learn before they can read your synopsis.
Make the expression survive a demanding page
Now apply the roles to the worst page you've got: the longest passage over the least forgiving image.
Take a page holding all five roles over a backlit shoreline crop — the kind where the water in the lower third is bright and full of detail, exactly where small text wants to sit. Build it twice. Same words — the qualification carries its "Status:" label in both — same image, same page dimensions. Change only the type, where it sits, and whatever support sits behind it. Everything else held constant, so that whatever difference you see comes from the design of the text, not from the copy.
The first treatment is the one decks tend to arrive at on its own. The display face carries everything. The title is enormous. The character name is set nearly as large, in the display face, so it reads as a section heading. The caption sits small and widely letter-spaced over the bright water. The qualification sits in the same pale grey, one size smaller, over the same bright water, its "Status:" label as faint as the wording after it. The synopsis is set in the display face at whatever size makes it fit.
The second treatment assigns the work. The synopsis sits in the reading face at the size and measure you already tested, where it can be read as a passage rather than decoded. The title stays in the display face, large enough to be the entry point but not so large it pushes the synopsis into the remainder. The name sits in the reading face above the synopsis, separated from any section heading by position and space instead of by more contrast. The caption moves off the bright water and sits at full contrast in the reading face. The qualification sits at or near body size, in full contrast, so that both the label and the wording register — and either a scrim behind it or a move to the dark part of the frame. The image crop is unchanged; what changed is where the text sits and whether it has something to sit on.
I'm describing a comparison to build, not a result anyone has measured. Nothing here has been laid out, exported or shown to a reader. Treat the two treatments as instructions and the questions below as the test.
Can the synopsis be followed as a passage, start to finish, without your eye wandering off? Is the name distinct from the heading, and distinct from the title? And can Status: Location proposed; access not confirmed be mistaken, even for a second, for confirmed access? If that last answer is "maybe," the page isn't finished, however good the title looks.
When something fails, don't reach for the font menu first. Competing emphasis is almost always a hierarchy problem — too many things at top volume. An indistinct caption is often a placement problem, because it's sitting on a busy part of a photograph. Crowded lines are leading or measure. A qualification that vanishes is usually a color decision or a copy-organization decision. And sometimes the honest fix is to stop trying to fit the synopsis, the caption and three status lines on one page. Change the page, not the typeface.
Inspect the deck in its delivered conditions
Check the file twice: once in the application you built it in, and once in the exported version, at the size people will actually view it. A deck gets opened on a laptop, sometimes projected in a room, sometimes skimmed on a phone. Those are different conditions. Your page has to survive the hardest one you can reasonably expect, not the one on your desk.
Then look for the things that break in the handoff. Fonts substituting. Embedding failing. Line breaks moving by a word and pushing a caption onto three lines. A qualification dropping to a new line and landing on the bright water you moved it away from. Breaks like these aren't hierarchy problems, and redesigning around them will leave you with a worse page and the same defect. A file that substitutes a font on export has a handoff problem with its own diagnosis — that's a separate job from choosing the type in the first place, and worth solving as one rather than rebuilding the page around a temporary fault.
Keep your numbers attached to your specimen and your viewing conditions. There is no universal minimum size. A setting that works for a title card doesn't transfer to a synopsis paragraph, and a print measurement doesn't transfer to a screen. What you record is what you tested, on what, at what scale.
Two more things that a good-looking export does not establish. A font file sitting on your machine is not permission to redistribute it inside a PDF, and its presence is not evidence that it will embed cleanly. Check the actual terms from the font's own source. And a successful visual inspection is a look at one file on one screen — it is not an accessibility certification, and it isn't a substitute for one.
If the export flattens your hierarchy, fix the export. Don't reflatten the design to match it.
The most readable deck isn't the one with the most beautiful title. It's the one where a reader on a mediocre screen, in a hurry, in a room where nobody is paying full attention, still finds the synopsis, still knows who Alexandra Bell is, and still knows the location isn't confirmed. Get those five roles right, test them on the page that resists you most, and the font question mostly answers itself.
Frequently asked questions
What five kinds of text should a film pitch deck be designed around?
The title; a synopsis passage at full length; a character name; a reference caption tied to an image; and a factual qualification. Between them they stress sustained reading, an entry point, an identifier, small type sitting on a photograph, and a limit on what is being claimed. Use the real copy, not placeholder text, including long surnames, numbers, parentheses, em dashes, and the actual qualification wording. If the qualification reads 'Location proposed; access not confirmed,' test that, not 'TBC.'
What order should you work in when choosing fonts and building the deck?
Collect the text the deck actually has to carry, at the length it actually runs. Decide what a reader meets first and what has to stay findable later. Assign size, weight, spacing, and position to those roles before getting attached to any typeface. Then test the arrangement on the hardest page. Choosing the font comes third or fourth, not first. Do editorial work before typographic work: if the synopsis is two hundred words and the page holds one hundred twenty, something has to come out; shrinking everything to fit is the wrong decision for both the page and the reading.
How can hierarchy be built without relying only on size?
Hierarchy is not only size. It also includes weight, case, letter-spacing, alignment, a rule, an indent, and—most underused—space. The body cites Matthew Butterick’s 'Headings' page for the case that headings work as structural signposts and distinct roles do not require dramatic differences in size. A character name separated from the synopsis by a line of air, set at the same size as the body but bold, can read as a name rather than a section heading. Case is a signal you can spend only once: if section headings are all-caps, setting the character name in caps too puts them in competition for the same meaning.
What does the stress-test page compare, and what does it not prove?
It builds the same words, same image, same page dimensions twice, changing only the type, where it sits, and whatever support sits behind it. The first treatment lets the display face carry everything: an enormous title, a character name nearly as large and reading as a section heading, a small widely letter-spaced caption over bright water, a pale-grey qualification, and a synopsis in the display face at whatever size fits. The second assigns work: synopsis in the tested reading face, title as entry point, name in the reading face separated by position and space, caption off the bright water at full contrast, and qualification at or near body size with either a scrim or a move to the dark part of the frame. The body says this is a comparison to build, not a result anyone has measured; nothing has been laid out, exported, or shown to a reader.
What should you check in the delivered file and its conditions?
Check the file twice: once in the application you built it in, and once in the exported version, at the size people will actually view it. A deck may be opened on a laptop, projected in a room, or skimmed on a phone, so the page has to survive the hardest condition you can reasonably expect. Look for fonts substituting, embedding failing, line breaks moving, and a qualification dropping onto the bright water you moved it away from. Those are handoff problems with their own diagnosis; don’t redesign the page around a temporary fault. There is no universal minimum size; keep your numbers attached to your specimen and viewing conditions. A font file on your machine is not permission to redistribute it inside a PDF, and its presence is not evidence it will embed cleanly. A successful visual inspection is not an accessibility certification.