Alex Wright

Alex Wright

Alex Wright is Phoenix Pop's VP of User Experience. His somewhat intimidating resume is peppered with references to such ivies as Harvard and Brown, not to mention his role as the chief UX guy at (a minor two-million page site) where he helped put on the Kasparov-Deep Blue showdown webcast. Especially interesting to me is that he's also a librarian by training, but I resisted the desire to talk Ranganathan and stuck to practical stuff instead: tools for information architects.
-- Lou Rosenfeld

Lou: Alex, sometimes the tools discussion on SIGIA-L really flummoxes me: by tools, it seems we only mean applications for developing blueprints and wireframes. Which is all well and good, but shouldn't we also be talking about the tools that directly impact our architectures, like search engines and content management applications? Why are we so fixated on what we should be using to draw instead of what site users or owners will actually interact with?

Alex: Well, many of us do spend a lot of time "drawing" in one way or another — diagrams, process flows, site maps, prototypes, or just scrawling on white boards. Visual communication is a big part of what IAs do. But there's no shortage of good drawing tools out there (Visio and Illustrator are probably the most commonly used in the IA world).

What we really need are tools to do the rest of our jobs, mostly involving the "semantic" layer of things — working with metadata, navigation and menuing systems, naming conventions, creating data dictionaries and thesauri — and tying all those layers of information together into the final realization of a product.

Lou: The concept of a "semantic" layer is very interesting because it addresses how an architecture conveys meaning. Can tools really help us design and develop this layer? It seems that they're more appropriate for everything but the messy ambiguous stuff you might find in the semantic layer.

Alex: Well, no tool is going to supplant the intellectual work involved in crafting an effective IA; but I can envision expert tools that would help. Imagine a tool that would let you structure a simple classification scheme or thesaurus, set up and manage a controlled vocabulary, then generate indexes and metatags based on it. A tool like that wouldn't necessarily ensure that the architecture "conveys meaning," but it would certainly make the whole process a lot more manageable.

Lou: Getting back to our tool fixation, how much of this desire for professional information architecture tools has to do with insecurity over what our deliverables look like? We're a new profession; are we afraid that graphic designers, producers, and, God forbid, clients won't take us seriously if we hand them pencil-drawn diagrams? Or is the real crisis that the current generation of tools won't allow us to communicate effectively?

Alex: Good question. I suspect that many of us may harbor a latent insecurity complex when it comes to Design (with a capital D). After all, the Web is a highly visual medium — yet precious few IAs have professional training in graphic design. But I firmly believe it's all Design in the larger sense, and that we should absolutely care about presentation quality. There's a lot we can learn from our Design colleagues about things like proportion, use of negative space, and basic lay-out and palette considerations … and most of all, about how to communicate ideas visually.

While I'm certainly not suggesting that IAs should start making pretenses at graphic design, I do find myself disappointed with the low standards of presentation quality that seem to prevail in the IA world. Which isn't to say I'm opposed to pencil drawings — some of the best designers I've worked with regularly use pencil sketches with clients.

Lou: So what tools are you actually finding to be helpful (and why)?

Alex: Lately we've been using Adobe InDesign. Even though it was originally targeted at the print designers' market as a "Quark killer," it has quite a few nifty features we've been able to adapt for use in creating project documents and working prototypes. InDesign lets you create templates and master documents using a consistent style that can be applied to large documents and deliverables. You can also create object libraries for commonly used graphics, content elements, and other reusable assets. It has a nice layering feature that lets you create multiple views of a project for different audiences — e.g., you can create a "zoomable" site map for client reviews, create simple wireframes for designers, add functional specs for engineers, implementation notes for production specialists, and then overlay the whole thing with the final visual design. It's really a very elegant tool, though of course we're using it for purposes it was never intended for …

Lou: How does the whole hew and cry over using a standard notation for IA play into the tools discussion? Does it mean that we'll need to continue using flexible and generic tools (a la Visio) until a standard is accepted?

Alex: I can't say I'm holding out much hope for standardization any time soon. Our field is too young, too anarchic, and too fuzzily delineated to lend itself to any serious attempt at standardization in the next few years. For now, voluntary best practices are the best we can hope for. Actually, I think there are some "de facto" standards evolving around common deliverables like site maps and process flow diagrams which tend to follow a simplified form of conventional software flowcharting (due in no small part to many IAs' continued reliance on Visio).

Jesse James Garrett recently put together a game attempt at a "visual vocabulary" for IAs. He's captured some common best practices for process flow diagrams and site maps, and tried to position a few simple guidelines around them. Who knows if anyone will take the time to try to adhere to something like this, but at least it's a start.

Also noteworthy is Vizbang's proposed IAML standard, an XML reference schema for 3D visualizations of "information spaces." While I'm not convinced that 3D interfaces are the wave of the future, I think they're trying to tackle the right issue: how do we develop a language for capturing the relationships between bodies of information, not just the contents of an individual Web page.

In my former life at IBM, we collaborated with Paul Kahn and Dynamic Diagrams in developing MAPA, a 3D visualization tool for mapping relationships between large bodies of information.

But the odd valiant effort notwithstanding, I'll be surprised to see any meaningful momentum towards standardization in the next few years.

Lou: So what's the Holy Grail for IA tool functionality? And when do you think it'll be here (or, at least, do you foresee tools getting better in specific ways soon)?

Alex: First, we need tools designed by IAs for IAs. All of us are continually struggling with tools that were fundamentally designed for some other purpose. And as long as we're toiling away with ad hoc solutions, we're going to waste a lot of time and effort re-inventing, re-purposing, and generally ad-libbing until someone comes up with a quality toolset designed for our profession.

The ideal IA tool for me would look something like this: WYSIWIG HTML prototyping a la Dreamweaver or NetObjects; flowcharting and site-mapping capabilities similar to Visio; tools for creating, managing, and applying metadata; presentation-layer facilities similar to those found in Adobe InDesign, especially the "layering" and asset management capabilities; and — most importantly — a vocabulary and user interface model built around the needs of IAs, not graphic designers.

Lou: So if you were investing a thousand bucks on tool vendors today, who would get what portion of your portfolio?

Alex: Macromedia 25%, Adobe 25% … hmm, maybe I'll save the rest of that money and try to build it myself …

Subscribe to our bi-weekly newsletter for notification of new ACIA articles and interviews.