Learning to haggle - moving information architecture from design to implementation

— Published June 30th, 2005  |  by David Moore

Louis Rosenfeld is a recognised expert in the fields of information architecture and user experience. Co-author of the influential book, Information Architecture for the World Wide Web (aka 'the polar bear book'), Lou has consulted for such clients as Accenture, AT&T, Caterpillar, the Centers for Disease Control, and Hewlett-Packard. Lou teaches a popular seminar on Enterprise Information Architecture in six cities annually.

DM: Clients often aren't clear on the difference between IA and usability - they just know they need help. What's the best way to make the distinction clear to them, or does it matter from their point of view?

LR: As a consultant, I find this issue to be a great way to divine just how savvy a prospect is. If they are confusing usability and information architecture, I steer the conversation away from those terms and toward getting the prospect to state their problems in their own language. If I determine that those problems are indeed ones of usability, I'll briefly explain why I'm not the right person to help, and make a referral to a usability engineer.

By doing so, I've implicitly explained the difference between the two fields. And at this particular point in our conversation, the prospect is far more open to learning the difference than they were at the beginning.

A lot of us — information architects, usability engineers, and other user experience practitioners — get hung up on terminology, and foist our labeling anxieties on an uninterested market. Better to educate prospects and clients when they're ready, which is rarely at the beginning of these discussions.

When people are ready to consider the differences, I find it useful to frame the discussion in terms of usability and findability, which is something of a friendly synonym for information architecture. You can have a site that's usable but isn't very good at helping people find information. And vice versa. I'm sure we can all think of usable sites with terrible search functionality. And we can probably come up with sites that have great labeling but putrid, unusable shopping carts. Usability and findability are not mutually exclusive, but they're not entirely interdependent either.

DM: Enterprise-wide CMS implementation seems like a classic situation for good IA work, but often IA professionals are only called in after organizations have spent huge amounts of money and have little to show for it. Why do you think that is, and is there anything we can do about it?

LR: The psychology of the Silver Bullet is a malady that's common to most enterprises. When confronted with huge and complex problems, like how to integrate content across departmental silos, it's comforting to read marketing materials that both correctly identify the problem and promise that their application will solve it. Of course, these materials are only half right; the diagnosis may be correct, but no single commercial tool can really manage and organize content, satisfy both content authors and users, and deal with the organization change and political wrangling that will necessarily be involved. That's where information architects and many, many other types of people come in, albeit too late, as you suggest.

The good news is that this is changing; organizations are getting much savvier, largely due to the pain they've recently experienced from failed enterprise application implementations. Certainly, some will throw out a failed solution and go out and acquire yet another Silver Bullet, but most enterprises aren't that rich and stupid.

One lesson from this painful school of hard knocks is to not place enterprise information projects solely in the hands of IT. Doing so sends the wrong message — that technologies drive solutions, instead of the reverse. Smart organizations are now partnering their IT staff with other departments like Marketing and Corporate Communications, as well as product managers, content authors, and customer specialists, to develop true solutions that are more than applications.

DM: Are senior executives becoming more aware of the role IA can play in a web or intranet project, or is it still a tough sell?

LR: It's happening, but slowly. There have been decades of propaganda about the Information Age, but clearly it hasn't sunk in with many senior managers, many of whom don't see their organizations' information as a strategic asset. Then again, a few years ago information architects listed "making a business case for IA" as their biggest headache; I rarely hear that concern anymore.

I'm hopeful that in a decade or two, the cream of today's information architects will be promoted to CEO positions because they're "information guys," just as today's companies often bring in new leaders because they're "numbers guys" or "process guys".

DM: If I'm responsible for running a large multi-national website that's having some problems, how should I decide whether to hire some in-house IA talent, or call in a consultant?

LR: You should hire both. You definitely already have employees that are working as information architects without realizing it. Recognize their work, formalize their roles, and bring in outside talent to train and mentor internal information architects, and to help make the case to senior managers.

