Subtitles and accessibility: the audience you are missing

Captions are how deaf and hard-of-hearing viewers, second-language viewers and anyone watching with the sound off follow your video. The text that makes a clip usable for a deaf viewer is the same text that makes it watchable on a silent train, so the accessible version is also the one that reaches the most people.

That is why filing captions under compliance undersells them. You make one version of the video, and it works for more of your audience.

Who captions reach

Three groups rely on captions.

People with hearing loss

The World Health Organization says 430 million people, over 5% of the world's population, have disabling hearing loss (WHO fact sheet). For many of them a video without captions is not a worse experience. They cannot follow it at all.

Non-native speakers

Plenty of viewers read a second language more easily than they follow it by ear at speed. Captions let them keep up with fast talkers, accents and unfamiliar terms. Making the words visible opens your video to people who would otherwise scroll past.

People watching with the sound off

People watch in public, at work, in bed and on commutes where sound is not an option. For them, captions are the only way the message gets through.

The growth case inside the access case

Each of those viewers watches longer, follows and shares when they can understand you. You do not have to choose between accessibility and engagement: the same captions serve both.

The accessible version of your video is the one that reaches the most people.

Captioning a video does not narrow it for a special audience. It opens it to everyone a soundtrack would have shut out.

Captions and transcripts are different things

People often confuse the two:

• Captions: text that appears in sync with the video, timed to the speaker. They serve viewers watching in real time.
• Transcripts: the full text of the video, available separately. They serve people who want to read, skim or search the content, and they help search engines understand what the video is about.

Short-form video needs captions. Long-form content benefits from both. A workflow that gives you accurate, timed captions also gives you the raw text for a transcript: an .srt file is a transcript with timestamps.

What usable captions look like

Captions only help if people can read them:

• Accurate: wrong words are worse than no words. Auto-transcription gets you most of the way. Read through once and fix names, jargon and anything the model misheard.
• Well-timed: text that lags or races ahead is hard to follow. Word-level timing keeps it on the voice.
• Legible: strong contrast and a clear font help everyone, including viewers with low vision.
• Clear of the interface: text hidden behind platform buttons helps no one.

The styling guide covers legibility in detail.

Make captions the default

Captioning every upload is a small, fixed step, not extra work for a small group.

Start with your next video. Run it through the free caption generator, read the transcript once, and export a version anyone can watch.