A year ago, I published an article hear called, “2014 In Review And Predictions For 2015 in which I discussed the different articles I posted last year, the different numbers of hits different articles received, I made some entirely unscientific conclusions derived from the statistics and made a number of predictions about what might happen in the blindness technology world this year. This article will discuss the analytics we got from Piwik over the past 24 months (the period in which I’ve been using it to track my analytics) and try to form some conclusions from these numbers with two full years of data.
I will also provide something of a scorecard regarding the predictions I made last year and make some predictions for 2016.
I also contain more specific numbers on a variety of statistics that I either didn’t mention last year or, when I did include them in the 2014 review, I did so only in summary. As this a pretty widely read blog among blind people and accessibility professionals and literally nobody else, I’m hoping that by including these numbers, others may be able to draw conclusions of their own about the habits of my demographic that help inform their decisions in the future. I certainly didn’t include all of the numbers just because I enjoy copying and pasting information from my spreadsheets into my text editor as it was a boring and time consuming process. I sure hope someone out there enjoys reading about them and finds some value in the information.
Before we get into the numbers, I’d like to thank those of you who helped me write and publish this blog in 2015. I don’t want to list a bunch of names here as I’m certain I’ll forget someone and don’t want to hurt their feelings by leaving them out and because some of the people who helped with a number of the pieces did so under the assumption that I would maintain confidentiality and, as I’ve never broken such before, I’m certainly not going to start now in a gratitude section. You know who you are and I hope you all know just how highly grateful I am to you for spending the time to help make this the most widely read independent blog in the blindness and technology space. You people rock!
This is also the reason I refer to the authors of this blog as “we” instead of just me. While it’s my name on almost all of the articles, Virtually all of them contain contributions from those whom I consider my (albeit informal) editorial team.
I’d also like to thank my friends Jackie McBride and Amanda Rush for rebuilding my server, installing a much newer version of WordPress and helping me fix up the look and feel of the site. While I had hoped to make a number of changes to the site, I was not in any hurry to do so but I got hacked badly and literally had to burn down my server and Jackie and Amanda rebuilt it from the ground up for me. I can do a bunch of stuff in WordPress but, when it comes to the server side of things, I’m an absolute novice and greatly appreciate the help these two terrific women did to get this site back up and running securely.
The 2015 Top 10
The following is a table listing the ten articles published in 2015 that received the most hits, it does not include the home page or articles published prior to this year:
|1||Who Are The Champions?||NFB||826|
|2||NVDA Remote Access||Hacking||751|
|3||What Did I Just Agree To?||AT||714|
|4||Accessibility And NFBCS||NFB||695|
|5||The Greatest Living Blind Hacker||Hacking||649|
|6||Apple’s 2015 Accessibility Rebound||Apple||577|
|7||Anarchy, Leadership and NVDA||Hacking||573|
|8||The Foggy 3rd Party Screen Reader Issue||AT||557|
|9||CSUN 2015: Traditional Leadership||AT||534|
|10||Data Breaches Plague AT Industry||Hacking||485|
The Predictions Scorecard
Before I dive into the details of the statistics we gathered this year and compare them to 2014, I thought I would start with a “predictions scorecard” listing each of the ones I made a year ago and whether and to what extent they may or may not have come true. I’m not going to repeat the predictions in detail so, if you want to read what I had written specifically, please go to the 2014 article.
- I predicted that Amazon would Fork Android Accessibility and start doing their own thing. This prediction came true when the Amazon Fire tablets released in October 2015 contain a new screen reader called VoiceView. I got my hands on one of the entry level, $50 units and, while I haven’t spent nearly enough time with it to draw any specific conclusions, I can say that VoiceView, led by Peter Korn and coded by Mark Mulcahy demonstrates improvements in the user interaction model. Simultaneously,, as described in the article published here, “Testing Android Accessibility: The Programmer’s Perspective,” they are running into problems with the poorly functioning Android speech synthesis API and, although Amazon has forked all of Android, they have not made the low level changes required to permit the screen reader to rid itself of right angle and other pathological gestures. I remind my readers, though, this is a version 1.0 of VoiceView, the team led by Peter Korn has had roughly a single year to do the work and, compared to TalkBack, it already shows some solid improvements. Point 1 for my predictive powers.
