Making website images accessible through useful descriptions

On a clean, text-led website, images can appear secondary, yet a map, diagram, photograph, or chart may be the fastest way to explain a topic. Alternative text provides a concise description when an image cannot be seen or loaded. It is stored in an image’s HTML alt attribute and read by assistive technology, allowing the visual and textual parts of a page to work together.

This matters across a diverse Australian audience. People who are blind or have low vision often navigate with screen readers, while others magnify pages, use reduced-data settings, or encounter image-loading problems. Written alternatives also help people understand a chart’s purpose without searching for a caption, especially when viewing a page on a phone.

The right description depends on the image’s job, not a formula. A photograph may need context, a chart may need its main finding, and decoration should usually be ignored by screen readers. Useful alt text is specific, economical, and written for the page rather than repeated from a filename or stuffed with search terms.

Alt text explains what an image is doing

Start by considering what the visitor should learn. If a photo identifies a place, say which place and any detail needed for recognition. For a diagram showing how a process works, summarise the relationship rather than merely naming the diagram. This purpose-led approach keeps the alternative tied to the information the page is trying to communicate.

For example, a picture of Sydney Harbour Bridge could use “Sydney Harbour Bridge viewed from the western side” when geography matters. If the same image only reinforces an adjacent heading, it may be decorative and use an empty alt attribute. Decorative images still belong in HTML, but they should not interrupt a screen reader’s reading of the content.

Screen readers need meaningful descriptions

When an image lacks a useful alternative, some screen readers announce its filename or nothing useful at all. That creates a silent gap: the visual page appears complete while the spoken version skips information. A concise description lets users decide whether the image is relevant and continue without losing their place.

Descriptions should add information, not narrate obvious presentation details. “Image of a person” usually says less than “A cyclist riding beside the Melbourne tram network” if the cycling context matters. Avoid duplicating nearby sentences, and keep complex charts, maps, or infographics available through a longer description or accessible data table.

Australian accessibility expectations

Australia’s Disability Discrimination Act 1992 makes accessible access to goods, services, facilities, and information an important legal consideration for many organisations. For Australian Government digital services, the Digital Service Standard requires conformance with WCAG 2.1 Level AA. Private, not-for-profit, and commercial sites should not assume that a different audience means a lower accessibility standard.

WCAG 2.1 AA treats text alternatives as a foundational requirement, including for images that convey meaning. It does not prescribe one fixed character count, which is why rigid limits can produce clipped or vague wording. The useful test is whether the alternative communicates the same purpose in the context where the image appears.

Accessibility also supports the Australian market’s broad device mix. Australians commonly encounter public information on phones, desktops, tablets, and community-run devices, and busy websites are often designed around mobile screens. An alt attribute adds little data, so it remains available when a visitor disables images or a large file fails to load.

Context beats keyword stuffing

Search engines can use alt text as one signal about an image, but repeating phrases such as “Australian travel guide” inside every description neither helps a reader nor improves relevance. On a general information site, descriptions should reflect the actual subject: a banknote, a map, a document scan, or an illustration of a service. Truthfulness matters more than promotional density.

Context also changes the correct wording. An image used as a link to a Canberra attractions guide should name the destination, such as “Canberra attractions guide”, while a standalone image of Parliament House can use “Parliament House in Canberra”. A magnifying-glass icon that launches search needs the label “Search this site”, not “magnifying glass”. These choices make controls understandable without sight.

Australian context needs Australian clarity

Local references should be precise enough to be useful. “A busy city street” may be inadequate if the page concerns Melbourne’s laneways, Sydney Harbour, Adelaide’s Central Market, or Darwin’s waterfront. Naming the right place helps readers connect the image with the surrounding copy. It also avoids implying that every Australian city looks like Sydney or Melbourne.

Australian English and everyday expressions need careful treatment. If visible signage includes the word “arvo”, a transcription may need to preserve it; otherwise, adding slang such as “no worries” to an image description can sound forced. Because Australia includes many linguistic and cultural communities, plain wording is usually more dependable than trying to imitate local speech.

Geography and connection styles make this especially relevant. People in regional and remote areas may rely on a phone to access public information, sometimes across slower or intermittent connections, so text alternatives must remain meaningful when an image fails to load. Descriptions should also handle Aboriginal and Torres Strait Islander artwork, Country, and language respectfully rather than making cultural claims from appearance alone.

A reliable alt text workflow

A workable review begins with an inventory of every image, including logos, icons, background photographs, charts, and images embedded in articles. For each item, record whether it is informative, functional, or decorative. Informative images need a text alternative; functional images need wording that explains their action; decorative images should usually have an empty alt attribute so assistive technology skips them.

Next, draft the description from the page’s purpose and read it without displaying the image. If it remains clear, specific, and useful, it is probably doing its job. Check that nearby text is not repeated, that essential text shown inside a meaningful image is included or otherwise available, and that detailed graphics have another route to the full information. Automated tools can flag missing attributes, but they cannot judge whether a description is accurate.

Finally, test navigation with a keyboard and at least one screen reader, then build the habit into content publishing. Authors benefit from examples, CMS prompts, and a clear rule for handling complex images. On a minimal informational site, this discipline prevents inaccessible gaps from weakening otherwise clear layouts, readable typography, and dependable navigation.