Video Accessibility: Captions, Transcripts and Social Media Tips

Video is the most common format online today, yet it's still often designed for people who can see and hear without difficulty. This guide brings together the technical practices that make a video accessible to everyone: captions, transcripts, audio description, caption contrast and fonts, plus the differences between the main social platforms.

Essential practices for accessible video

An accessible video lets anyone follow it, regardless of disability. For people with hearing disabilities, that means accurate captions and transcripts. For people with visual disabilities, it means clear audio description and sufficient contrast between text and background. The practices below cover both fronts, from the text content to how captions actually look on screen.

Captions and closed captions (CC)

Captions translate or transcribe the spoken dialogue on screen.

Closed captions (CC) also add the scene's relevant sounds: a door closing, applause, background music changing tone. For people with hearing disabilities this difference matters: a caption that only reports dialogue leaves out part of the information.

A well-made caption meets a few technical conditions:

• The text must appear in sync with the audio, with no perceptible lead or lag.

• It must stay on screen long enough to be read in full, calculated on an average reading speed rather than the speaking pace.

• It must indicate who is speaking when more than one person is in the scene.

• It must not leave gaps without captions when something relevant to understanding happens in the scene.

Transcripts

A transcript is the full text of everything said and heard in the video, offered as a separate document or included on the page that hosts the video. Unlike captions, it isn't synchronized: readers can scroll through it at their own pace, search for a specific passage, or consult it without playing the video.

For someone with a hearing disability, the transcript is a complete alternative to the audio content. It also has a secondary benefit for organic and AI-assistant visibility: a search engine reads the text of a transcript far more easily than it can read a video file, and that helps the content get found.

Audio description

Audio description is an additional narration that describes out loud what's happening on screen: a facial expression, a gesture, a change of setting, on-screen text that isn't read aloud. It's needed whenever visual information is necessary to understand the scene and isn't already conveyed by the dialogue or sound effects.

It becomes essential when a video communicates something purely through images: a tutorial showing a movement, a chart that appears without commentary, a product demo where the interaction is entirely visual. In these cases, without audio description, a person with a visual disability loses part of the content that no other element of the video makes up for.

Audio description fits into the natural pauses in the dialogue. When those pauses aren't enough, you can produce a slightly longer version of the video, with extra pauses built in specifically to make room for it.

Caption contrast and readability

Caption color matters as much as caption content. The Web Content Accessibility Guidelines (WCAG) call for a minimum contrast ratio of 4.5:1 between text and background for normal text: by extension, the same threshold applies to captions overlaid on a video.

In practice, this means avoiding light text on an equally light background, or dark on dark, and being wary of combinations that rely on color alone to stand out (red on green, for example, is hard to read for people with color blindness). A semi-transparent background behind the text — even just a dark band at 50-60% opacity — solves most readability problems on video backgrounds that keep changing.

Position matters too. Captions should sit where they don't cover faces, hands signing in sign language, or informative graphic elements like captions, logos, or on-screen interfaces.

Caption font and size

A sans-serif font stays more readable on screen than a serif font, especially at the small sizes typical of captions. Size should stay proportional to the screen the video is most often watched on: what's readable on a monitor can look tiny on a phone, where most social content gets watched.

Two lines at a time is generally the limit beyond which a caption becomes hard to read in the time available.

Use ALL CAPS sparingly: it slows down reading compared to natural upper/lowercase, which the brain recognizes by word shape rather than letter by letter.

Alt text for social captions

On social media, video accessibility doesn't stop at the player. The post caption plays a role too, especially when it describes visual content that the platform's automatic captions don't cover. We apply this in our own content: in a recent Reel about the importance of alt text, the caption includes a description of what the video shows, in addition to the post's text.

This kind of description helps someone using a screen reader understand what they'd see if they could see the video, and it complements — without replacing — on-screen captioning.

Video accessibility on social media: differences between platforms

Video is only one piece of the picture: for the full set of practices — captions, alt text, hashtags, and more — see our guide to social media accessibility.

Every platform handles captions and descriptions its own way, and the differences matter when the same video gets published across multiple channels. Knowing them means not relying solely on automatic captioning, which remains a good starting point but always needs review.

Instagram (Reels)

Instagram generates automatic captions for Reels at the moment of publishing, which can be turned on with a tap. Automatic speech recognition still makes mistakes with technical terms, proper nouns, and regional accents, so it should always be reviewed before publishing. The post's alt text, separate from the video, remains the right place to describe visual elements that captions don't cover.

