Why testing is more than a checklist
Following guidelines gets you a long way, but guidelines describe what is generally helpful. They cannot tell you whether a particular person, using a particular device in a particular situation, can actually get what they came for. Accessibility testing is the step that closes this gap. It combines quick technical checks with feedback from real people.
Think of it in three layers: automated checks, your own manual checks, and testing with users. Each catches different problems, and none is a replacement for the others.
Layer one: automated checks
Free checking tools and built-in browser features can scan a page and flag items such as missing image descriptions, low color contrast, empty links or form fields without labels. They are quick and are good for catching obvious slips.
Their limits are important. A tool can tell you that an image has a description, but not whether the description is useful. It can tell you headings exist, but not whether the outline makes sense. Experienced practitioners generally say that automated tools find only a portion of the possible problems, and the exact share depends on the tool and the content. Treat a clean report as a good start rather than proof of accessibility.
Layer two: simple manual checks
You can do these yourself in an afternoon.
- Keyboard only. Put the mouse aside and use the Tab key to move through the page. Can you reach every link, button and form field? Can you see where focus is? Can you leave a menu or pop-up without getting stuck?
- Zoom and resize. Enlarge the page to a very large size, or use the browser’s zoom to magnify text. Does anything overlap, get cut off or force awkward sideways scrolling?
- Sound off. Play your videos with the sound off. Do captions carry the message? Then listen without watching a video: does the audio still make sense, and is there a transcript?
- Screen reader skim. Most devices include a basic screen reader. Listen to how your page is read. Pay attention to the heading list, link wording and image descriptions. It takes some practice, and your own experience will differ from that of an experienced user, so use this for rough insight only.
- Grayscale. Switch to grayscale to see whether color alone is carrying meaning.
- Motion check. Look for flashing or auto-moving elements and check there is a way to pause.
- Read it out loud. Clunky or unclear passages tend to reveal themselves.
Write down what you find, fix the most serious problems first, and recheck.
Layer three: testing with real users
Real feedback tells you what you could not predict. People who use assistive technology every day have habits, shortcuts and expectations that a newcomer would not guess, and they can quickly tell you whether a page is merely technically compliant or truly usable.
Involve disabled readers respectfully
- Recruit thoughtfully. Look to community groups, organizations and networks that connect with people who use assistive technology, and ask whether people are willing, not whether they are obliged. A single person does not speak for everyone with a given disability.
- Offer fair compensation. Their time and expertise are valuable. Pay them if at all possible, and be clear upfront about the amount and the length of the session.
- Be flexible. Offer a choice of format, whether a video call, a phone call, written feedback or an in-person visit, and let people use their own device and settings.
- Explain what you are testing. Say that you are testing the content, not the person. Make it clear that difficulty is a sign of a problem with the content.
- Respect privacy. Ask permission before recording or quoting, and keep personal details confidential. Do not ask for medical details you do not need.
- Credit and thank them, if they wish to be named, and tell them what you changed afterward.
Keep sessions simple
A handful of participants will usually reveal repeated problems, so you do not need a large study. Give each person a few realistic tasks, such as “find the opening hours” or “watch this video and tell me what it is about,” and ask them to think aloud if they are comfortable doing so. Then observe without helping too quickly. Silence while someone is working it out is normal.
Ask good questions
Useful prompts include: “What did you expect to happen there?”, “What would you call this section?”, and “Was there anything that slowed you down?” Avoid leading questions such as “Wasn’t that easy?” Short reader surveys and audience interviews can complement this by collecting broader opinions over time.
Making use of what you learn
Sort feedback into three groups: things that blocked someone completely, things that caused frustration, and preferences. Fix blockers first. When feedback from different people conflicts, which happens often because needs differ, look for solutions that offer choice, such as adjustable caption size or a text alternative to a video.
Record what you changed and why, so that future contributors do not undo it. A short note in your style guide is enough.
Make it routine
Testing is not a one-off event. Content changes, templates change and assistive technology changes. Build brief checks into your publishing routine, and schedule a deeper review now and then. Accept that you will not find everything, and that accessibility is a continuing practice. The aim is steady improvement, informed by the people your content is meant to serve.