Web Accessibility Overlays Fail Practitioners, WebAIM Survey Finds
WebAIM survey data shows 67% of practitioners rate accessibility overlays as ineffective; a new guide details heading structure, alt text, and contrast fixes.
- AuthorTom Whitfield
- ChannelAccessibility
- Filed
- Read4 min

- WebAIM survey: 67% of accessibility practitioners rate overlays as not at all or not very effective; among respondents with disabilities, 72%.
- American Libraries published the guide on December 17, 2025, updating it April 23, 2026, after the US Department of Justice extended its web accessibility compliance deadline.
- WCAG permits new-tab links only in limited cases, such as preventing logout from secure content or avoiding disruption of a multistep workflow.
A WebAIM survey of web accessibility practitioners found that 67% of respondents rate automated accessibility overlays as "not at all or not very effective." Among respondents with disabilities themselves, disapproval rises to 72%. The findings anchor a practical guide published by American Libraries Magazine on December 17, 2025, and updated April 23, 2026, to reflect the US Department of Justice's extension of its web accessibility compliance deadline.
The guide's central argument: teams should stop treating overlays — pop-up widgets that promise accessibility compliance without code changes — as permanent solutions. Overlays may serve as temporary stopgaps during remediation, but the evidence from practitioners who rely on assistive technologies daily points the other way. The more durable work happens in HTML itself, starting with heading structure.
Headings are navigation, not decoration
Heading tags run from level one, the highest-level content, to level six. Content management systems typically assign each level a distinct style — different font, size, or color — which tempts designers to choose headings aesthetically rather than to reflect document structure. The cost falls on screen-reader users, many of whom navigate by jumping from heading to heading to skim a page's outline before selecting a relevant section. Without clear, concise headings, their only option is having the entire page read aloud from start to finish, which slows navigation considerably.
Alt text: brief, purpose-driven, never redundant
Alternative text conveys an image's content to screen-reader users when the HTML image tag is read aloud. The guide acknowledges the inherent tension between completeness and brevity, and offers two techniques. First, cut extraneous words — never begin alt text with "picture of" or similar. Second, describe the image's purpose in context: Van Gogh's Self-Portrait with a Straw Hat might warrant different alt text depending on whether the site presents it as a likeness of the artist, a study of his color use, or an example of his technique.
Functional images demand the most attention. Footer social media icons frequently lack meaningful alt text, rendering them invisible to assistive technologies. Conversely, alt text should not duplicate captions or surrounding text, or screen readers will repeat the same words consecutively — inefficient and confusing. Purely decorative images, such as a logo repeating the page title, should carry empty alt text so screen readers skip them entirely.
Color and contrast
Color can signal success or failure of actions, build brand recognition, and direct attention. Used carelessly, it creates barriers for users with color-vision deficiency, low vision, or certain assistive technologies. The guide's rules: never rely on color alone. Red error messages should pair with explanatory text. Colored links, particularly non-underlined ones, need another distinguishing signal. Data visualizations should add text labels and patterns as secondary channels.
Contrast between adjacent colors — text against background — determines whether users can read at all. Many accessibility testing tools now run automated checks that confirm whether contrast meets WCAG standards.
Links: descriptive text, same tab
"Click here" fails a basic screen-reader workflow: assistive technologies let users jump directly to links and hear each link's text read aloud. If every link says "click here," selection becomes guesswork. Raw URLs as link text are equally bad — screen readers read every character. Best practice is unique, destination-specific link text, such as "visit our interlibrary loan page."
Links should open in the same tab by default. New windows complicate back-navigation for blind and low-vision users, hinder users with mobility disabilities, and disorient users with cognitive disabilities. WCAG guidance permits new tabs in limited cases — such as linking from secure content that would otherwise log users out, or where a same-tab link would "significantly disrupt a multistep workflow" — but the default should let users decide.
Plain language and prioritized workflows
Readability guidelines include logical organization from general to specific, common vocabulary over jargon, active voice and present tense, and evaluation of reading level, font choice, size, contrast, and white space.
For teams unsure where to start, the guide recommends prioritizing by impact: fix barriers users have already reported, combine automated testing tools with actual assistive technologies, begin with the most popular content, and prioritize pages likely to attract assistive-technology users — accessibility information pages and feedback forms. Training everyone who creates or edits content prevents new problems from accumulating.
Test with representative users
Even libraries with established UX testing programs underrepresent people with disabilities. The guide recommends deliberate recruitment through disability services offices, student groups, or local organizations run by and for individuals with disabilities. Remote testing lets participants work from their own setups with familiar devices, operating systems, keyboards, and assistive technologies. Interview questions should capture those exact configurations, and the protocol itself must remain flexible enough to accommodate them.
The DOJ's deadline extension buys organizations time, but the practitioner data on overlays suggests where that time is best spent: on structural fixes and representative testing rather than widget licenses.
Original: americanlibrariesmagazine.org
More from Tom Whitfield
Show full bio
Correspondent covering companies, deals and strategy in Product and interaction design, design systems at Interface Wire.
65 articles