Accessibility for the visually impaired

A directive came down requiring websites to be accessible to people with visual impairments. The easy path: add a standard plugin and call it done. I proposed doing it properly: actually understand how visually impaired users use websites.

Self-initiated the research. With the project manager, arranged a visit to an institution teaching computer literacy to visually impaired users.

Conducted a series of observations and sessions: watched how people use browsers and screen readers, where they get lost, which navigation patterns they use — keyboard navigation, zoom, High Contrast Mode.

Key observations: — Most users work with screen readers, not accessibility plugins — semantic HTML matters more than a 'visually impaired version'. — The main problem isn't font size but focus order and alt-text quality on images. — The 'accessibility version' toggle confused users: unclear what it enables and how to go back.

Based on observations, built a toolkit and implemented it on concert and theater organization sites:

— Fixed semantics: headings in correct order (h1→h2→h3), navigation via <nav>, ticket buttons as <button> not <div onclick>. — Added alt texts to posters, artist photos, banners — concise, without starting with 'image of'. — Improved focus order: Tab should follow logical reading order, not DOM order. — Reworked the accessibility toggle: made its function obvious, added 'active' state.

My first real hands-on UX research and inclusive design experience — 2017, well before accessibility became standard practice in the design community.

Key insight: you can't design for accessibility from guidelines alone without seeing how people actually use the product. A standard plugin would have given the appearance of a solution — not the solution.