Accessibility usually gets framed as a compliance obligation. For creators it is the opposite: it is the cheapest audience expansion available. Around one in six people worldwide live with significant disability, and a large share of everyone else watches social video with the sound off — commonly cited estimates put that near 80% for feed-based platforms. Captions and alt text serve both groups with the same work.
The good news is that most of this is not a redesign. It is a set of habits you bolt onto the edit and the upload: accurate captions, described images, readable contrast, predictable sound, and a page people can actually navigate. Do it consistently and you will notice the effects in retention, saves, search visibility and brand deals long before you notice the extra twenty minutes it costs.
TL;DR: Accessibility is a growth channel. Auto-captions are a starting point, not a finished product — fix names, jargon and music cues before you publish. Add alt text to every image using "subject + action + context", keep overlay text at a 4.5:1 contrast ratio and on screen long enough to read, mix dialogue forward and label non-speech sounds in the caption file, and provide transcripts for video and audio. Then test it: watch your own content muted, read your captions at 1.5x, and navigate your link-in-bio page with a screen reader. Build the checklist into your export routine once and it becomes permanent.
The case, in numbers you can act on
Accessibility arguments tend to stay abstract, so start with the operational version:
| What you fix | Who it helps | What you get back |
|---|---|---|
| Accurate captions | Deaf and hard-of-hearing viewers, muted scrollers, non-native speakers, anyone watching in public | Higher watch time, better retention, more search surface |
| Alt text on images | Screen reader users, low-bandwidth users, broken-image loads | Indexable image context, better accessibility scores |
| Transcripts | Deaf viewers, readers, search engines | A second entry point into the same content, ranked differently |
| High contrast and readable type | Low-vision viewers, everyone on a phone in sunlight | Fewer drop-offs on your landing pages |
| Dialogue-forward audio mix | Hard-of-hearing viewers, commuters, laptop speakers | Fewer early exits; audio is the most common reason people bail |
| Clear, consistent navigation | Everyone, especially screen reader and keyboard users | More clicks on the links you actually care about |
Almost none of these are "accessibility features" in the narrow sense. Captions, transcripts and alt text are also SEO infrastructure: they are text, they are indexable, and they turn your media into something search engines and AI answer engines can quote.
Captions: the fastest win, and the easiest to do badly
Auto-captions are close to free and about 70–85% accurate on clean audio — which sounds fine until you realise that accuracy collapses on exactly the things creators care about: names, brands, technical terms, product names, and accents. A caption stream that says the wrong product name is worse than no caption, because it misleads the viewer and misrepresents the brand you were paid to feature.
The working routine:
- Let the platform generate the first pass. YouTube, TikTok, Instagram, LinkedIn and CapCut all do this now.
- Edit the words that matter. Your own name, your guests' names, brand names, products, numbers and jargon. This takes five minutes on a ten-minute video.
- Keep lines short. Aim for roughly 32–42 characters per line and no more than two lines on screen. Long lines force eye movement and break comprehension.
- Match reading speed to speech. If you talk fast, do not compress into one-word flashes. Two full lines held steady beats four flickering fragments.
- Label non-speech audio.
[laughs],[door slams],[music swells],[engine revs]. Without them, a viewer who cannot hear loses the joke, the tension, or the context. - Identify speakers in interviews with names or
[host]/[guest], at least on first appearance. - Export a subtitle file (
.srtor.vtt) and keep it, so you can reuse captions across platforms and burned-in versions instead of re-doing the work every time.
Burned-in captions and platform captions do different jobs. Burned-in text is part of the video and survives every repost, which matters because the majority of watchers see your content stripped of platform captions on someone else's feed. Upload a proper caption file too, so the text is selectable, searchable and screen-reader accessible.
Alt text that actually describes something
Alt text replaces an image for someone who cannot see it. Two failure modes dominate: the empty alt ( or "image1.png") and the keyword spam ("biolinky creator link in bio tool biolinky"). One helps nobody, the other reads like a scam.
Use a simple formula: subject + action + context, in one sentence, without "image of".
| Bad alt text | Better alt text |
|---|---|
IMG_4823.jpg |
Creator filming a product review on a phone mounted to a tripod in a kitchen |
link in bio tool screenshot |
Screenshot of a link-in-bio page showing four large buttons for a newsletter, shop, latest video and contact |
chart |
Bar chart showing watch time rising from 2 minutes to 7 minutes over eight weeks |
Three rules that cover most cases:
- Describe what a sighted viewer needs to understand the content, not every pixel. If the image is decorative, leave alt empty and let assistive tech skip it.
- Put the words in the image into the alt text — headline text, numbers on a chart, the offer on a graphic. That text is invisible to everyone else.
- Write alt text when you create the asset, not at upload time for a dozen platforms. Store it in your content spreadsheet next to the filename.
Contrast, type and overlay text
Design choices decide whether people can read what you made. WCAG — the standard most accessibility law and corporate procurement points back to — asks for a contrast ratio of at least 4.5:1 for body text and 3:1 for large text (roughly 24px, or 19px bold). Light grey text on white, thin fonts on busy photos, and pastel-on-pastel all fail that bar.
