IA Summit 2006: Gathering of the Tribe

Take-Home Messages

Because my interests are eclectic, I attended talks and seminars covering most of the major themes of the conference. A couple of identity-related themes for IAs permeated many of the talks:

• expanding the influence of IAs—which I could generalize to expanding the influence of user experience professionals
• IAs as storytellers and keepers of the vision

Here are some nuggets of wisdom from many different speakers—not just food for thought, but practical information I can use on my projects right now. One caveat: Because so many excellent talks happened at the same time, I was unable to attend some important topics with outstanding speakers. I look forward to learning about what I missed.

How IAs Can Influence Business Strategy

Many speakers touched on the topic of IAs and business, even if that was not the primary subject of their talk.

What the Speakers Said

The Management Innovation Group (MIG) pre-conference seminar, “Enhancing the Strategic Influence of IA,” which Victor Lombardi, John Zapolski, shown in Figure 3, Scott Hirsch, Harry Max, and Mark McCormick presented, was a gold mine of new ideas, along with ideas I’ve heard before, but needed to hear again.

A few of the key messages from the MIG seminar:

• To provide value at higher levels in a company, IAs should help executives reframe and understand their companies’ strategic challenges. As keepers of the vision, IAs can tell stories, reframe business problems, and graphically visualize a company’s products or goals. For example, IAs might come up with one diagram that says it all.
• In this light, we can see the modeling of business processes as a strategic IA role. IAs and interaction designers don’t just design Web-application user interfaces; we design work. Or as the panelists in the session “Enterprise IA: Not Just for the Web Any More” told us, we don’t just design Web sites; we design the processes that create them.
• When engaging in strategic decision-making, we should remember that the value of an IA is not telling executives how to run their businesses. Rather, the IA’s value is in deeply understanding the business, translating business issues for various stakeholders, and helping executives see different business opportunities.
• We should all be able to explain how our work affects the overall profitability of our companies. Imagine giving an elevator speech to a CFO—you’d need to explain your value proposition.
• The value proposition for an IA might go something like this: “To help [person or company] to achieve [goal], as an information architect, I [purpose]. Unlike other information architects, I [differentiator]. I do this by [process].”
• Over the years, there’s been a lot of discussion about the financial return on investment (ROI) of UX work. But ROI can be a red herring—that is, people may be focusing on the wrong issue. In general, there is no consistent numerical ROI for UX activities. The ROI depends on the project, the product, and the organization. Also, there are usually too many variables to claim that a product succeeds purely because of its user experience. Thus, ROI matters more as a process or a journey.

A slide from Scott Hirsch’s presentation, “Your Role As Change Agent,” appears in Figure 4.

Building on what to do to entwine IA with business, Samantha Starmer’s “Selling IA” showed us how to engage with executives. She recommended that, when speaking with executives, we match our presentations to the executives’ style: police cop (needs to control), visionary (wants to get excited about the project), scientist (wants to analyze details), or therapist (desires consensus and good feelings).

My Take

The ability to speak the language of business and finance is becoming more and more crucial for IAs. IAs can find common ground with executives—we have complementary skills, and we both want to create business value.

Tools, Techniques, and Deliverables

There were many practical sessions on tools, techniques, and deliverables.

What the Speakers Said

Often, what seems like an IA problem is really a management problem—or a change-management problem. In the session on Enterprise IA, we also learned that each department within a company wants to solve its own problems, without looking at the big picture across all departments. One change-agent technique is the role-playing workshop, in which the hidden agenda is encouraging empathy. Group A describes Group A’s problems. Then—and this must be a surprise—Group B must solve Group A’s problems.

In “Game Changing,”PDF Jess McMullin suggested a different group activity to help a team gain consensus on product requirements: Ask participants to design the box for the new product. The box must include a product name, tagline, a short description, long description, and so on. The team can actually lay out the box afterward in Photoshop® or Illustrator®, print it on heavy coated paper, and glue it together.

Karen Loasby described how “Metadata Games” at the BBC offer ingenious tactile ways of explaining structured-content authoring to naïve authors. For example, to communicate the concept of templated content elements in a content management system (CMS), the BBC asks its authors to create a new article for an online newspaper from pieces of existing articles—not by actually using a CMS, but by assembling laminated printouts of content elements such as titles, paragraphs, and photos into a new article. Creating the new article is literally like putting puzzle pieces together. Each piece of the puzzle is a content element that can fit only in a certain place.

Steve Mulder’s “Bringing More Science to Persona Creation” explained how to build personas that incorporate quantitative, statistically significant data, not just stories that are made up. Even so, for a persona to feel like an archetype, we must oversimplify and exaggerate the persona’s characteristics. Ignore small differences between personas; they get in the way of telling the story.

The panel discussion “Wireframes: A Comparison of Purposes, Process, and Products” offered an overview of the full range of IA deliverables and advice on when to use them. One key point was that the larger and longer the project, the more documentation there is; and the less trust that exists between team members, the more you need to write down. Panelists also noted that IA deliverables should generally be printable, because in many cases, stakeholders like to draw on them.

