Screen reader software usage shifts on desktop and mobile

WebAim released their 2009 Screen Reader Survey last week, a follow up from last years Screen Reader survey. Very good reading it makes too but of particular interest are results around screen reader choice on the desktop and increased screen reader access on mobile.

For years it’s felt like screen reader users have mainly used IE on the desktop in combination with the major screen readers Jaws by Freedom Scientific and WindowEyes by GW Micro. It’s not that other platforms don’t support screen readers (we have Orca on Linux, VoiceOver on Mac) it’s just that IE seems to have dominated.

As such what types of content and web technologies users can and can’t access has very much been driven by what the three software vendors Microsoft, Freedom Scientific and GW Micro have supported. This has made access to the open web a bit lopsided cutting down on choice for the end user, competition and by extension innovation. SVG is an example of a web technology that has possibly suffered by not being supported by IE and in turn by Jaws and WindowEyes.

What’s interesting to see in this year’s survey is that Jaws and WindowEyes – while still the most used – have some stiff competition at snapping at their heels from open source, free screen readers (NVDA and  SAToGo ) and VoiceOver which is available with Mac:

  • JAWS 75.2%
  • Window Eyes 23.5%
  • VoiceOver 14.6%
  • System Access or System Access To Go 22.3%
  • NVDA 25.6%

While this year’s stats show little shift for Jaws and WindowEyes usage overall there is a significant leap forward for NVDA (NonVisual Desktop Access) and VoiceOver:

Of the 1121 respondents, 74% use JAWS, 23% use Window-Eyes, 8% use NVDA, and 6% use VoiceOver. While several other screen readers were reported, these were the most prominently reported.

The upsurge in VoiceOver could be explained in part by iPhone now providing VoiceOver support; all of a sudden there is a very real reason to switch to Mac if you can use a screen reader you are familiar with on both your desktop and mobile.

This could also explain the increase of screen reader users on mobile reported this year with 53% of survey respondents with disabilities confirming they use a screen reader on a mobile device. This is up from 12% last year (although last year’s survey doesn’t distinguish disabled from non-disabled users).

I wonder how much this is to do with the ‘iPhone Factor’ but also can’t help thinking that social networking has done for the mobile web what Kylie Minogue did for Agent Provocateur – everybody wants some. And for me at least 2009 feels like the year that we all sat up and paid attention to the potential of mobile for people with disabilities.

We’re still faced with one massive problem with mobile access however and that’s the lack of an open, cross platform accessibility API that mobile screen readers can hook into. On desktop we have IAccessible2, MSAA and UI Automation (amongst others) but on mobile users are tied into one platform often only supporting one browser (such as iPhone, Blackberry RIM and others) so while desktop has opened up we find ourselves in a 1990′s type impasse with users left with little room to choose on mobile. Opera works well with VoiceOver but we have no way of telling if it works on the iPhone as it’s not supported. My hope is that with more users there’ll be more momentum behind breaking this stand off and opening up the market and ultimately giving users not only choice but portability between platforms.

It’s good to also see the free, open source NVDA on the up. They’ve worked hard to include WAI-ARIA support and are becoming a key tool for web developers when testing. The Jaws demo version used to be popular among developers but with the license becoming less developer friendly NVDA is becoming a safer option.

It’s early days but the rise of VoiceOver and NVDA combined with alternative browsers such as Opera, Safari and Firefox may break the hold that screen reader giants Jaws and WindowEyes have over the market, helping to open up competition and with it how fast screen readers innovate in supporting new technologies such as HTML5, SVG and so on. This will be a win all round for both users and developers.

Update 12 November, 2009

Jared’s presentation The Legend of the Typical Screen Reader User summarises many of the points above. He also includes quotes from screen reader users at the end in response to the question “What suggestions do you have for developers/manufacturers of screen readers?” which he mentions in his comments below.

Mobile accessibility resources:

14 thoughts on “Screen reader software usage shifts on desktop and mobile

  1. As awesome and innovative as VoiceOver and NVDA are, that’s only part of the reason for the shift. The truth of it is that JAWS and Window Eyes just flat out suck! They are the Window ME of assistive technology – bloated, expensive, slow, and buggy.

    We asked survey respondents for feedback for screen reader vendors and will be publishing samples of the responses soon. In short, consumers are very frustrated with the cost, lack of responsiveness, and general suckiness of modern commercial screen readers. There’s an overall theme that if they don’t improve and decrease costs very soon, they’ll be bailing en masse for NVDA and VoiceOver. And I predict that this is precisely what a VERY large percentage of them will do in coming years – which is a primary reason why increased support and focus is needed for inexpensive, bundled, and open source screen readers.

