For many of our partners, screen readers are a bit of a mystery. Even those who are familiar enough with how screen readers work are often unsure how screen readers should be incorporated into their testing workflows. There are at least 5 major screen readers along with a number of other options; companies understandably want to know the best places to focus their limited resources. WebAIM, an accessibility nonprofit based out of Utah State University, produces a survey of screen reader users that can help inform our decisions and priorities in this area.
More about the WebAIM survey
This year (2024), WebAIM produced their 10th screen reader user survey. These surveys are conducted every 1-2 years and are one of the best sources of data on screen reader market share, OS and browser preferences of screen reader users and other insights into this population of users. (You can also supplement your understanding with the new Web Accessibility Survey sponsored by AmericanEagle.com.)
Why is this important for you
Our approach to accessibility testing using screen readers is based on earlier versions of this survey data. Obviously it’s important to validate our assumptions and make sure that they still are still accurate. Web technology changes and user behaviors change along with it.
But more broadly, as a digital professional, it is important to rely on actual data. Web professionals use the web differently than most people and have different preferences in their browser and OS usage. Even accessibility professionals (and even those with disabilities) that know how to use screen readers and other assistive technologies are often different than the typical screen reader user. Having user data can help them actively correct for these biases and make informed decisions that account for actual usage patterns.
It’s worth reading the full set of results and findings for yourself. But here are our big takeaways that we think are important for digital professionals to understand.
Subscribe to our Newsletter
Actionable testing insights
Don’t assume users are like you
Digital professionals are disproportionately Mac users with large high resolution monitors. They also make up a pretty significant portion of Linux users. Even setting accessibility aside it’s important for us to keep in mind that the vast majority of desktop users (over 70%) are on Windows. For screen reader users, those differences are magnified.
- 86.1% of respondents use Windows as their primary operating system
- Among screen reader users, those without disabilities are 3 times more likely to use MacOS
As long as we are on our Macs, it’s easy to over rely on the built-in VoiceOver screen reader. That can’t be the full extent of screen reader testing. It’s critical to test in Windows and to use Windows screen readers like JAWS or NVDA.
Screen reader users have different preferences
While we all need to get out of our bubbles, it’s important to realize that screen reader users may also differ from the typical internet user in important ways. We’ve already touched on how screen reader users are a bit more likely to be Windows users. On mobile, it’s a different story.
- 70.6% of screen reader users use an Apple iOS device as their primary mobile platform. (In contrast, Android has 70% market share globally.)
- Within North America, iOS is even more popular with 84% share among screen reader users (vs 57% among the general population).
For desktop users, Chrome has been the top browser for many years. Roughly 65% of users worldwide use Chrome but for screen reader users that is only 52.3%.
It’s important not to overstate these trends. There are still significant numbers of users that are on Android and it is important to do OS-specific accessibility testing. But it does make sense to use an iPhone as your primary accessibility testing device. In contrast, it is a good idea to weight your desktop accessibility testing a bit more towards using Firefox, Edge or Safari than you otherwise might.
Screen reader and browser testing combinations
NVDA (37.7%) and JAWS (40.5%) are neck-and-neck as the most popular primary screen reader. And over 60% of all respondents commonly use both NVDA and JAWS. Voiceover rounds out the top 3 with 9.7%. (Interestingly, screen reader users without disabilities use MacOS and VoiceOver 4 times as much as screen readers with disabilities.)
Our usual recommendations when doing comprehensive cross-screen reader testing is to use these pairings.
- Firefox with NVDA
- Chrome with JAWS
- Safari with VoiceOver
However, it’s worth noting that for both NVDA and JAWS screen readers, Chrome is still by far the most popular browser. That means if you only have budget to test on NVDA, you should consider focusing more of your testing on Chrome (over Firefox).
How to prioritize accessibility fixes
The survey also includes some interesting data on how screen reader users navigate the web and what they consider to be the difficult or frustrating issues they encounter. Keep these findings in mind as you consider what to remediate first on your site.
Heading structure and order
On lengthy web pages, screen reader users overwhelmingly (71.6%) first attempt to use section headings to find what they are looking for. Along those same lines, 88.8% of respondents find heading levels (e.g. H1, H2, etc.) useful. You can make your page much more navigable by taking care to maintain a heading hierarchy that is logical without skipping levels. It’s also important to make sure that your section headings accurate describe the contents of each section.
Ranked problematic items
According to the survey, more than half of all screen reader users say they almost never contact site owners about accessibility barriers. If you are lucky enough to receive direct feedback, you may already have some anecdotal data about where users have difficulty with your site.
If not, you can especially benefit from a more systematic understanding of where most screen reader users struggle. Respondents were also asked to select their top 3 most problematic items from a list. The order and difficulty of these items has remained constant over all 10 surveys that WebAIM has conducted over the past 14 years.
While addressing these items would not render a website fully accessible, you can make significant progress in removing large accessibility barriers and annoyances for users. The good news is that many of these items (such as adding correct alt text, updating headings or reducing the number of navigation items) can be addressed easily without necessarily requiring any code updates. Other items may depend more on development updates but are global issues that can be fixed in one place for the entire site. Review the ranked list below and consider which might apply to your site:
- CAPTCHA
- Interactive elements like menus, tabs and dialogs not behaving as expected
- Links or buttons that don’t make sense
- Screens (or parts of screens) that unexpectedly change
- Lack of keyboard accessibility
- Images with missing or improper descriptions (alt text)
- Complex or difficult forms
- Missing or improper headings
- Too many links of navigation items
- Complex data tables
- Inaccessible or missing search functionality
- Lack of “skip to main content” or “skip navigation” links
Need more help?
Access Armada specializes in building custom accessibility testing plans tailored to your requirements, development processes, and testing budgets.