My Take

IAs typically create different deliverables for different purposes and different audiences.

The last few years have seen the pendulum swing away from massive, exhaustive project documentation toward fast, iterative prototyping. Depending on the project, the most productive approach may be somewhere in the middle of this continuum.

IA teams are using creative, innovative techniques for discovery and education. The challenge is to know when to use them, with whom, and how to couch innovative techniques in business terms.

Web 2.0 = Ajax + Social Networking + Collaborative Tagging

Wondering what the heck Web 2.0 is and why we should care? Several talks addressed this issue.

What the Speakers Said

Dave Heller’s “What Do Ajax, RIAs, and Web 2.0 Really Mean for IAs?” provided lucid definitions of all these terms. Dave pointed to the Yahoo! User Interface Code Library, which includes JavaScript utilities and Ajax controls. The complementary UI Design Pattern Library shows examples of auto-completion, breadcrumbs, drag and drop, and so on. Dave advised that we pay attention to metrics, security, accessibility, and performance in Ajax applications; we’ll need to work very closely with developers.

Panelists in the session “Wireframes: A Comparison of Purposes, Processes, and Products” recommended that we learn Flash for prototyping Ajax applications—though Flash is not so well-suited for printing and documentation.

In the session “Tagging and Beyond: Personal, Social, and Collaborative IA,” we learned that the people who add the most tags are a small percentage of users, but they seem to be the most social users and the most obsessive about tagging. Of course, tagging also gives us insights into how regular people categorize things. On Flickr™, the owners of photos add ninety-nine percent of the tags. People tend to add these kinds of tags:

• descriptions—blue
• categories—cat or cats
• opinions—lame
• actions—read this
• relationships—friend
• cryptic insider jokes
• spam

But like other human activities, tagging and collaborative content are part of a social context. During David Weinberger’s keynote address, audience member Rashmi Sinha pointed out that tagging reflects the social point of view of the people doing the tagging—for example, the Wikipedia entry for Hinduism reflects the Northern California view of Hinduism. Later, in her talk “Sorting in an Age of Tagging,” Rashmi told us that categorization is a matter of cultural consensus. As Donna Maurer explained in a separate talk on the same theme, this is one of the main points of George Lakoff’s book Women, Fire, & Dangerous Things.PDF

Rashmi also mentioned tag clouds, which are getting a lot of attention lately. Rashmi noted that discussing the usability of tag clouds misses the point. The question is: “Why are tag clouds popular?” Perhaps people like tag clouds because the currently popular tags bubble to the top of a cloud or because they can add tags quickly without thinking about them too much.

D. Grant Campbell and Karl Fast gave us the big picture in “From Pace Layering to Resilience Theory: the Complex Implications of Tagging for Information Architecture”.PDF From a theoretical perspective, change often occurs in a patchy, non-linear, discontinuous way. If this is true, collaborative tagging may have little impact for a long time, then suddenly have a big effect. Whether this will be a good effect or a bad effect, it is too early to say.

My Take

There was some interesting discussion about collaborative tagging—more or less what we were calling folksonomies last year. The aggregate, emergent behavior of taggers really is fascinating. For example, on some collaborative tagging sites, we might see how synonyms and related terms from hundreds of people tend to cluster around a concept. To what extent would this mimic an individual’s mental model of a canonical concept that is surrounded by related terms? How are the semantic distances different for individuals versus groups?

Many of the collaborative-tagging Web sites have a social-networking aspect. From a business perspective, social networking Web sites are growing rapidly, according to a comScore® Media Metrix study the Washington Post commissioned and published after the IA Summit. Many businesses will notice this growth and will try to capitalize on it; thus information architects need to pay attention to this trend.

Though Web 2.0 is interesting and important, I didn’t hear a lot of radically new ideas in the presentations I happened to attend. Like many others, I’m interested in seeing whether and how collaborative tagging would work in a corporate environment such as an intranet. In his keynote address, David Weinberger said something like, “We can’t get from messy folksonomies to pristine hierarchies, but that’s okay, because folksonomy is good enough.” Yet I’d like to hear practitioners’ views on what’s good enough in their corporate environments. Unfortunately, if this topic was addressed in this year’s conference, I missed it.

I also did not see any new, killer Ajax applications in the presentations I attended. (If I missed some, please let me know!) If one were designing a first Ajax or RIA (Rich Internet Application) user interface, one could learn from the IA Summit presentations. As with much of IA practice, the take-home message is: Learn how other IAs are specifying and prototyping Ajax and Rich Interaction Applications. Then modify those techniques to create your own methodology.

Content Management, Taxonomies, and Metadata

Speakers offered practical advice on content management and creating taxonomies and metadata.

What the Speakers Said

As a panelist stated in “Enterprise IA: Not Just for the Web Any More,” enterprise IAs can look at metadata and controlled vocabularies as glue. By creating a shared set of terms, we can teach sales and marketing how to talk to each other.

