Give every graphic a brief text description
Some images cannot be reduced to a sentence, and G95 is the honest minimum for those. A short text alternative that briefly describes the content, with a longer description supplied separately by another technique. W3C counts it as sufficient for the non-text-content rule at Level A, using a more specific technique, which in HTML is the alt attribute. Read the scope carefully, because this one is easy to misuse. G95 is for content whose meaning genuinely will not compress, like a complex chart or a photograph carrying an argument. It is not a licence to write photo on everything and call it finished. Where the meaning does fit in a sentence, the fuller technique applies and the alternative has to carry the whole of it.
How we find it in an audit
Every image gets read for whether its alternative describes the content or the file. Photo, image, banner and chart are the four we find most, and not one of them describes anything. Where an image is genuinely too rich for a sentence, we check a longer description exists somewhere and that the short one points at it. Then a reviewer looks at the image beside its text to settle the only question that matters, which is whether somebody who cannot see it learns what it was doing on the page.
How affected users experience it
Alt text reading image tells a screen-reader user that something is here and nothing whatsoever about what. They cannot tell whether it mattered, so they either skip it and risk losing the point or stop and try to reconstruct it from context. Both cost time, and the second costs confidence as well. A brief description ends the question in six words. Rescue workers carrying sandbags along the flooded high street. Not everything about the photo, and enough to know what it says.
Passes vs. fails
Passes
Alt text reading Rescue workers carrying sandbags along the flooded high street. Brief, and the situation lands.
Fails
A news photo carries alt text reading photo, so the subject, the place and the relevance are all withheld.
How this gets tested
The W3C publishes test rules that define what a checker looks for here.
- Element marked as decorative is not exposedA tool can check this
- Image accessible name is descriptiveA tool finds candidates, you decide
- Image button has non-empty accessible nameA tool can check this
- Image has non-empty accessible nameA tool can check this
- Link has non-empty accessible nameA tool can check this
- Object element rendering non-text content has non-empty accessible nameA tool can check this
- SVG element with explicit role has non-empty accessible nameA tool can check this
- Image not in the accessibility tree is decorativeA tool can check this
Other ways to satisfy this rule
39 guides on this site are filed under 1.1.1 Non-text Content. W3C lists this one as sufficient for that rule when it is implemented through a more specific technique, so the condition is part of the test rather than a footnote to it.
- ARIA6sufficientName icon-only controls with aria-label
- ARIA9sufficientAssemble one label from several text pieces
- ARIA10sufficientName images and charts with aria-labelledby
- ARIA15sufficientLink complex images to longer descriptions
- C9sufficientPut decorative images in CSS backgrounds
- G68sufficientDescribe the purpose of live-only streams
This guide is our interpretation of W3C technique G95: Providing short text alternatives that provide a brief description of the non-text content. W3C publishes its techniques as guidance rather than as the standard, and says so on every one of them. The success criterion is what conformance is measured against, and a technique is one documented way to meet it.