One stop shop for free articles & web content
88 articles on writing web content and intranet content
The Hey, You! user-test in action
When I talk usability testing, I often advocate the very simplest, on-the-run, repeated little tests, performed by somebody outside the loop of site owner, designer, and expert consultants.
This type of test is just one item in the complex process of developing a web site. But if I stress this modest type of test, rather than inhouse usability tests or testing by experts, I have my reasons.
First, I'm especially conscious of small business people and do-it-yourselfers. People with small budgets need to know there's something significant they can do for peanuts.
Secondly, I notice that big organisations often sincerely believe that once they've got a good focus group and a good designer and a professional to perform a site critique, that's all they need.
Or they skip the tiny, iterative, "hey you!" user tests, and go straight to the mammoth professional usability testing company with its mammoth invoice. (There's a place for major usability tests, obviously.)
I have just learned first-hand a great deal about this very basic form of usability testing: the process, the psychology and the results.
It was almost ready: my training CD-Rom, most unoriginally titled "Quality Web Content". (Now out of print.)
But hey — haven't I been urging you to do usability tests on web sites? And this training tool is structured exactly like a web site on a CD-Rom.
The fearful designer
Suddenly I understood why people do *not* do user tests. Why they "forget", or postpone it, or convince themselves it's not necessary.
Hadn't I been using this thing myself for a year? Hadn't I checked and double-checked everything in sight? What could they possibly tell me, these outsiders?
I'd grown to dread working on this darn thing. I had been making it for over a year. Alone, and with sympathy from only my two trusty helpers, Dale Copeland and Miraz Jordan. Without feedback, because every time I answered the question "What are you working on these days?" everyone abruptly remembered a chiropractor appointment. (They'd rather I said "a novel".)
Would my bÍte noir turn out to be a white elephant? User testing would also be a Rachel-test, or so it seemed. Very threatening.
Strangely, I never imagined other people would find it hard to navigate. It just couldn't be simpler, I thought. More fool me. (See "Foolish assumptions" below.)
But I did have secret fears. I was afraid:
The latter seemed horribly likely. Long ago, the slideshows and tutorials had seemed uncomfortably slow to me, until I copied them on to my desk top. By now, all the images were cached on my machine, so the thing ran swiftly — and I could not give the CD-Rom a true speed test.
Do it anyway
But I'm a vicar's daughter, so I did the decent thing. I grabbed two neighbours, Laura and Graham (names changed to protect their sensitivity), and asked them each to do two hours' testing for $50 an hour. It seemed a pitiful incentive for staying indoors on a sunny weekend, or sitting at a computer instead of blobbing out at night.
Their brief was to use the CD-Rom any way they liked, and comment on places where they got stuck, or lost, or confused.
This was only a semi-usability test because I used a cheat's method: I didn't observe. Laura and Graham just jotted down their own notes. They preferred to work alone anyway, and Graham took the CD-Rom off to the beach for a weekend.
The vulnerable user-tester
Laura is an illustrator who has recently moved into computer graphics, and Graham is a scientist. Both are competent, intelligent, and educated. Graham doesn't use the Web much. That's great, because some of my customers are experienced with the Web, and some are not.
Laura and Graham fitted the typical user-tester psychology: in this role, they became unexpectedly lacking in confidence, apologetic, ashamed of their "ignorance". They assumed it was their fault when they misunderstood something, although clearly it was my fault.
Results: the good f-words
The first thing my users said was, it's very fast. Phew! That is largely due to the use of cascading style sheets. Some of the 400 genuine sample web pages on the CD-Rom may be slow, but my tutorials and slideshows zip along like Cassidy, so they say.
After hearing that, I could take any criticism on the chin. But first Laura and Graham had more f-words: they said the CD-Rom was fun, fascinating, and full of variety. Good. I thought so too. (Phew again.)
Major fixable flaw: structure
Lisa was puzzled by the way you could just plunge into the CD-Rom. She wanted more guidance about what to do first. Sure, there were odd pages giving help and guidance. But they were in the fourth and last column of the home page, and labeled "Extras".
Light dawned: I should simply rearrange the columns on the home page, change the last column name from "Extras" to "Basics" and place it first.
I did that, and slightly rewrote some of the help files. Simple.
Why had I positioned the slideshows first? Through an historical accident: the CD-Rom grew from my original slideshows. My journey started with slideshows. Therefore the CD-Rom started with slideshows.
This is absurd, because the slideshows are for teachers and trainers - experts. The slides can be useful as revision, but they are not self-explanatory: they need a commentary. For most of my target audience, the slides are an afterthought, so now they're in the last column.
A similar problem is very, very common on the Web. What is placed first on your web site? The first item should be what's first priority for your target audience.
Other easily fixable problems
Laura and Graham picked up broken links and missing images. The images were visible on my own screen, but were named jpgs when they were really gifs. Simply fixed.
On the slideshows, some links were enigmatically labeled — again, because I made the slides for my own use. Easily fixed.
Graham pointed out something glaringly obvious: my tutorials had radio buttons that were unnecessary, because whichever button you selected, the answer pop-up was the same.
Again, history is to blame. I began the tutorials intending to provide a different answer according to which button people chose. As I progressed I thought it better to provide an inclusive answer. But whoops, I left the buttons there.
I removed almost all the radio buttons. Now the tutorials are clearer and better-looking.
Foolish assumptions about navigation
We quickly forget our own previous ignorance. We foolishly assume that everyone else knows what we know. That's why we need user-testers.
Neither Laura nor Graham realised that the top logo was always a link to the home page. They thrashed their back buttons. I was shocked — at myself, for making assumptions, and for not offering an alternative route home.
I assumed that anyone would know that a list of words in a narrow left-hand column was a menu of links. But to Lisa, that wasn't clear, so again, she used the Back button. The simple fix was to insert the word TUTORIALS at the top of the list of links.
The test underlined the rule of keeping the Back button functional: thank goodness I never decided that the 400 sample web pages should open in new windows!
More odds and ends
I won't bore you with the details, but as I fixed the problems found by my users, I inevitably found other spots where I could simply make things better. So I did.
Stop here - for now
One problem Graham found was that in some but not all of the tutorials, the text was too wide for his screen.
I'd noticed this inconsistency before, and thought, "bother, too tricky, not that important, do it later." As you do.
I reckon if I study the code and style sheet again and again and again, the problem and the solution will become clear. Meanwhile, it won't worry most people. (I hope.)
Laura wanted to know, what was the Glossary about? I decided that didn't matter, as long as people could find it with one click.
The value to me and you
To me, the test was invaluable. I was stumbling in the dark, making this thing without any feedback from outsiders. My user-testers found:
The test did more: it freed me from my fear of finishing. Know what I mean?
Now I know the CD-Rom is usable, interesting and -- well, it's a goer, so why not let it go?
For you, the potential user, the test eliminates many moments of puzzlement and confusion. And your Back button will last longer.
The usability test represented a tiny fraction of the total work, but made a great difference to the product.