F74: Why Unlabeled Media Alternatives Fail Users

No label. That's the failure. A synchronized media alternative — a video with spoken instructions and visual demonstration — sits on a page, fully produced, technically functional, and completely disconnected from the text it was built to replace. Users who need it can't find it. Users who encounter it don't know where its corresponding text lives. The content exists. The relationship doesn't.

This is WCAG Failure F74 (opens in new window), and it's a failure pattern that reveals something important about how teams think about media alternatives: they treat production as the finish line.

The Failure

F74 describes a specific breakdown in how synchronized media alternatives are presented. The W3C's official description is precise: synchronized media alternatives "need to be clearly labeled with the text for which they substitute, so users can find them and so users who normally expect text alternatives to synchronized media know not to look for them."

The canonical example from the W3C document is instructive. A tax form page provides prose instructions — fields, data requirements, completion steps. It also provides a synchronized media alternative: a video with spoken instructions and a person demonstrating the process on screen. Both exist on the page. But the video appears elsewhere, disconnected from the text it replaces, with no label indicating what it substitutes for.

This fails two success criteria simultaneously:

The structural logic matters here. WCAG creates a category exemption: synchronized media that is itself an alternative to text doesn't need its own text alternative. That exemption only holds when the relationship is declared. Remove the label, and the exemption evaporates — along with the accessibility benefit the media was supposed to provide.

Why This Matters

Two distinct user populations are harmed, in opposite ways.

Users who would benefit from the synchronized media alternative — people with cognitive disabilities who process audio-visual information more effectively than dense prose, or people with reading disabilities who find a demonstrated walkthrough easier to follow — encounter the text version without any signal that an alternative format exists. They may struggle through content that was never designed to be their primary access path. The alternative was built for them. They never find it.

Users who encounter the video first face a different problem. Without a label identifying it as an alternative to specific text content, they don't know whether they need to find a text version. Screen reader users in particular may spend time searching for a text alternative to the video — because that's the expected pattern — when no such alternative is required or intended. The label isn't just wayfinding. It's a signal that changes what users need to do next.

This is the kind of failure that automated testing tools are poorly positioned to catch. As our analysis of automated versus manual testing methodologies documents, automated tools can detect the presence or absence of captions, but they cannot evaluate whether a piece of media is functioning as an alternative to adjacent text — or whether that relationship is communicated to users. The semantic relationship between a video and the prose it replaces is invisible to a scanner.

The Fix

The correction is structural, not cosmetic. The synchronized media alternative needs to be:

1. Positioned proximately to the text it replaces, or
2. Explicitly labeled with a heading or descriptive text that identifies it as an alternative to specific content

The label needs to do two things: tell users who need the alternative where to find it, and tell users who don't need it that they don't need to search for a text version of the video.

A minimal correct pattern looks like this:

<!-- FAILING PATTERN -->
<section id="instructions">
  <h2>How to Complete Section 3</h2>
  <p>Enter your total income from all sources in Box 3a. 
  If you received income from multiple employers...</p>
</section>
<!-- Video appears elsewhere on page with no connection to the text above -->
<div class="media-player">
  <video src="tax-instructions.mp4" controls>
    <track kind="captions" src="captions.vtt" srclang="en">
  </video>
</div>

<!-- CORRECTED PATTERN -->
<section id="instructions">
  <h2>How to Complete Section 3</h2>
  <p>Enter your total income from all sources in Box 3a. 
  If you received income from multiple employers...</p>
  <!-- Label explicitly identifies this as an alternative to the text above -->
  <div class="media-alternative">
    <h3>Video Alternative: How to Complete Section 3</h3>
    <p>This video provides a spoken and visual walkthrough 
    of the instructions above.</p>
    <video src="tax-instructions.mp4" controls>
      <track kind="captions" src="captions.vtt" srclang="en">
    </video>
  </div>
</section>

The heading "Video Alternative: How to Complete Section 3" does the work. It names the relationship. Users who need the video know it's there. Users who encounter the video know what text it corresponds to and that they don't need to locate a separate text version of the video itself.

If the media alternative must appear elsewhere on the page due to layout constraints, an in-text link pointing to it — with anchor text that identifies it as an alternative — satisfies the labeling requirement.

Applying This

Code review for this failure requires a content-level check, not just a markup check. The question isn't whether the video has captions (though that matters separately). The question is: does anything on this page tell users that this video is an alternative to specific text content?

Practical checks for development and QA teams:

• Proximity audit: Is the media alternative adjacent to the text it replaces? If not, is there explicit labeling?
• Heading scan: Does the media element have an associated heading or label that uses the word "alternative" or equivalent phrasing?
• Reverse navigation test: Starting from the video, can a user identify which text content it replaces? Starting from the text, can a user find the video alternative?
• Screen reader walkthrough: Navigate the page with a screen reader and note whether the relationship between text and media alternative is announced or discoverable

This failure type is a strong candidate for inclusion in content author checklists, not just developer checklists. The labeling decision is often made during content creation, not during development. If content management workflows don't surface this requirement, it will be missed consistently — regardless of what the code review process catches. As our research on compliance framework implementation shows, failures that live at the intersection of content and code tend to fall through organizational gaps.

CORS Perspective

F74 is a community-first failure. The synchronized media alternative was presumably built because someone recognized that a subset of users would benefit from it — people with cognitive disabilities, reading disabilities, or learning differences for whom a demonstrated walkthrough is genuinely more accessible than prose. The operational failure is that the production process completed without the labeling step, which means the community benefit never fully materialized. From a risk standpoint, the dual failure against SC 1.2.2 and 1.2.8 creates compounded exposure — but the more significant issue is that content teams may not understand why the label is required, making this a training and workflow problem, not just a remediation task. Strategically, organizations that invest in producing synchronized media alternatives have already made the harder commitment; ensuring those alternatives are properly labeled is a small additional step that determines whether the investment actually reaches the users it was designed to serve.