Christina: Actually, I stumbled into taking on any sort of career later than many. I went to art school and majored in photography and video. I graduated and immediately started painting, waiting tables and traveling, which I did for nearly ten years. I was in grad school (for creative writing, oddly enough) when a friend of a friend suggested I quit waiting tables and review web sites for this new portal CNET was putting together called snap.com. What a way to discover the Internet! I reviewed 175 web sites a week. I saw the best and the worst of the Web. Pretty soon I figured I could probably make web sites and taught myself HTML.

From Snap I went to Egreetings, holding various positions in the company until I settled into Information Architect. I had been reading everything I could get my hands on for some time to try to figure out how to make a better web site. I read everything from Lynda Weinman to O'Reilly's Dynamic HTML, looking for a hint. But at this time there really wasn't much out there on advanced web design principles, so I turned to other more mature professions--including architecture, cognitive psychology, software development, graphic design, marketing and usability--trying to understand how to make the experience better for the humans who were visiting our site.

I had just finished your book, actually, and felt that information architecture was the place where my oddball mix of interests and knowledge could be combined into a cohesive whole. So when my Creative Director, Dave Rossi, asked, "What do you want to do next?" I had to say, "I want to be an information architect." Of course I spent the next hour explaining what I meant, but he was up for letting me try. Thanks, Dave!

From there I went to work with some great mentors (Maria Guidice and Kim Ladin) at Hot Studio. Finally I met Noel Franus and Gabe Zentall, and they made me an offer I couldn't resist: join their Information Architecture Group, Carbon IQ. Now I'm a partner and I couldn't be happier.

Lou: As someone with a fine arts background, how can you possibly stand how the rest of us IAs constantly beat up on designers?

Christina: If you've ever seen a wireframe, it's not too surprising there is a lot of friction between IAs and designers. Most wireframes are grayscale designed web pages. And then the IA says there is no design... Well, I can't say I'm surprised that designers are mad at us. We've stolen the fun part of the problem, and left them with nothing more than color and font selection. Not exactly satisfying.

When IAs talk about designers as if they were children to be handed coloring books and told not to color outside the lines, I get pretty angry. Design is a much older profession and it has had hundreds of years to evolve. Admittedly there are a fair number of graphic designers who came late to the game and are under-educated in their art, and don't understand that design is more than a slick-looking homepage. But if we are going to effectively work together, we have to educate each other--not malign the whole design profession. After all, designers have a lot to teach us about the very things we struggle with everyday: balancing user needs with business requirements, communicating large amounts of information clearly, and designing experiences.

Lou: And how does your arts background support your focus on information architecture?

Christina: HAH! If anything my artist background negates it. When I was a fine artist, the one thing I knew is I didn't have to think about my user. Heck, the user of my paintings might not even have been born yet--think of poor Vincent Van Gogh. The message I got when I was in school was that "thinking about user means selling out your vision." But I don't have the luxury of ignoring the end user as an IA.

Recently at the ASIS&T Summit, Peter Morville listed a bunch of jargon including "user-centered design." Unfortunately I'm afraid he is right. It's gotten to be a selling tool without much meaning behind it. People who say they do user-centered design too often just jam some usability testing on the end of their process and call it good. Usability is not QA.

At Carbon IQ we practice User-Driven Information Architecture. It's pretty simple stuff: we involve the potential end user in every stage of our design. When we gather requirements, we also interview the target audience to understand their needs, and watch them use the product to understand their desires and frustrations. We then involve users in the design process, using techniques such as rapid prototyping to make sure we are on the right track. Finally we finish with usability testing, to assure accurate labeling, clear design and messaging.

Lou: It sounds like a great idea to involve the users in the design process. Then again, isn't it dangerous? Users don't always know what's best for them: they don't know what web technologies and designs are capable of. They know what they want, which is important, but they don't always know what they need. Doesn't this cloud the picture?

Christina: Ah... yes and no? Users are not designers. Heck, when I'm a designer, I stop being a user. And vice versa. The act of creating and the act of using each require completely different mind sets.

When you practice User-Driven IA, you do not ask the users of the product what they would like, or how something should be laid out. This is asking for trouble. Instead you practice a type of scaled back ethnography, and watch users perform the tasks, asking them what they think they are doing and why. You then note their mental models of the tasks they are trying to accomplish, and how they think they will be accomplishing them, and try to craft an experience that most closely maps to that model.

Lou: Do you ever encounter resistance to this approach?

Christina: Lots of clients assume this methodology will be slow and expensive, or lacks business sense. The second concern is pretty easy to dispel. We don't practice User-Driven IA because we are philanthropists: we do it because a happy customer is a loyal customer. It's cheaper to retain a customer than acquire one. This is one of the core tenets of business. User-Driven IA helps create the kind of satisfying experience that makes both the business and the customer happy. I was at a Tufte talk, and he said the average time of a web site visit is less than the average time a web site takes to load. Why would you spend thousands of dollars on something people will spend five seconds with, then abandon? Doesn't make good business sense to me.

As for being long and expensive, well, there are lots of different ways to get users involved in your design process. You can go the long complicated route: rent a lab, fly to far off destinations to get different demographic samples and poll hundreds of people. Or you could talk to users in a room in your office--or better yet, in their office--and save the lab costs. You can choose not to do the full formal analysis and instead simply gather the major findings from the testing from each session that you can immediately apply to your design. You can test less often, applying it at focal points. You can use smaller samples. Use your imagination: there are dozens of ways to meet a deadline and still involve the people who are going to have to suffer through your design.

Lou: What are common mistakes?

Christina: Don't test with your friends or coworkers. Do hire someone who knows how to moderate and design a test. Try very hard not to do your own testing (it's very difficult to remain objective when you did the design or architecture).

I hope that everyone will move to this sort of information architecture eventually. I recommend Contextual Design: A Customer-Centered Approach to Systems Designs (by Hugh Beyer and Karen Holtzblatt) if you want a jump-start on how to make it work.

Lou: Let's change gears: you're really involved in the Bay Area Information Architecture group/cocktail hour. Many at the ASIS&T Summit wanted to know how to assemble local information architecture groups. Based on your experience, what advice can you offer?

Christina: It's a funny thing; for months I was talking about holding an information architecture cocktail hour. Peter Merholz got sick of me talking about it and said, "Would you just do it."

Peter and I invited all the IAs we knew to come to the Hot Studio offices, and it was a terrific success. We decided to hold one every month. It's gone from eight people to over a hundred, and from just IAs to the other user experience disciplines, including HCI and interface designers. My partner Noel Franus added in his nascent information architecture group and we've been going like gangbusters ever since.

My advice to anyone wanting to start a group like this? The best way to get something done is to just do it. Have a meeting. Set a date, and invite people. Then set another. Keep it up. IAs are looking to talk to each other. This is the time to start a group.

Lou: It's been exciting to watch so many local IA groups get started; thanks for sharing how you did it!