Alt Text and Accessibility Labels

What this parameter measures

Two things decide this score: whether your images carry alt text, and whether your interactive controls carry accessible labels. friendly4AI scans the HTML, counts every <img> and how many have an alt attribute, then counts interactive controls (<input>, <button>, <select>, <textarea>) and how many are labelled through aria-label, aria-labelledby, a title attribute, or, for non-input controls, visible text content. It then computes two coverage ratios and averages them into one overall accessibility-coverage figure.

Why it matters for AI-readiness

Alt text and ARIA labels are how AI systems read non-text content. An image without alt is opaque to a crawler; a button labelled only by an icon is meaningless to a machine. Strong labelling lets answer engines understand what your visuals depict and what your controls do, which improves how completely they can describe and cite your page. Accessibility coverage also doubles as a quality signal — well-labelled pages read as well-built and trustworthy to both assistive technology and AI.

How we score it

Under the v4.4 methodology, this Content Structure parameter is a gradient based on combined coverage. The processor calculates image alt coverage (images with alt / total images) and control label coverage (labelled controls / total controls), then takes their mean. A page with no images or no controls in a given dimension is treated as full coverage (1.0) for that dimension, so it is not penalized for absent elements. A page passes (100) when the combined coverage is at least 90%, earns a partial (50) when it is between 60% and 90%, and fails (0) when it is below 60%. These bands match the published rubric directly.

How to fix common issues

• Add a descriptive alt attribute to every meaningful image, for example <img alt="Product dashboard screenshot">.
• Use alt="" (empty alt) for purely decorative images so they are correctly skipped rather than left unlabelled.
• Give every <button> and icon control an aria-label or visible text describing its action.
• Label form inputs with aria-label, aria-labelledby, or an associated <label>; inputs are not credited for visible text alone.
• Audit for the lowest-coverage dimension first — a strong image score cannot offset many unlabelled controls because the two are averaged.
• Re-scan to confirm combined coverage reaches 90% or higher.