Rough and ready web site usability testing

Wrong, wrong, wrong.

Sure, a basic usability test performed by experts in a dedicated usability lab can cost between US$5,000 and US$50,000. No sweat for large corporate and government bodies: they should include usability tests in the budget from the very beginning. Without blinking. (Still small voice: "But do they? Even in 2008, do they?")

For small and medium-sized enterprises, however, the cost is a worry, plus we imagine we need a PhD in the subject to even tackle testing. In the immortal words of Scarlett O'Hara, "Why, I hadn't thought about that yet. I'll, I'll think about that tomorrow."

Tomorrow is not too late. Some serious problems on a web site can be fixed fairly cheaply and easily, even after launch. But unless you test, you'll never know how people attempt to use your site, where they fail, and therefore what needs fixing.

In short, any usability testing is 100 per cent better than no testing. Duh.

Top of page

You can do it
The king of usability, Jakob Nielsen, said some cheering words in 1998:

Even though experts can do the work more efficiently (and usually with better results), it is encouraging that utter beginners could complete a full Web usability project in less than one week. This truly proves that "limited budget" and "lack of time" are not valid excuses for inflicting difficult sites on your users.
http://www.useit.com/alertbox/980503.html

Hey, what's this "less than one week?" You can do some damage to your web site faults in just a few hours.

Top of page

Usability — what's that?
Here are some words of wisdom from Bruce Tognazzini about the first principles of usable web site interfaces.

Effective interfaces are visually apparent and forgiving, instilling in their users a sense of control. Users quickly see the breadth of their options, grasp how to achieve their goals, and do their work. [...] Effective applications and services perform a maximum of work, while requiring a minimum of information from users.
http://www.asktog.com/basics/firstPrinciples.html

Top of page

Why test and test and test?
Because some problems are so serious that they prevent your users finding other problems. Fix problem A, then you will be able to perceive and (we hope) fix problems B and C.

The more extensive and elaborate the test, the more likely you are second-guessing your web site users. If you already know all the problems, why bother testing? A more simple, open-ended approach may get better results.

A series of small do-it-yourself usability tests should reveal most of the latent catastrophes and user-confusers on your web site. Things that are broken, hidden or contradictory. Problems that slow users down to the point of frustration.

Only testing will sort out which are issues of usability, and which are matters of opinion. While your web team sits around the table in glorious isolation, opinions and feelings (not facts) are the currency. When opinions clash, the CEO's opinion, no matter how stupid, will usually prevail.

Get out of the boardroom and find out what's really important. Watch outsiders use the site. Watch and wonder. Then fix. Then watch your test users again. Then fix. Then watch them again... Whew. Finally they get the point that you thought was blindingly obvious. Finally they get what they want from your site. Finally they appreciate its glory.

That's all there is to it.

Top of page

Rough and ready (sweet and simple) think-aloud tests
The goal is to observe and record a few real users doing real tasks and thinking aloud all the way.

1. Find 5 'real users' and contract them to be your testers.
2. Set up a quiet room where they can explore the web site, one at a time. Allow about an hour with each user.
3. Welcome and thank them, explain the process, ask them to look at the home page and say what they're thinking.
4. Suggest a task they might do, but let them choose. Example: find and order a product -- they choose the product.
5. Watch them do the task, while you observe and sometimes question them.
6. Analyse the web site faults revealed by the user, and prioritise them for fixing.

Note that 'real users' of your web site are almost anybody except your company's employees. Ideally, testers are the target customers of the web site. But hey, if you can't find the ideal group of 'real users' for every test -- no worries. Use your neighbour, commandeer some students, drag people off the street. Reward them for the hour they will spend sitting at your computer, trying to figure out how your web site works.

Top of page

Rock bottom basics of testing
See a photo of a typical lab if you want to get frightfully serious about usability testing. Interesting stuff.

But the lab could just be you with a pen and clipboard, quietly watching and questioning one person as they explore the web site on your computer. There's even value in showing an outsider a colour printout of the home page: can they figure out what the page is all about? What would they click on?

Usability testing is an imperative -- but not an all-or-nothing imperative. It is far, far (to the nth degree) better to attempt a short, modest, simple, amateurish, pen-and-paper test than to ignore this incredibly important obligation. And you're certain to get some big surprises.

Don't be intimidated by test mythology or technology. Rough-and-ready beats head-in-the-sand every time. Good luck!

Top of page

P.S. use a checklist first
Way before any user testing, designers should do their best to check that the site follows the basic principles of usability. Every site is different, but, er... for example:

Can people tell whose site this is?
Can people tell what the site is for?
Can people find what they are looking for?
Can people achieve what they want to?
Do all the links work?
Does every page link to home?
Is the code correct?
Is downloading fast enough?
Is the site accessible?
Is the text legible, even to poor-sighted or colour blind users?
Is the text easy to skim-read?
Are menu names succinct and accurate?
Are there several options for navigation?
Is the company's identity obvious on every page?
Does every page have an obvious location: can people see what page, section and area of what site they are on?
Do all the forms work?
Can users reverse their actions?

O.K. Now test.

Top of page

(c) 2008 Rachel McAlpine Copy or quote if you like but acknowledge the author and web site please.