  2. The truth of it is that JAWS and Window Eyes just flat out suck! They are the Window ME of assistive technology – bloated, expensive, slow, and buggy.

    Well said Jared – thanks for voicing what my misplaced English reserve wouldn’t quite allow me to voice!

    Looking forward to reading the survey respondents comments on screen readers too.

  3. Reminds me of Hixie’s entry about making better products being one of the prime strategies to increase market share.

    We’ve observed this before, such as better browsers challenging Internet Explorer. Also the way a few better mobiles overtook the established brands, after which those brands started copying those ideas.

    Indeed, there’s been a great deal of cross-pollination amongst web browsers and other technological products in recent years. As you say, this innovation is good for users – especially when every product starts copying the good ideas. (Tabs in web browsers, large screens on handheld devices and so on.)

  4. These results differ a lot from what I really think is happening in the UK. From the WebAIM results we are not seeing a representation one of the most used screen readers in the UK and Europe which is Hal (speech and mag – Supernova). Is this because the survey is predominantly US based? Very few people use WindowEyes in the UK for example.

    Also with my contact and experience with blind users I know very few that use NVDA. That is not of course because it is not a good product. Those that do use it are more techical people that don’t mind the voice and lack of training and support materials.

    Many people that use System Access and other free/cheap products are people at home who can’t afford the cost of JAWS or Hal. I think the reality is if you have someone else buying a product for you, through Access to Work or Education funding, it is likely you will still be buying something like JAWS, Window Eyes or Hal.

    I also think that we are always going to see a difference in screen reader use country to country.

  5. These are all valid points Sally and ones that echo what I saw anecdotally when I worked with you at RNIB. WebAim also back up your points as they flag that usage for some free screen readers are from devs and more technical people too. They also make the distinction between what users have as a primary and secondary screen reader.

    I still think that the underlying point remains (from my unscientifically proven small corner of the web!) and there is a shift and opening up within the market and as such I agree with Jared’s prediction. Perhaps we are just now touching on the tip of the iceberg.

  6. I doubt the “Iphone factor” has much to do with the leap in screenreader users uptake of mobile. sure there are some bleeding edge users trying it out but most in my experience are using commercial offerings predominantly Talks and Mobile speak. both have been around for a good 5 years but have only in the last couple started working well with mobile browsers, this couple with the growth of non visual dependant apps that run on the platforms which support mobile screenreaders is what account for the surge, e.g. twitter clients, m.facebook, mobile IM, the Email and other productivity apps integral on simbian only in the last year or so becoming decently accessible with screenreaders. oh and not forgetting that in the UK vodafone give away Talks to VI customers, or for those who choose to buy it it’s around a quarter the price of desktop packages (around £150.

  7. Some good points Sally but I think we could unpick further. really I think webaim should bundle figures for Hal and supernova, they are the same technology the latter is as you say just screenreader and magnifier bundled. similarly we shouldn’t confuse system access and system access to go. the former is a commercial package with a price tag around £200+, and the former being free but tethered to an open internet conection, and minus all of the bundled services that make system access attractive to many less technology mature users such as it’s own email client, discussion forums and directory style of pre selected content. we do have a significant user base of hal / supernova users in the UK and to a lesser degree in europe but these are largely in younger groups especially school and university students where the more attractive pricing than jaws possibly with zoomtext or magic is probably at least 1 factor in the product selection when it isn’t the user picking up the tab. in turn this group are probably also less likely to be completing online surveys for webaim, they mightn’t have the skillset yet and likely as not aren’t aware of or motivated to participate in such communities. or to put it another way as fantastic as the survey is and long may it live it probably isn’t a truly accurate picture of screenreader market share – only of trends in useage patterns in a certain sector of the audience. NVDA is innovating in proving the concept of access to technologies such as aria but I suspect there are very few users using it as their only screenreader. it’s the paradox of free – you haven’t invested cash so you don’t invest the time to learn to make use of the full potential.

  8. Another good point raised Adrian about education and demographic. I don’t think WebAim claim the survey to be the the definitive catch all write up for screen reader usage today but is, as you mentioned, and really good trend spotter. So yes, long may it last!

  9. Many companies are turning to the online survey as a tool to judge the effectiveness of their products and sales pitches through existing customers. Others use them on any site visitor, whether customer or not as a way of finding areas of their website to tweak to make more effective.

  10. Pingback: The Blind Buzz on Accessibility « The Blind Buzz

  11. Pingback: More on WebAIM’s Screenreader Survey « The BAT Channel

  12. Pingback: More on WebAIM’s Screenreader Survey « AccessTech News

  13. Pingback: More on WebAIM’s Screenreader Survey - For Learning Sake

Comments are closed.