TikTok

TikTok works in a similar way: automatic captions can be turned on during editing, with the same need for manual review. The main difference from Instagram is in text styling, which can be customized in font and color directly in the app, making it easier to apply the contrast rules covered above.

(Note: Instagram offers the Edits app for creating and editing captions.)

YouTube

YouTube supports uploading separate caption files, in .srt, .sbv, or .vtt format, the most reliable option because it allows a full review before publishing. Since 2020, YouTube no longer offers community-contributed captions: responsibility now rests entirely with whoever uploads the video.

YouTube Shorts

Shorts inherit the vertical, short format of Reels, but caption handling still follows the YouTube player: you can upload .srt files for Shorts too, an option Instagram and TikTok don't offer in the same way for their own short formats.

LinkedIn

LinkedIn's audience is professional, and the style of video content follows a more restrained register, consistent with the measured visual pace we use in our own LinkedIn videos. Automatic captions are available here too, but the same rule applies as on the other platforms: they need review, especially for the technical terminology that comes up often in B2B content, because the transcription is frequently not faithful to the spoken audio.

Tools and processes for accessible video content

Many automatic transcription and captioning tools do a good job as a starting point. None of them removes the need for human review: speech recognition still gets proper nouns, technical terms, and overlapping voices wrong, and a caption error remains a barrier even when the source video was designed to be accessible.

A simple workflow helps you avoid skipping steps.

• Step 1: plan the content. Write a simple script, with one person speaking at a time and a clear message per scene.

• Step 2: record clean audio. A dedicated microphone and a quiet environment cut down on the correction work later.

• Step 3: generate captions and a transcript. Start with an automatic tool, then manually correct names, technical terms, and punctuation.

• Step 4: add audio description if needed. Only when the visual content communicates something the dialogue doesn't already cover.

• Step 5: test the result. Watch the video with captions on and no sound: if it still makes sense, that's a good sign.

We treat this kind of work as part of the same ongoing practice we apply to websites and documents: not a one-time check, but a process that repeats every time new content gets published.

Why invest in video accessibility: the numbers

Two figures help frame the scale of this issue.

According to CENSIS, around 7 million people in Italy — 12.1% of the population — live with a hearing difficulty (source: CENSIS, 2020). According to ISTAT, 18.6% of the population over age 15 lives with a mild or severe vision limitation (source: ISTAT, report "Condizioni di salute e ricorso ai servizi sanitari in Italia e nell'Unione europea").

A video without accurate captions excludes the first group. A video without audio description, when the visual content is central, excludes the second. Combined, these are percentages no business can afford to ignore if the goal is genuinely reaching its whole audience.

There's also a less direct but concrete benefit: transcripts improve how video content gets indexed by search engines, because they provide text a crawler can read in full. The same principle increasingly applies to AI-based assistants, which rely on text to understand what a video is about.

Inclusive video: where to start

Three things apply to any video (plus a bonus one), regardless of platform.

1. Automatic captions are a starting point, never a finish line, and always need correcting before publishing.

2. Contrast and text size aren't a cosmetic detail — they're the difference between a caption people can read and one nobody can follow.

3. Audio description needs to be planned, not bolted on at the last minute, because it requires pauses built into the edit on purpose.

4. There's also a fourth, less obvious point: thinking about accessibility while writing the script, not after editing, reduces the correction work downstream. A video designed from the start for a broad audience needs fewer follow-up fixes to captions, audio description, and contrast, because those choices are already part of the script.

If your team handles video production and wants to apply these practices systematically, our Academy includes training paths dedicated to multimedia content accessibility.

Frequently asked questions about making video accessible

What's the difference between captions and closed captions?

Captions report the spoken dialogue. Closed captions (CC) also add the scene's relevant sounds, like applause or a door closing, which help people with hearing disabilities.

Is audio description required for every video?

No, it's needed when the visual content communicates information the dialogue doesn't already cover, like a tutorial showing a movement or an entirely visual product demo.

What's the recommended minimum contrast for captions?

The Web Content Accessibility Guidelines (WCAG) call for a minimum contrast ratio of 4.5:1 between text and background, the same threshold used for normal text on the web.

Are Instagram's and TikTok's automatic captions already accessible?

No, automatic speech recognition still makes mistakes with proper nouns and technical terms, so captions should always be manually reviewed before publishing.

Why add a transcript in addition to captions?

A transcript offers the full text, readable at your own pace, useful for people who prefer reading, and it improves how video content gets indexed by search engines.