DM: You mentioned recently that the search engine vendors still don't seem to think a whole lot about how their products work for end users. What features/issues would you like to see them paying more attention to?

LR: 1) Replacing "Advanced Search" interfaces, which are usually miscellaneous and almost meaningless buckets of powerful features, with interfaces that support query revision by introducing those features *when you need them*.

2) Best Bets functionality should be offered by all search engine vendors. Providing Best Bet search results for the most common search queries can greatly improve overall user experience.

3) Spell-checking of queries, as Google does. We know from search log analysis that many queries are full of typos; another great way to improve user experience.

4) Basic reporting tools to help us perform search log analysis. Our design decisions should be informed by basic data on how our users search, such as their most common search queries and the queries that are retrieving zero results.

DM: What do you think are the biggest IA mistakes being made? Are they different from the mistakes we were making a few years ago?

LR: Information architects used to focus on design. Nowadays, we have to focus as much on how to get our designs accepted and actually implemented. The politics of design acceptance are tricky, and require negotiation skills. And as a profession, we're simply not very good at horsetrading.

Information architects can prepare themselves for increasingly inevitable stakeholder negotiations by asking themselves some basic questions, such as:

• Which aspects of a design are most likely to raise the hackles of a stakeholder?
• Which of these battles are worth fighting, and how should the case be made to defend an important position?
• When a battle is lost, what consolation can we ask of the "winner"?

So often we enter political discussions assuming we'll lose. I think we'd do better to learn to haggle and be opportunistic; leaving a design meeting with something sure beats walking away empty-handed.

DM: People can be very protective of their domains. Do you have any tips for implementing IA solutions across a politicized enterprise?

LR: Start with low-cost, under the radar actions that can have a high impact. For example, creating Best Bet search results for your 50 or 100 most common search queries will probably not ruffle any feathers, and can have a large and beneficial impact for many users. Look for other ways to improve search, such as adding a spell checker, because search improvements usually don't require changes to your site's content and therefore don't create much political fallout.

Small, sneaky, yet beneficial and successful actions like these can help build the momentum required for more ambitious, better-resourced plans.

DM: We're not all working in large organizations - do the lessons learned from large enterprise IA solutions have applications for smaller organizations and sites?

LR: All organizations deal with politics, and so much of enterprise information architecture is about creating paths around political obstacles. If you think like an enterprise information architect, you'll have greater success in other types of IA projects because you'll be more politically aware and attuned than those information architects who believe that their work is all about generating wireframes and blueprints.

DM: My wife's an architect, and she gets angry when I mention information architecture, feeling it cheapens her august profession to be associated with such an ill-defined and disreputable pursuit. What can I tell her to try and change her mind, or is she right?

LR: Well, your wife should know how professionally cheapened I feel when I pass by the "architect" - designed monstrosities popping up all over the US.

Seriously, I don't really understand why some structural architects feel so threatened. Is she similarly up in arms over software architects? Should film directors picket the next convention of creative directors? Should engineers of the usability, civil, and sanitation varieties rename themselves in deference to the folks who operate trains? Sorry, it's too late; the label is already out there, and there's no going back.

It may be second nature for architects to design structures, but I think setting up this particular structure — a wall of separation — will hack apart a growing bond between different varieties of architects that could be mutually beneficial and intellectually exciting.

DM: What areas are you currently most interested in? What should we be looking out for?

Our two biggest challenges are enterprise IA, which we've already touched on, and designing information architectures for global web sites. Other fields have tackled the implications of internationalization and localization, but there are few resources to help information architects design for multi-lingual, multi-regional, and multi-cultural audiences. The world's only going to continue to shrink, so those pioneering IAs who are working on global sites should find themselves quite busy for decades to come. And the rest of us should take note.

Related Links:

• Louis Rosenfeld can be found online at - http://louisrosenfeld.com
• Buy Information Architecture for the World Wide Web by Louis Rosenfeld from Amazon.co.uk