- I predicted that Google’s Accessibility would Improve Dramatically. GoogleDocs and Chrome (on Windows at least) have improved greatly and there’s a lot of chatter around the accessibility rumor mill that Google has started funding NVDA with large scale financial contributions. At the same time, in 2015, Google released an entirely new and entirely inaccessible framework and they deprecated the mostly accessible Eclipse IDE and replaced it with an entirely inaccessible one. Android accessibility improved slightly which should also be seen as a good, albeit half assed step forward. Thus, as Google had a very mixed year regarding accessibility, I’ll give my predictive powers a half point for this one, I’m up to 1.5 out of 2 possible points.
- I predicted that NVDA and VoiceOver would See Continued Marketshare Growth. Sadly, I can not give myself a point for getting it right nor can I say that this prediction was wrong, we simply do not know. For screen reader market information, like many others, I rely on the annual WebAIM survey and, when they published their latest results, I wrote and posted this blog article about their numbers. In the 2015 WebAIM results, we observed the Window-Eyes and ZoomText shot up in marketshare over the previous survey, that NVDA and VoiceOver remained more or less flat and that JAWS saw a dramatic drop in its share. What I did not know when I published that article was that AISquared, publishers of both Window-Eyes and ZoomText, ran an active campaign on its mailing lists and social media to drive its users to fill out the survey. AISquared did nothing illegal, unethical, immoral, unfair or just plain wrong; they saw a marketing opportunity and, as WebAIM is a self selecting survey, they took advantage of a publicity bump they could achieve if they could make their numbers look better. The unfortunate side effect of the AISquared campaign, though, is that this year’s WebAIM results, already difficult to use due to the nature of self-selecting surveys, was polluted badly with Window-Eyes and ZoomText being disproportionately represented as none of the other screen reader publishers ran similar campaigns of their own. This made me very sad as the article I wrote about WebAIM Survey #6 took me a lot of time and effort, largely trying to find a reason that the numbers had changed so dramatically from year to year, only to learn after I published my findings that I was working with data that AISquared had skewed badly. So, no points for or against my predictive powers on this one.
- I predicted that Samsung would Also Fork Android Accessibility as has Amazon. To date, I’ve no solid evidence that this is true so I’ll score it as an incorrect prediction, thus, the score now stands at 1.5 correct out of 3.
- I predicted that we would see small Businesses forming To Provide NVDA Technical Support. While I’ve heard rumors of such, I haven’t seen evidence of this trend actually starting. Score: 1.5 out of 4.
- I predicted that I would Publish Something Here that would get Me Slammed by some group of blind fanboys who insist on spewing bullshit no matter what the actual facts say. This did not happen, probably because I hardly wrote about Android and other religious obsessions. Score: 1.5 out of 5.
- I predicted that we would see low cost Windows tablets emerging as competition to Apple in the mobile space. While I continue to see new and relatively low cost Windows tablets hitting the market, I can’t say that I’ve seen a whole lot of blind people take up using them. Score: 1.5 out of 6.
- I predicted that “something important would happen regarding accessibility and math,” which was a largely insincere prediction as it was based entirely on insider information. I had seen the work that my good friend Sina Bahram had done with DesignScience in NVDA and knew that it would be released publicly in 2015. But, I’m an insider and get to enjoy the benefits of insider information. Score: 2.5 out of 7.
- I predicted that Apple would continue to dominate the mobile space among blind users (a truth) but, in the same prediction, I stated that I believed that its accessibility would continue to deteriorate as it did in 2014. I was absolutely wrong on the second half of this one and wrote “Apple’s 2015 Accessibility Rebound” about the remediation and improvements that we saw in their 2015 product releases. I’ll take a half point for this one so the score is 3 out of 8.