Practical rules for social design:
- Never place text directly on a busy photo. Add a solid or semi-opaque panel behind it, or a strong gradient scrim.
- Do not rely on colour alone to convey meaning. Pair it with a label, an icon or a shape — about 8% of men have some form of colour vision deficiency.
- Keep text on screen long enough to read. A rough guide: about 0.3 seconds per word for text-heavy screens — a five-word overlay needs around 1.5 seconds minimum, more if the viewer is also processing audio.
- Avoid flashing content faster than three times per second. This is a genuine seizure risk, and it is one of the few accessibility issues that can harm someone in the moment.
- Keep type large in feed video. Assume the smallest screen, not the editor monitor. If you have to squint on a phone, it has failed.
- Use a dyslexia-friendly pairing for long-form text: generous line spacing, no justified alignment, moderate line length (roughly 50–75 characters).
Audio: the mix decides retention
For hard-of-hearing viewers and for everyone in a noisy room, your audio mix is an accessibility feature.
- Mix dialogue forward. Voice should sit clearly above music. Duck the bed 6–12 dB under speech rather than pushing the voice up into clipping.
- Cap your peaks and normalise. Aim for a consistent loudness across videos so your viewers are not adjusting volume every upload.
- Avoid music under important information. If the voice-over lists a price, a rate or a deadline, bring the music down to near-silence.
- Caption the environment. A door slam, a notification chime, a rising crowd — if it carries meaning, put it in the caption file.
- Offer an audio described version for visual-only storytelling. For a heavily visual video (a tutorial where the important action is off-mic, a travel piece), a second version with narration describing what is on screen opens the content to blind viewers. This is a bigger lift, so reserve it for flagship content.
Platform-by-platform checklist
| Platform | Where captions live | Where alt text lives | The one thing creators forget |
|---|---|---|---|
| YouTube | Subtitles tab — upload .srt, edit the auto-generated track |
Alt text lives in the description for community post images; for thumbnails, repeat the on-screen text in the title and description | Edit auto-captions; default accuracy will miss product names |
| TikTok | Auto-captions + text overlays; downloadable .srt in some markets |
Alt text field appears after adding a photo or on the video's "more options" | Alt text on photo posts, not just video |
| Auto-captions on Reels; edit before publishing | "Write alt text" under Advanced Settings when posting | The off-screen caption text is what screen readers read — write it deliberately | |
| Auto-captions on video; edit in the post flow | "Add alt text" on every image | Baked-in text plus a caption file beats baked-in alone | |
| Podcasts | Show notes plus a full transcript page | N/A | Publish the transcript as its own HTML page, not a PDF |
| Newsletters / blog | Plain-language headings, descriptive link text | Alt text on every image | "Click here" tells a screen reader nothing — use "Read the rate card" |
| Your link-in-bio page | Readable default text size, high contrast | Alt text on every button image and background | Background images with text baked in and no alt text |
Test it in five minutes
You do not need an audit tool to catch most problems:
- Watch your last video muted. Can you follow the whole thing? If a step disappears, the step needs on-screen text or clearer visual language.
- Read your captions at 1.5x. If you cannot keep up, neither can anyone else.
- Turn on your phone's screen reader and navigate your own link-in-bio page. Are the buttons announced with meaningful labels, or does it read "image, link, image, link"?
- Zoom your page to 200%. Does the layout hold, or do buttons overlap and text clip?
- Look at a thumbnail on a phone at arm's length in daylight. If the headline vanishes, the contrast is failing.
Making it automatic
Accessibility collapses when it depends on remembering. Two habits prevent that:
- Put the checks at the end of the edit, not at upload. A fixed export routine: captions corrected, alt text written, transcript exported, thumbnail contrast checked, link page labels verified.
- Keep an accessibility section in your content spreadsheet with columns for alt text, caption file path, and transcript URL. Whoever uploads — you, a VA, or an agency — copies from the sheet instead of improvising.
Compliance is also becoming a commercial factor. Accessibility requirements for public-facing digital products have tightened across the EU, UK and US, and brands increasingly ask partners whether their content meets baseline standards. Being the creator who already does captions, alt text and transcripts correctly is a small, quiet advantage in a pitch.
Your link-in-bio is content too
The page your entire audience clicks through to is the most overlooked accessibility surface creators own. Text baked into background images, low-contrast buttons, tiny tap targets, and icon-only links with no labels all exclude people at the exact moment they tried to act.
A Biolinky page lets you build that page with real text labels instead of image buttons, descriptive alt text on every visual element, strong default contrast, and link names that make sense when read aloud — so "Read the rate card" and "Join the newsletter" are announced properly by a screen reader instead of "link, link, link". It is a small change with an outsized effect: the audience you were accidentally turning away at the final click can finally get through.
Fix the captions, describe the images, raise the contrast, mix the voice forward, and test your own work with the sound off. Do it once in your export routine and accessibility stops being a task and becomes a property of everything you publish.