United Nations Global Audit of Web Accessibility

Few leading Web sites worldwide meet basic accessibility guidelines.

Larisa Thomason

The United Nations recently commissioned a Web accessibility study that validated the anecdotal evidence people with disabilities have offered for years. Most Web sites have accessibility problems. In fact, many sites don’t even fulfill the most basic accessibility requirement: ALT text on images.

The UN study audited the accessibility of 100 leading Web sites from 20 countries. (Request an executive summary of the report.) The results were startling to many – but not to people with visual or physical problems. For years, they’ve been frustrated by inaccessible sites.

Their aggravation is understandable once you look at some of the results.

  • ALT text descriptions:93% did not provide adequate text descriptions for graphical content, causing problems for visually impaired people.
  • Poor contrast: 78% used foreground and background colour combinations with poor contrast, making it difficult for people with mild visual conditions such as colour blindness to read information.
  • Improper (or no) header tags: 89% failed to use the correct technique for conveying document structure through the use of headings, making page navigation awkward for many visually impaired people.
  • Inadequate link text: 97% used link text that did not clearly indicate the destination of the link, causing confusion for people with learning difficulties.

Ok. So some of the other findings – like using JavaScript for important functionalities – may be harder to bring into compliance. But ALT text descriptions? Header tags? Descriptive link text? Readability?

What were the designers thinking when they put these sites together? This is basic design that affects every user and costs sites money in lost sales and customer goodwill. If people can’t access your information and use your shopping cart, they sure won’t buy your products.

And, as AOL found out, they may just sue you for good measure. In 1999, the National Federation of the Blind sued AOL because, among other problems, the company’s software was incompatible with screen reader technology. The NFB withdrew the suit after AOL agreed to make changes. In October 2006, the NFB sued Target, citing several issues including the fact that the site requires the use of a mouse to make purchases.

On December 3, 2006, the UN observed an “International Day of Disabled Persons,” and described the importance of accessible technology, noting that:

Persons with disabilities are at a considerable disadvantage by not being able to access information technologies. For instance, as education becomes increasingly dependent on information technologies, not being able to access the Internet for example limits the learning potential of persons with disabilities.

Whether or not the United Nations effort actually has any benefit remains to be seen. But at least it may help focus attention on the problem – and enlighten recalcitrant designers about just how easy it is to meet basic accessibility requirements.

Usability Testing in the Real World

In his humorous 1968 novel, Heaven Help Us, author Herbert Tarr describes how a young rabbi (Rabbi Small) helps a bar mitzvah student give a stellar performance. The student is so nervous that even the sound of someone sneezing is enough to break his concentration.

Rabbi Small’s solution is to so inure the boy to noise and distraction that he can continue the service unflustered – no matter what’s happening around him:

“I rambled around the sanctuary, sneezing, coughing, rattling candy wrappers, stamping my feet as Sandy spoke. I also broke paper bags, sneaked up on the boy and yelled in his ear, though it embarrassed me to be caught throwing Silly Putty at a bar mitzvah boy while he spoke earnestly about God.”

I recall this scene whenever I conduct usability tests. Rabbi Small tried hard to create real-world conditions for the boy. It’s a good idea. You’ll get better data from your usability testers if you provide some common distractions during the test.

Think about it: how often at home or in the office do you have the luxury of peace and quiet while you work? The office environment is filled with water cooler chatter, ringing phones, and that annoying co-worker who blares polka music on her CD player. Home isn’t much better – particularly if you have pets or children. Just try to be productive with a child hanging on your leg or a cat reclining on your keyboard.

That’s the real world, and it’s something rarely reflected in a formal usability test.

It’s like those EPA fuel economy tests where nobody ever speeds, gets stuck idling in traffic, or floors the accelerator with an enthusiasm usually reserved for NASCAR drivers. Your mileage may vary because you don’t always drive as if there’s a DMV inspector in the passenger seat.

In a standard usability test though, the user knows she’s being watched. Nobody wants to look stupid in front of the observer, so the tester is motivated to concentrate on the task, go slowly, and pay attention to details.

That’s almost completely opposite of how most people actually use Web sites. As Steve Krug points out in his excellent usability manual, Don’t Make Me Think, a misconception we have about users is that: “we picture a more rational, attentive user when we’re designing pages.” Furthermore, we optimistically assume that: “users will scan the page, consider all of the available options, and choose the best one.”

After all, that’s what our eager-to-please usability testers do, right? But be honest. Is that how you use a Web site when you’re alone? No normal person has that kind of time.

I’m not suggesting that you should yell at your test subjects or throw Silly Putty at them. But a bit of background noise and a distraction or two does give the test a more realistic ambience. Arrange for your phone to ring; ask a co-worker to interrupt with a question; drop your clipboard. Or try this: cats don’t care that they’re opaque; have one stand in front of the screen at least once during the test.

Nobody uses the Web in a perfect environment, so try not to be so obsessive about creating one when you do usability tests. A more realistic testing environment will help you get better results. And those will help you design a better, more seamless user experience.

Could A Caveman Use Your Web Site?

By Larisa Thomason

My family loves the new GEICO Car Insurance commercials where the hapless caveman defends his species against ignorant assumptions that all cavemen are idiots. Apparently, we aren’t alone. News and commentary about the commercials is all over the Web:

This is great viral marketing, and we’re all envious. But the point of the campaign is something that every Web designer and Web site owner should engrave on their keyboards:

“So easy a caveman can do it!”

Yep. GEICO is making usability their value proposition – at least in this ad campaign. If only every other Web site made the same commitment to ease of use.

Unfortunately, if you start discussing “usability” and “accessibility” with many site owners and designers, their response is less than animated. You can generally count on silence, glassy-eyed stares, and gentle snores of boredom.

Actually, usability is anything but boring. It’s a vital ingredient in customer satisfaction. Review these three scenarios. Chances are, you’ve had similar experiences.

  • You’ve spent hours scouring the Web, and finally found the perfect gift for your hard-to-please significant other. Then you can’t complete the transaction because the shopping cart isn’t compatible with your browser.

    Are you bored?

  • Or what if you’ve forgotten to mail your credit card payment and it’s due right now? You go to the card Web site to pay online, but can’t read the information because the text displays in 8px type.

    Have you accomplished your task?

  • How about this situation? You subscribe to a satellite TV service and get a notice that prices are rising. So you sign into your account to see what service you have now and compare costs. But the Web site doesn’t give you access to that information. All you can do is pay your bill and order movies.
  • Does this site offer what you need?

These situations are a minor irritation for most users, but can be major obstacles to users with physical or cognitive disabilities. People using assistive technologies generally have one way to access information. If the site doesn’t accept that method, the disabled user is just out of luck.

Maybe you think: “Hey, that’s just a few people and I can blow them off.

But think back to those users who were irritated or frustrated. They have options, and one of them is to click straight over to your competitor’s site and breeze through his shopping cart to buy that special gift.

GEICO has the right idea. “Easy to use” isn’t just goal; it’s a requirement for a successful Web site. Good usability by itself can’t make your site a success. However, poor usability – by itself – can make your site a failure.

There are many Web usability resources that cover everything from basic usability and accessibility principles to legal requirements for site owners. Check them out:

Then invite that caveman over and ask him to do a usability test.