Conga!

Ok. I know what you're thinking: "Jazzy song for a Friday morning. But what does it have to do with IA?"

Gloria Estefan's "Conga" is great way to get a jump-start on the weekend, no doubt. But dance also provides a lesson for information architecture.

I've been interested in dance since I watched old musicals on late-night movies as a kid. In fact, I wanted to grow up and dance like Fred Astaire. Now, I appreciated Ginger Roger's smooth, graceful, moves -- and her ability to do everything her partners did except in heels and backwards. But it was Fred Astaire who really awed me. And when he danced with the hat rack and appeared to dance on the ceiling in "Royal Wedding," I knew I had to learn to dance like Fred Astaire. I practiced incessantly -- once even managing to leap atop a kitchen chair back and rotate it on one leg. Once only. The wall behind the kitchen table, the chair, and my body suffered from many attempts. Outside the chorus in a high school version of Camelot, I never did dance. But I did learn something from those hours on end of trying to imitate Fred Astaire's inimitable moves. First, most things require practice in order to do them well. And second, small improvements are both important and satisfying.

Let's talk about the this second idea with regard to information architecture. It is ideal to completely identify all of an organization's information before beginning an architecture. That is, identify and then map out sources, topics, articles, audiences, audience, areas for growth, authors, SMEs, contributors, goals, etc. etc. etc. Ideal. However, unless your organization is brand new and has a very small, manageable amount of information and yet a clear idea of the organization's near-term and long-range goals and direction, it may be impossible to identify, much less gather, all of your information. Furthermore, work does not stop and therefore information flow and creation does not stop during IA exercises. So unless your organization has deep pockets and exceptional content management practices, these sorts of "forklift" IA exercises with a goal of creating a completely new information architecture while maintaining current operations are difficult. That's not to say that they are impossible, but they are impractical for most small to medium organizations.

What is more practical is to expect small, continuous improvements. (Sound familiar? Continuous improvement is the hallmark of the Kaizen movement that spread from Japan to the US via the work of W. Edwards Deming. Continuous improvement also is the goal of dancers and other creative types when they "practice, practice, practice.") Continuous improvement (CI) requires standardization first, then measurements of performance and improvements of the standardized processes. An argument against this method is that because CI always measures against the standard, it does not allow for creative leaps of inventiveness in the process or product. In information architecture, however, the new architecture is already a creative leap.

A good IA approach does not (necessarily) replicate what you have now, but provides an architecture that works: it is appropriate for your goals, objectives, current work, and future plans. The key to a successful architecture for many organizations is to build one into which your existing data can be managed as is and migrated into the architecture over time while allowing new information to be created so that it makes full use of the new architecture.

A wonderful measurement to use for success of an IA implementation is usability -- and not only for your customers or external audiences but also for the folks who contribute to your knowledge base. Wikipedia has been in the news this month as it celebrates its 5-year anniversary. Wikipedia was begun on September 7, 2004. As the History for the site says: "The early history of Wikipedia was characterized by much chaos and well-meaning strangeness." Over three million articles in more than 100 languages strong and virtually every college freshman's first search source, Wikipedia did not exist five years ago when the Fall semester began for many colleges and universities. Had Wikipedia not taken the continuous improvement approach, it would have died out with the anguished cries of college professors who cried out for students to use academic sources. (Students take heed: Wikipedia still is not a vetted source; rather, it is a good source of additional sources and for a quick overview.) This brings us to a second measurement for IA success: content integrity.

The quality, completeness, and appropriateness of content is what I mean by content integrity. And the best IA structure in the world does not guarantee content integrity. Early Wikipedia articles are the poster-children for excellent IA and poor content -- which with practice and CI became better. However, a successful IA will be able to make the best of content, fully utilizing, linking, and cross-referencing content for easy search and retrieval. The rest? Producing excellent content requires excellent writing. That's a talent that's up to the authors and editors. And I'll save that conversation for another day. Right now, I'm off to practice my conga with Gloria...