Screen Reader User Survey #5 Results

The results from our most recent Screen Reader User Survey are available at http://webaim.org/projects/screenreadersurvey5/. There was much interesting information provided by 1465 respondents this year.

A few notable or surprising items:

  • Screen reader usage on laptops (81% of respondents) now outpaces desktop computer usage (78%).
  • Windows usage continues to decrease, but it is still used by 82.8% of respondents.
  • JAWS usage as a primary screen reader remains stable at around 50%. WindowEyes saw its usage cut almost in half over the previous 19 months, from 12.3% in May 2012 to only 6.7% in January 2014. NVDA and VoiceOver continue their increases in usage.
  • Internet Explorer is the primary browser of 58.7% of respondents. Chrome is primary browser for only 2.8% of respondents.
  • Respondents with JavaScript enabled remains extremely high at 97.6%.
  • The positive perception of free or low-cost screen readers continues to increase. 74% of respondents see free or low-cost screen readers (such as NVDA or VoiceOver) as currently being viable alternatives to commercial screen readers.
  • An astounding 82% of respondents use a screen reader on a mobile device. iOS devices remain the most popular primary mobile platform with usage four times that of Android or Nokia.
  • 79% of respondents are very or somewhat comfortable with allowing web sites to detect whether they are using a screen reader, and 87% are comfortable with screen reader detection if it results in a more accessible experience. In response, Marco Zehe has authored a post about why screen reader detection on the web is a bad thing.

There is much, much more information available in the survey results.

Comments

  1. Chip Orange

    Even your summary I find offensively biased against Window-Eyes; pointing out “only” in your usage figures, while the others which have a similarly low usage you simply stated “continued their increase”, while not giving any true figures for comparisons. I find your entire survey a farse when everything you do shows your bias.

  2. Jared Smith

    Chip-

    You’ve made a very strong statement. I think it notable that Window-Eyes has seemingly lost half its users in such a short time period. A definition of the word “only” is “relatively few of the kind”. It is commonly used to compare large and relatively smaller numbers. Surely its usage does not automatically suggest bias or that the data is farcical. The word “only” is used many times in this way in the full survey results (which contain the “true figures” you suggest we have not provided).

    Stating that only 6.7% of respondents use WE as their primarily screen reader is not farce – it’s data.

    I presume that much of your hostility (and I remember that we’ve hashed this over before on e-mail) is based on our disclaimer on the original survey that Window-Eyes users should use Browse mode or use another screen reader to complete the survey due to a long-standing issue in older versions that makes standard form fields inaccessible otherwise. Some in the WE community (of which I know you are very active) took exception to these instructions. Interestingly, several users of older versions of Window-Eyes users had issues with our very basic survey due to this issue, and some even thanked us for the Browse mode tip so they were able to complete it successfully. While you have suggested that these instructions have biased the survey against screen reader users, the comments and data suggest that it has instead allowed us to collect a more representative sample of its users.

  3. Truth

    Chip’s just being a baby because he realizes that old laggy WE isn’t the future. I mean come on. To state a survey is a farce simply because one doesn’t like the stats is something you do in middle school.

  4. Tim Connell

    Hi Jared,

    just wondering if you have given much consideration to the overall size of the market? If the market was expanding and new players entering, then a drop in market share could be happening at the same time that sales of a particular product are increasing. People can easily interpret a drop in market share as meaning solely that a particular product is in decline, whereas it could just be a reflection of an expanding market.

    Thanks,
    Tim

  5. Jared Smith

    Tim-

    Yes, this is certainly an interesting consideration. I suspect that the overall screen reader market is probably expanding, but probably not more than maybe a few percent annually. It would take notable growth in the market for the screen readers that are losing market share to not also be losing customers. For Window-Eyes, for example, to lose 1/2 it’s market share but not lose customers would require that the market itself has doubled in 19 months.

  6. Eric Damery

    Jared -

    It would be interesting to hear from GW Micro directly. They could help validate your findings.

    Since the respondents in your surveys have indicated that JAWS has dropped from 66% to 50%, your logic would suggest that JAWS customers have decreased at a significant rate (about a 30% drop since survey 1)? My data can confirm that this is not the case and I’m in a good position to observe it.

    I think that your respondents are mixing mobile with PC and there is a large sample of web developers and testers in your surveys (and growing) who may not represent true screen reader users getting an education or working a job where accessible web design and using an iPhone are not the main objective.