Keyboard and Switch Users
If it works with a mouse but not with a keyboard, it does not work. This is the single highest-yield thing to test, and it takes ten minutes.
Unplug your mouse and try to buy something on your own site.
A switch user may need several seconds and several activations to reach a single control, so every unnecessary tab stop is a real cost and every missing one is a wall. The most common catastrophic failure is the focus trap: a modal or a custom widget the keyboard can enter but not leave, which ends the session because there is no mouse to click the close button with. The second most common is the invisible focus ring — the page is fully operable, but there is no way to see where you are.
Scale: About 1 in 8 U.S. adults has a mobility disability, and keyboard operability also serves screen reader users, switch users, and anyone with a temporary injury. (Source: CDC. Survey-based disability figures undercount — treat this as a floor.)
What they're using
- A standard keyboard, used with Tab, Shift+Tab, Enter, Space, and arrow keys
- Single- and dual-switch devices driven by head, hand, or breath, which emulate scanning through focusable elements
- Sip-and-puff systems and head pointers
- Ergonomic and one-handed keyboards
7 barriers we find most often, and the criteria behind them.
Controls the keyboard cannot reach
Click handlers on divs and spans, custom dropdowns, carousels, and drag-and-drop interfaces built without keyboard equivalents. If it is interactive, it must be focusable and operable.
Success criteria: 2.1.1 Keyboard A
Focus traps
Modals, embedded players, and third-party widgets that take focus and never give it back. The user cannot continue and cannot leave without closing the tab.
Success criteria: 2.1.2 No Keyboard Trap A
No visible focus indicator
Removing the default outline without replacing it is one of the most common lines of CSS on the web, and it makes a site unusable for sighted keyboard users.
Success criteria: 2.4.7 Focus Visible AA · 2.4.11 Focus Not Obscured (Minimum) AA
Tab order that does not match the page
Positive tabindex values and DOM order that disagrees with the visual layout send focus jumping around unpredictably.
Success criteria: 2.4.3 Focus Order A · 1.3.2 Meaningful Sequence A
No way to skip repeated navigation
Without a skip link or proper landmarks, every page begins with the same forty tab stops before the content starts.
Success criteria: 2.4.1 Bypass Blocks A
Targets too small, and actions that fire on the wrong event
Small controls are hard to hit for anyone with a tremor or limited dexterity, and controls that activate on mousedown rather than on release remove the ability to abort a mistake.
Success criteria: 2.5.1 Pointer Gestures A · 2.5.2 Pointer Cancellation A · 2.5.8 Target Size (Minimum) AA
Timeouts that assume typing speed
A session or a form that expires on a fixed timer penalizes anyone who takes longer to enter information, with no warning and no way to extend.
Success criteria: 2.2.1 Timing Adjustable A
The changes that actually remove the barrier.
Build with real interactive elements
Buttons, links, and form controls are keyboard-operable by default. Most keyboard failures are divs that should have been buttons.
Design the focus indicator on purpose
Do not remove the outline — replace it with something that meets 3:1 contrast and is visible on every background it will appear on.
Manage focus around dialogs
Move focus in on open, cycle it within, restore it to the trigger on close, and close on Escape.
Add a skip link and use landmarks
One skip link at the top of the DOM and correct header, nav, main, and footer elements.
Give people time
Warn before a timeout and let the user extend it, or do not impose one.
Three tests you can run before you call anyone.
- Put the mouse away. Complete your highest-value task with the keyboard alone.
- Watch for the focus ring the entire time. If you lose track of where you are, so will everyone else.
- Open every modal on the site and try to leave it with Escape and with Tab.
If any of those go badly, you have your answer and you did not need a vendor to get it. What a vendor is for is the part after: knowing which of the 55 criteria that maps to, how deep it goes, and what it costs to fix properly rather than patch.
Run the Free CheckerWhere this shows up in practice
Other people the standard was written for.
Voice Control Users
Voice control users say what they see. When the visible label and the accessible name disagree, the command fails silently.
Read the profilePeople With Cognitive and Learning Disabilities
The largest disability group and the least served by automated testing. Almost nothing here can be caught by a scanner.
Read the profileAll 55 Success Criteria
Every Level A and AA criterion in plain language, with the failure we see most often for each.
Reference