Thus, I got 38% of my predictions correct. Now, I challenge you, find year old copies of tabloids that publish psychic predictions and see if any of them came up with a .375 batting average. I’ll put my predictive skills up against John Edwards or Sylvia Brown any day.
Analyzing The 2015 Numbers
The Broad View
This year, the blog received just over 21,000 unique page hits, a drop from a hit count greater than 25,000 in 2014. The single biggest difference from year to year was, in 2014, DaringFireball, a very popular mainstream technology site, put a link to one of my articles on its home page that led to approximately 6000 hits on that single article. In 2015, I didn’t get any incoming links from major mainstream sites so didn’t enjoy a bump in my hits that such will cause.
When I look at my Piwik statistics, I use arbitrary numbers to evaluate the relative success of an article. If I get 250 hits, I consider the article having been worthy of the time I spent researching, writing and editing it; if a piece gets 500 hits, I consider it a success; if an item reaches 750, it’s a win in my mind and, when we break 1000, I consider it a big win. Among the articles that first appear here 2015, we had zero big wins, two wins, seven successes, seven that were at least worthwhile and one absolute clunker. This year, it was my home page that got the most hits, almost all coming via search engines, with about twice as many as the article that received the most attention. With the home page excluded, we had a mean number of unique hits per page published in 2015 of roughly 500 compared to a mean of 1100 published in 2014. I should note that a number of the articles posted here continue to get hits over time and all of those posted in 2014 would, therefore, have had more time online to be discovered on search engines and read after the first week it’s live, the period in which a piece will get the majority of its readers.
Why did we see our mean hit count drop year to year?
First and most obviously, none of the articles we published this year reached the big win level and only two were wins at all. In 2014, we had four articles pull down more than 1000 hits with one grabbing over 6500 to date. Next, while we published 17 new articles this year, the two that received the most hits, “Testing Android Accessibility: I Give Up” and “The Death Of Screen Reader Innovation” were published in 2014 and 2013 respectively but, in 2015, combined for roughly 3000 hits and both would have been big wins on this year’s numbers alone as they’ve been in each year since they were published. , As with my home page, it appears as if readers find both through web searches. In fact, five of the top twenty articles this year were published in previous years, a new phenomena for this blog.
What About The Median?
Interestingly, though, is that while the most big wins were published in 2014, the median hit count for articles published in 2015 was just over 525 where same for 2014 was roughly 350. Thus, we had some huge wins published in 2014 but, overall, the typical article received more readers this year than last. I was disappointed that we had so many fewer winners this year but am happy to see the median hit count growing year to year.
Where Did The Readers Come From?
This year, of all visits to the site, 48% were direct entries, 29% came from search engines and 21% came from links on other web sites. This is a substantial change from 2014 where 75% of the visitors came as direct entries and less than 15% from search engines. I suppose my SEO is getting better.
Looking at which web sites were the source of the traffic, we find that it actually correlates with the direct entries, almost all of the traffic on this web site comes from Twitter, the only place I advertise the articles when they’re published. I tend to post articles at 9AM Eastern and I can usually tell how well an article will do by the number of retweets I see in the first few hours after publication. I don’t usually advertise my pieces on FaceBook (I did twice this year) but we did get a fair amount of hits from FB when others posted links to our articles there. Otherwise, links included on other web sites came to about 15% of the referrals and I appreciate it every time a reader posts a link to the site.
Thus, we can conclude that this blog’s readership is driven almost entirely by my Twitter feed (I’ve approximately 750 followers) and those of the kind people who retweet my announcements, some of whom are among the most prominent individuals in the accessibility business.
Piwik tracks all sorts of things about the traffic on a site. Here are some of the others that I found sort of interesting:
- 78% of all visitors came from desktop devices, 15% from smartphones and 5% from tablets.
