Usability - Lessons Learned From A Redesign

November 05, 2003 | 2 Comments

This is the second in a series of “Lessons Learned” posts regarding the seattlechildrens.org redesign efforts. The first covered development process and Web standards.

Before I get into the lessons learned from out usability efforts I want to point out a few things.

First, we’re still learning how best to conduct our usability sessions and with every on we learn quite a bit, not only about our users, but about how best to go about our usability efforts. While I’ve got quite a bit of experience with various forms of usability testing as well as user research, I’m by no means a usability guru — I’m just a Web designer and developer who happens to think that users need to come first and be advocated for with every design decision.

That brings me to my second point. If you are involved in designing, developing, writing for or managing Web sites or applications you owe it to yourself to get involved in some kind of usability process. It’s a real eye opening experience. No matter how much you feel you’ve mastered you craft, or how much you know you “get” the Web, watching a few real people use your sites will teach you loads and could really open your eyes to some things you’d never before noticed.

With that said — on to the lessons learned.

Why Usability?

This seems like an obvious question with an obvious answer, until you think about how uncommon usability really is. There are many Web designers, developers and teams that don’t have the time, money, or know-how (or think they don’t) to bother with usability.

We decided that even though we couldn’t commit the type of time, money and effort we’d need to get do it really right, we could do quite a bit of “Gorilla” usability and make as much effort as possible to get to know our users, understand their needs and set goals that helped satisfy those needs.

We do some sort of usability on the majority of project we work on, large and small. For the redesign of our external site we felt it was very important to devote as much time and effort as possible to user research and user testing as we could

Getting to know our users

The first thing we did was conduct simple user feedback and testing sessions with various audience groups. These took several forms, tailored to the type of audience we were talking with.

For example, with our community referral doctors we conducted and online survey in which we asked them many questions about their Web use, what they found helpful, what they thought of our existing site and things we could do to help them. We also asked them to volunteer for more in-depth sessions at their convenience. This was important as they are busy folks.

We then set up one-on-one meetings with them where we drove out to their clinics and ran them through various tasks on the site, as well as talked to them about their particular situations and gathered feedback. This was very helpful, we learned quite a bit, for example: many clinics shared one computer, and one didn’t have access to a printer and used legacy photocopied referral forms. These were things we didn’t know and were trying to solve via the Web, and things we’d have never know had we not taken the time to go out to the clinics and talk to them.

With our parent and family group we brought them in and had forums where we got to know them, gathered feedback and ran findability tests. We watched them use our existing site as well as some related sites and then feedback what they felt was hard or easy as well as what they liked and disliked. We also ran them though Bipolar Emotional Response tests to develop a matrix of likes and dislikes. That was fun and informative.

Here are some forms we used in these sessions, in Word format. The first is our script (Word doc), complete with findability tasks given to our users and the second is a generic feedback form (Word doc) we had the participants fill out.

All of this information was gathered and turned into goals, requirements documents and mulled over and discussed in all the various meetings we had for the redesign.

Personas

Once these initial sessions were completed, we documented our findings and created personas to which we referred to when working on the design and information architecture of the site. Conceivably these personas could have been used in developing our tone and style and in marketing as well.

We made these findings available to everyone else involved with the hope that we could shape everything we did to our users needs. They proved invaluable at times when discussing functionality of the site, or making a decision about the Web sites features or structure.

Some more on personas:

An ongoing process

We conducted various feedback sessions throughout the process of the redesign. We tested our IA extensively with every iteration and these tests resulted in further refinement of a very complicated IA as well as many design and functionality decisions that ultimately helped our users get the most out of our site.

The most recent session was conducted a week ago, three weeks or so after our launch. We held a feedback session with a new parent group and we focused on the redesigned site to check our previous work and help us define goals for ongoing progressive enhancement.

For the most part the site passed with flying colors, however, we learned quite a few more things we can take back and use to make our site even better. Usability is an on going process, you never stop. As I’ve said, there is no such thing as the perfect Web site. There is always something you can do better.

Some common observations

While I can’t share (for obvious privacy reasons) any specifics of our findings with you, I can share some general observations from the sum of all of our sessions. Keep in mind the type of site we are testing (an information heavy hospital site) as that is very relevant.

Quick access to information was important. A speedy download and clear navigation structure was very important.
Our users are busy and don’t have much time for the Web
They do go to the Web for information as well as other things, shopping and email being the highest hit.
A nice look and feel was a bonus, but not important (to the users).
Accessiblity was important.
Current, relevant content (and lots of it) was high on the list of priorities.
The users didn’t like anything too “flashy” — heavy graphics was a big no-no.
They preferred a “professional” look.
There are still quite a few folks on dial-up.
Our users tend to browse first, search as a last resort.
Overall they seem pretty savvy and know, in general, the conventions of the Web
They don’t like spam or pop-ups.
Used the back button quite a bit to get to the homepage, didn’t touch the logo or “home” very much
They did lots of “pogo-ing” — going from one page, then back, then to another, then back.
The did much better with unique item searching as opposed to when they weren’t sure about a term

This list could go on, but you get the idea. There are obviously quite a few specific issues that our user research uncovered over time. You really learn a lot about how people interact with the Web in these tests, even if you don’t go about them in a scientific way. Simple observation can reveal all sorts of things you either didn’t know, or thought you knew.

Again, I’ll reiterate that if you work on Web sites or Web applications in just about any shape or from you really owe it to yourself to do some kind of user research with real users. You’ll be happy you did and I’m sure your work will be better for it.

Filed under: IA and Usability

Comments

1. TOOLman said:

Very interesting! The observations mostly coincide with my “gut feeling,” except that I was somewhat surprised that most people prefer to browse rather than search.

It is very interesting to note that those requirements are very different from what most designers prioritize when they create a web site.

Posted on November 5, 2003 10:45 PM | #

2. jimmedee said:

At my company, we make every effort to educate the client on the need to conduct useability studies and focus groups. Too often, the client has a preconceived notion about what a “good” website looks like.

Just the other day a new client directed us to a site he really wanted to emulate in his new site redesign. It was your typical, badly designed Flash site. He really liked the pretty colors and stuff moving all around. As your focus group findings dictate, the user’s main desire is to get to the information in a timely manner and is not impressed graphics-heavy sites. A lot of clients just don’t know any better.

Posted on November 6, 2003 07:51 AM | #