Panelists also noted that it’s hard to get people’s attention by talking about taxonomies and content management. What does get executives’ attention are business problems—for example, “We can’t sell our products on the Web,” because people can’t find them.

Taxonomy means different things to different people. If executives say they want a taxonomy, this does not mean they really need one.

In her “Metadata Games” talk, Karen Loasby mentioned that the BBC is using a semi-automatic CMS authoring system that suggests controlled-vocabulary keywords authors can accept or reject. Authors can also add new keywords in an ad hoc manner, which the editorial team later reviews.

Ann Rockley offered this tip in her talk on “Object-Oriented Design,” If you’re thinking about moving unstructured content into a structured format, model the content first and create templates before choosing the format, XML standard, and so on.

In “New Approaches to Managing Content,”PDF Dan Brown noted that all CMSs assume that producing content is like producing widgets in a factory—in other words, that authoring is a rigid, linear process with clear, invariant intakes and outcomes. In real life, producing content is messier and less linear. People collaborate, skip steps, and invent new processes. Is there a better metaphor than a factory?

My Take

The trend toward lightweight governance of metadata and vocabularies seems to be continuing. Using a content management system well and making sure authors are tagging properly takes time and money. In the world of CMSs, there is no magic.

Who Are IAs and What Is IA, Anyway?

Yes, we’re still navel-gazing. Though it’s annoying, this introspection is important, because we need to be able to explain IA to our clients, recruiters, and employers.

What the Speakers Said

As Richard Dalton and Dave Heller pointed out in the “Defining the Damn Thing BOF,” it may be helpful to focus on information architecture rather than on information architects.

Here’s a concept from the “Enhancing the Strategic Influence of IA” pre-conference seminar: IAs are gap-fillers. One panelist in the session “Enterprise IA: Not Just for the Web Any More” noted that, as a group, we may not have entrepreneurial personalities, but we are opportunistic in seeking out complex problems—oops, challenges—to explain. We’re generalists with special skills. We can do the tasks that people in other roles might do, but often don’t.

In a poster, Laurie Gray presented an informal survey of information architects’ Myers-Briggs personality types. The most common single personality type of the respondents was INTJ—that is, introverted, intuitive, thinking, and judging. INTJ people represent only one percent of the general population, but constitute about twenty percent of all respondents among IAs.

My Take

The two core personality traits of UX professionals are the ability to analyze and the ability to empathize. Wonder how that fits in with Myers-Briggs?

It’s not just what we know, what we create, how many books we read, or how many conferences we go to; it’s who we are that draws us to user experience work in general and to information architecture in particular. And it’s who we are that makes our clients, managers, and colleagues trust us.

But How Does It Feel?

The conference itself is, of course, an experience: overwhelming, exhausting, stimulating, boring. For some people, it feels like a high-school reunion—meeting with friends they haven’t seen for a while. For other people, it feels like there are cliques of really smart, accomplished people who all seem to know one other, but don’t really talk to anyone else.

For me, attending my third IA Summit, it still feels like a wondrous experience. Coming from a secondary market—Colorado, USA—I hadn’t known there are so many IAs in the world. Some are hipper than thou; others fade into the background. Some IAs work very differently from the way I do—with different deliverables, different design approaches, and different business goals—but other IAs do the same kind of work I do. It’s reassuring to discover that many of us face similar problems. Despite our differences, when I attend the IA Summit, it feels like we are all part of the same tribe.

My Thoughts for Next Time: IA Summit 2007, Las Vegas, Nevada, USA

To the organizers—Continue encouraging non-USA attendees. May a thousand IA flowers bloom all over the world, may this conference help them grow, and may cheap Las Vegas airfares be a boon to everyone.

To the Fasteners Committee—How about giving first-time attendees the bolts and everyone else the nuts? Experienced attendees could donate their nuts to the people holding the matching bolts. That way, first-time attendees could feel like they are getting something, not giving something away.

To the Program Committee—Limit the number of talks by any one presenter to, say, two. Try to recruit more speakers who are implementing collaborative tagging and/or social networking on corporate Web sites.

To presenters—Please, for the love of all that is holy, do not read verbatim from your PowerPoint® slides. Please…

To nursing mothers—Bring your babies—whether you’re a celebrity IA or not. New IAs need to come from somewhere, don’t they? A new parent could write a paper some year: “Incipient IA Skills in Toddlers: Nature and Nurture.”

For introverts like me—We’ll get a lot more out of this conference if we actually speak to people in our physical vicinity. Dislike small talk? Fake it! That’s what I do. Thanks to IA Summit conversations, I’ve learned an incredible amount from some fascinating people and even got a job offer.

To people who feel out of place—Volunteer with ASIS&T or the professional group of your choice. It works.

Rat pack wannabes—While you immerse yourselves in Las Vegas’ slick shtick during conference breaks, I’ll be hitting the trails a few miles outside of town. I’ve heard there are some red rock hills to climb. Have a martini on me, though. 

Photos by Pabini Gabriel-Petit, Andrew Hinton, and Kyle Pero