- Piwik reports that the brand and model of the devices used to access this site could not be determined in more than 80% of the cases so we’ll skip this statistic as it’s entirely useless.
- Just over 60% of visitors were using Windows, 11% Macintosh, 10% iOS, 3.5% used Android and less than 2% GNU/Linux.
- 29% of visitors used Internet Explorer, 27% used Chrome, 21% used FireFox and 21% used Safari.
- Not at all surprisingly, 87.3% of our visitors had their browser language set to English and no other language represented more than 1.5% of such.
- Just over 50% of visitors come from within the United States, the UK comes in second with 8%, Canada next with 7%, 3.3% of visitors come from India, 3% from Australia and none other represents more than 1%.
- No single city represents more than 1% of visitor locales and approximately a third could not be determined.
What Did We Discuss This Year?
In my 2014 end of year summary, I was able to divide up my articles into a handful of fairly distinct categories. In 2015, the blog was less specific and, while broad themes like “access technology” and “screen reading,” were discussed, I can’t pigeon hole articles as easily into categories like “Android” or “Apple” as I tried to discuss issues in a broader context this year, hence, the articles are harder to organize by subject.
I definitely spent more time criticizing NFB this year than in the past. The article published in 2015 that got the most hits (roughly 825) is, “Who Are The Champions,” a piece that detailed the accessibility histories of NFB conference sponsors and demonstrated the incredible hypocrisy of the organization that claims to represent our community. This article surprised me in the comments section: I received absolutely no posts defending NFB on these issues. In the past, when critical of NFB, I’d at the very least get a token number of comments from the faithful federationists but, this time, based on a lot of private conversations I’ve had with loyal NFB members, they agreed with what I had written and chose to remain silent publicly.
This year, I wrote a number of articles that one might assemble under the loose heading hacking and/or hackers. In this group, I’m including articles I had written about the community driven NVDA Remote Access project, leadership in accessibility (which dovetails with the articles critical of NFB) and a nice bio piece about my buddy Tyler Spivey. These articles all beat the 500 *success level hit count but only one ached the level of being a win. I enjoyed writing these pieces and I’m happy they were successful among readers.
The Flatter Hit Count
In 2014, we had some articles perform incredibly well while others completely tanked. Those that were successful were easily categorized; in 2015, the article with the most hits was only a couple of hundred over the median and, excepting the one true loser, the worst was only a couple of hundred hits below it. No specific category jumps out from which I can draw much of a conclusion other than, “my readers are really not fond of the NFB technology strategy.” Otherwise, most articles were successful and, as seeing my hit count is the only reward for which I write the blog, I’m happy that people are reading the articles, tweeting out the links, posting comments and coming back year after year.
In early August, I posted an article called “My Favorite Rock And Roll Cover Songs” that generated a handful of really entertaining emails from readers which rarely happens here when very few people actually read an article. The article was fun to write and, based on the emails, the people who did read it seemed to enjoy it but, damn, very few of you read that one. Maybe take the time now to click through and give it a read, I still think it’s pretty good.
Predictions for 2016
It’s time for me to pull out the bong and listen to the sounds it makes to help me with my predictive powers. As I mention at the top, I got a little more than a third of my predictions for this year correct, maybe I’ll do better this time.
- My first prediction is to double down on our buddy Peter Korn and predict we’ll see more positive steps toward solid accessibility at Amazon. While we haven’t witnessed a major accessibility turnaround at Amazon yet, we need to remember that they’ve only had a person leading accessibility for two years and that Peter spent most of that first year trying to hire a staff. Amazon has a ton of different technology products, 180,000 employees and only one Peter Korn to lead this massive remediation effort.
- I will predict that, as we did this year, we’ll see both good and horrible accessibility come out of Google. I expect they’ll continue improving accessibility in their Chrome browser (on Windows and Android at least), that GoogleDocs will get even better and that Android accessibility will improve but not by a meaningful amount. I also predict that, as they did this year with their new framework and IDE, Google will release at least two new technology products with virtually zero accessibility included.
