New Challenges Retreat: ideas, discussion, and a call to action

Posted by
“On a global scale, there are only a handful of us, and yet our work affects millions every day.”

“We’ve got a lot of questions and not so many answers,” according to opening session speaker Peter Van Dijck as he and Jorge Arango addressed the issue of global information architecture (IA). As the IA field grows, conferences and retreats provide IAs with the opportunity to ask questions and discuss the possibilities.

This year’s IA Retreat, “New Challenges in Information Architecture,” took place at the Edith Macy Conference Center, just north of New York City, October 7-9, 2005. Of the many themes discussed at the retreat, those that stood out revolved around the challenges of enterprise information architecture (as in very large enterprises, such as government agencies, and Fortune 100’s), cross-cultural IA issues, and designing user experiences for evermore complex, and increasingly less, web-centric systems.

Peter and Jorge gave a “broad-strokes” perspective on the daunting challenges faced by cross-cultural multi-language sites, where even something as seemingly innocuous as how country/language is selected can set off a firestorm among users. In one great example, they explained that Flemish users protested loudly when a product site chose to set French as the default option if users selected Belgium as their country. The design was changed to accommodate the users.

Language and culture challenges in IA were also very much at the center of Mairi Willis’ session on designing intranets in large, distributed organizations. She described her task of deploying intranets across some 50 countries, each of which incorporated the country’s own language and cultural idiosyncrasies, all while remaining consistent with the overarching enterprise framework. Mairi told of having to mix diplomacy with sprinkles of dictatorship to provide countries as culturally and geographically disparate as Kazakhstan and Indonesia with enough freedom to ensure that locally specific needs were met without compromising enterprise-level objectives. An example of one specific issue was deploying information from headquarters to individual countries.

Dan Brown and James Melzer each took stabs at addressing enterprise-level challenges. Dan led a workshop that proposed leveraging the work of George Lakoff and the thinking behind prototype theory toward designing content management systems-which too often are more expensive than useful. He guided us through several group exercises, where workflow, roles, and content were used as the core building blocks of a CMS. Some of us felt that “context” deserved being a separate building block.

James’ session, which was based on his work for large government agencies, started with a nutshell description of Enterprise Architecture and how (surprise, surprise) user experience had mostly been left out. He presented a great diagram, which placed user experience at the center of the enterprise architecture. A discussion that simmered throughout these presentations was the absence of a formal IA theory, and how we instead find ourselves cobbling together theories from other disciplines.

Is there a need for a Ph.D. in information architecture? Is information architecture just an umbrella for other disciplines, making the need for a theory of its own moot? As Peter said, a lot of questions, not so many answers.

Marcelo Marer, Mary-Lynne Williams and Victor Lombardi took more of a strategic focus in their sessions. Marcelo and Mary-Lynne described navigating design concepts through their client’s various acquisitions of other companies, in which much of the heavy lifting in the design work, conventionally associated with the user-visible interface, took place in back-room strategy sessions with executives. They also provided a great example of high-level design “from the hip;” when, in the twelfth hour of this enterprise-wide effort they decided to discard the strategy they’d worked so hard to sell to stakeholders and instead adopt the infrastructure already in place at a newly acquired company. It was a gutsy, but ultimately, successful decision, and a great testament to the power of being open to change course, even after a fundamental design choice has been made.

Victor talked about how advertising space and value on a page-level basis can affect page design. He showed methods for calculating the value of an ad, and discussed how size and placement of the ad on the page are key factors. This led to a discussion of the trade-offs between choosing content quantity on a page and how difficult choices need to be made by stakeholders regarding usability versus income. The focus of IA is often on the structure of the content, while the big swatches of ad space, which have a direct impact on usability and usefulness (and in some cases the ads turn out to be more useful than the content), are easily forgotten. Todd Warfel pointed out that navigation and orientation are also critical factors when incorporating ad space, describing how users sometimes click on an ad link and are not aware until several pages deep in the advertiser’s site that they no longer are in the originally visited site.

As part of his presentation, Todd outlined methods for improving IA deliverables, including how to leverage grid systems. Using two different newspaper layouts with similar amounts of content as a discussion springboard, Todd showed how the one that made better use of a grid had a cleaner and more relaxed look. He explained that it improved usability during page transitions, since only content that is unique to the new page moves or changes.

Todd also showed how pattern libraries can increase productivity both for IAs as well as developers, who are keen to leverage the same code base across multiple solutions. Reminiscent of the Yahoo! pattern library presentation at this year’s IA Summit (March 2005), Todd showed several examples of solution reuse across multiple designs. As with wireframes, the ins and outs of sitemaps are a matter of continual discussion among IAs. Into that contentious mix, Todd added his hub-based model, which he described as more closely matching the user’s mental model of the site. In contrast to the tree-based approach, page clusters with user persona information associated with them radiate from a central homepage. Todd showed how the model works for content-heavy as well as interactive sites, and that many of the clients he had shown it to understood the model without need for introduction. He noted that the hub-model may be more work to maintain than other models. A lot of discussion emerged surrounding Todd’s concepts, such as questions about the role of IAs in defining page layout, or the usefulness of sitemaps. At the same time, there was agreement that there is a dire need for more innovative ideas, such as Todd’s work, and Adam Greenfield certainly brought that message home with his “Everyware” presentation.

“From the complexities of designing across cultures and languages to managing information in massive organizations, the challenges facing information architects seem to be growing more daunting every time we look around to assess them.”Due to scheduling conflicts, Adam was only able to join the retreat at very end, but as it turned out this was the perfect closing session for the retreat. Adam’s presentation was both a detailed depiction of a ubicomped world in which we’ll find ourselves in an impossibly near future, as well as a call to action for IAs to apply their skills to the mind-boggling challenges of designing user experiences for that future (which, to paraphrase William Gibson, is already here, just not evenly distributed.)

Very methodically and with sometimes frightening detail, Adam described a technological eco-system in which the web as we know it vanishes into the pervasive ether, in which privacy and ethics become paramount factors, and systems become so complex that attempting to design them at the atomic level may simply not be feasible. Instead, we may find ourselves functioning like gardeners tending plants, guiding and overseeing but not controlling detailed behavior.

As the leader of the committee organizing this event, I can say that a major motivating factor for making the retreat happen was the sense of a need for more events and opportunities for IAs to come together and discuss our work. Peter Merholz, president of the IA Institute, has raised similar sentiments about the need for more events. Too many IAs work in isolation from other IAs. On a global scale, there are only a handful of us, and yet our work affects millions every day. I know that efforts are underway to organize a retreat in the Chilean Andes sometime next year, and I hope to learn about other similar efforts. The catalyst that set me off on the path to make this event happen was Christina Wodtke’s call to attendees of last year’s amazing “Future of IA” retreat to make retreats happen in our local areas.

I can only hope that this year’s retreat will serve as an inspiration for others to create similar opportunities for IAs to come together, share and vet ideas, and continue evolving as we take on the challenges that lie ahead.

More retreat resourcesRetreat event page
Retreat wiki (where Chiara Fox and others added session notes, links to presentations, and more)
Flickr stream
Podcasts of the sessions will be posted soon…

Anders Ramsay organized this year’s IA Retreat. He is an information architect in New York City, as well as founder of New York City IA Meetups. His personal site can be found at: