Plenty of people start watching shorts with the sound off. Others are in a noisy place, or simply read faster than they listen. Captions catch all of them. Done well, they also add rhythm and emphasis to the edit. Done badly, they're a distraction that makes your video feel cheap.
Burned-in vs platform captions
Most platforms can auto-generate captions. They're fine as a fallback, but you don't control the style, timing or accuracy, and viewers can turn them off. Burned-in captions, rendered into the video during editing, look the same everywhere and become part of your brand. For most shortform content, burned-in captions are the standard. Uploading a caption file as well helps accessibility.
Style: readable first
- Size. Big enough to read on a small phone without squinting, small enough not to cover the face.
- Font. A clean, bold sans-serif is the safe choice. Decorative fonts are hard to read at speed.
- Contrast. White text with a dark outline or shadow, or text on a solid box, stays readable on any background.
- Highlight colour. One accent colour for emphasis words. Not five.
- Consistency. The same style across every video helps people recognise your content in the feed.
Timing and chunking
How you break the text matters as much as how it looks.
Short chunks
Two to five words per caption is typical for fast-paced shorts. Long sentences on screen force people to read instead of watch.
Break on natural phrases
"I stopped posting / every day" reads better than "I stopped / posting every day". Split where you'd pause when speaking.
Sync to speech
Captions should appear as words are spoken, not before and not trailing behind. Even small delays are noticeable and irritating.
Word-by-word highlighting
The popular style where each word lights up as it's spoken can help with fast talkers. It can also be tiring. Use it if it fits your tone, skip it if your content is calmer.
Placement and safe zones
Every platform overlays its own interface. Usernames, post captions and buttons sit along the bottom and right side. Keep captions in the middle of the frame, often just below the speaker's face. Check them on a real phone in the actual app, not only in the editing software.
Accuracy matters more than you think
Auto-transcription gets close, then mangles names, brand terms and jargon. A typo in a caption is the first thing commenters notice. Every caption should be proofread, ideally against a list of your common terms: product names, people, technical words. Keep that list in your style guide so your editor gets it right every time.
Emphasis without clutter
Captions can carry emphasis: a key word in your accent colour, a slightly larger size for a punchline, an emoji where it adds meaning. The trick is restraint. If everything is highlighted, nothing is. A good rule is one emphasised word or phrase per caption at most, and often fewer.
Captions and the hook
The first caption is part of your hook. If the opening line is "So yeah, um, I wanted to talk about...", the caption makes that weak start even more obvious. Edit the opening so the first words on screen are interesting. Some editors add a separate headline at the top of the frame for the first few seconds to frame the topic.
Captions for different content types
- Talking-head: full captions, styled, synced word by word or phrase by phrase.
- Tutorials: captions plus step labels so people can follow along.
- Music-led or visual content: fewer captions, maybe just a headline and a few key lines.
- Ads: captions that work as a standalone script on mute. See our ad creative guide.
Accessibility
Captions aren't only an engagement trick. They make your content usable for people who are deaf or hard of hearing. Clear, accurate captions with good contrast are the basics. Adding a caption file on upload, where the platform supports it, helps even more.
Quick caption checklist
- Readable on a small phone?
- Inside the safe zone on each platform?
- Synced to speech?
- Proofread, especially names and terms?
- Consistent with your brand style?
- Not covering faces or key visuals?
Building a caption style guide
Write down your caption rules once: font, size, colours, outline, position, how many words per line, when to use emphasis and emojis, and a list of names and terms with correct spelling. Share it with every editor who touches your content. It's a single page, and it keeps hundreds of videos consistent.
Review it every few months. If a style isn't working, change the guide rather than giving the same note in every revision.
Good captions are invisible in the best way: people read them without noticing the effort behind them.
Related guides
Part of our shortform editing series. Read shortform hooks, Reels editing and Shorts editing next. See our work or get in touch.