- I predict that the recent merger between Freedom Scientific and Optilec will result in absolutely nothing of significance to end users of the new company’s products. After this latest merger, the FS CEO was quoted as saying that separately Optilec and FS took much longer to innovate than the combined new organization will. As neither Optilec nor FS has done anything innovative in the past decade, I suppose that doubling the rate of innovation at the merged business will mean that it will produce absolutely zero innovations but do so in only five years. Users will be able to enjoy nothing new or interesting in half of the time they previously had to wait. . Doubling the pace of non-innovation doesn’t impress me but you may disagree.
- We will see more community driven and funded projects delivering useful technology for people with vision impairment and other disabilities for profoundly less money and in less time than under the traditional model of developing and distributing AT to we consumers.
- The highly competitive and incredibly insular accessibility consulting industry, where there are a fair number of similar companies doing similar tasks for similar prices will begin to experience some consolidation. I’ve no insider information on this subject but I predict at least one fairly high profile merger in this industry coming soon.
- I predict that Microsoft’s accessibility will improve on both Windows and the mobile version thereof. I made this same prediction for 2015 and was very wrong but, as Windows 10 was released with so many accessibility regressions, I find it hard to believe that they won’t improve at least somewhat this year.
- I predict that Apple will maintain its large lead in out-of-the-box accessibility and will get even better in 2016. Excepting the unfortunate iOS/8 release, every iOS since the iPhone 3GS has seen accessibility improvements and, although OS X saw regressions in some releases, the latest demonstrated a lot of fixes and some nifty new accessibility features as well.
- I predict that NFB will remain entirely pathological regarding technology and will maintain its strategy of selling endorsements to reward companies that do accessibility poorly while publishing at least one article in an official NFB publication trashing a business that does it reasonably well to being the very best. I will continue to write about NFB hypocrisy in this area but I don’t think that anything will ever cause them to stop the corrupt practices of shaking down companies for large contributions with a pile of legal threats. This year, they named Google “Accessibility Champions” which defines NFB as being nothing more than a pay to play, money talks, accessibility walks organization.
- Some large corporate and other institutional accessibility teams will read the WebAIM survey results and, because it’s the only marketshare data we have in this field, add Window-Eyes to their user test suite (virtually all already use ZoomText for low vision testing) believing that it saw actual market growth and not realizing how badly the data was polluted by a single marketing campaign. Others will look at the same WebAIM survey results and, as none stand out as obvious leaders, they will change their accessibility testing strategy to one that’s entirely standards based and includes little actual user experience. Adding Window-Eyes to a test suite won’t hurt users but will add more cost and longer schedules to accessibility quality assurance engineering, thus discouraging some from doing it at all.
- The long planned “Gonz Blinkoverse” podcast will start posting episodes in the first quarter of 2016. As this is entirely under my control, I can be pretty certain I’ll get this one right.
- A new, screen reader agnostic dictation system that works with both the built-in Windows dictation utility and most if not all versions of Dragon will be released as free, libre, open source software (FLOSS) at some point in the first half of 2016. This new utility will work with JAWS, Window-Eyes and, of course, NVDA and will be funded through a crowdsourcing campaign that the team will be announcing soon.
When I wrote the article summarizing 2014 a year ago, I felt it was pretty self serving. I was writing about the performance of my own blog and didn’t think too many people other than I would care to read it. A friend with whom I shared an early draft of the article suggested I add a “predictions” section and publish the article “as is.” When it turned out that the piece was fairly popular, I thought I might make an early December tradition out of writing such and this is the second one. I hope readers get something of value out of the piece, find it entertaining in some way and maybe be able to draw some conclusions of your own. Please do share your thoughts in the comments section below as I enjoy hearing reader’s opinions, good, bad or indifferent.