Boxes And Arrows : The Design Behind the Design

How to Architect Sites Across Cultures Without Losing Your Mind

by Adam Greenfield on 2002/03/11 | [22 Comments]

Ever since I started working formally as an information architect, somewhere back there in the antediluvian mists of 1999, I’ve clung to the belief that there’s a universal set of conditions that we’re trying to achieve in our work. These would be things like simplicity of structure and clarity in labeling: attributes that not only tend to further a user’s understanding of a given website, but also serve to enhance the web experience in general. But what I’ve slowly begun to believe over my time working here in Japan—and this verges on heresy for me—is that there is simply no such thing as a universal good.

I got my first clue that I might be headed for a comeuppance on my very first project in Japan. The client, a book site much along the lines of Amazon.com, insisted on providing what they called a “hon no sommelier” function, or “book sommelier.” This was a tool that was supposed to solicit a few lifestyle preferences from site visitors, after which it would recommend a book they might like.

Any conceptual problems with this idea aside, I pointed out that—in the States, anyway—only wine connoisseurs, foodies and restaurateurs were likely to be particularly familiar with the term “sommelier,” leading to a situation in which an unacceptable percentage of the user base quite simply would have no clue what that button up on the navigation bar was supposed to mean to them.

Further, and again based on my American experience, my guess was that a healthy segment of those that were familiar with “sommelier” would regard it as a thoroughly pretentious way to describe what, after all, was a simple recommendation tool.

I explained all this to the client’s representative, as patiently and objectively as I knew how, begging them to dispense with the idea. No dice. It was explained to me on more than one occasion, and with a fair amount of passion, that ever since the orgy of wine consumption the Japanese enjoyed in the Dionysian depths of the Bubble Economy, “everyone” knew what a sommelier was. Not only that, but that it would lend a “classy air” to the site.

From that point on we agreed to disagree, sort of. I provided schematics that called out an “o-susume,” or recommendation, function; the client continued to refer to it as the “hon no sommelier.”

There were other, far deeper problems with this site (some of which I’ve detailed in an article), and for a variety of reasons—only some of which had to do with matters of IA—my firm never completed the project. Eventually, it was subcontracted out to a production house that worked, apparently, without any IA input; the result is predictably turgid.

In the end, I believe nobody won. By inflexibly holding the line on “best practices” regarding a label, and a few other similar disagreements, I contributed to a situation in which a site was built that shafts the hapless user far more thoroughly than any we might have created.

The lesson here is really not a difficult one; it’s merely hard for a headstrong person like me to accept. And that is to slowly back out of the picture and do what I claim I’ve been all about from the beginning: listening to what the user wants. It so happens that, in the States, this is easy for me because “what the user wants” may mesh quite well with all those High Modernist values I hold dear. That is, there’s a happenstance overlap between the crisp grids and clearly articulated navigational schemas I personally like, and defensibly good usability practice for an American audience.

But what does “usability” or “clarity” mean in a culture like Japan? Have you ever ridden a Tokyo subway? If you have, I’m sure you’ll remember those ads, stuffed to the nonexistent margins with bright yellow copy against black backgrounds, sporting celebrity headshots, bikini girls, cute mascots and entire forests of exclamation points (the one I’m thinking of is an ad for a news weekly). How about all the consumer goods, including more than a few otherwise high-end efforts, overprinted with nonsensical Japlish slogans and cartoon characters?

Is it possible that in such an environment this is what your user expects, this is what your user is comfortable with—ultimately, this is what your user demands? Does it follow that, claiming as we do to be user-centric in our practice, this is what we should deliver to them?

Well, yes and no. In some cases, it may well be that the claims I found so ludicrous—”It will lend an air of classiness to the site”—are actually better approximations to the real user’s mindset than any I am capable of offering. Maybe, in context, “sommelier” isn’t such a hard thing to swallow.

In fact, this is what my Japanese superiors urged me to accept, in a slow and steady campaign aimed at securing my acquiescence. There may even have been one or two mornings, after absorbing this discourse, when I’d wake in a hungover daze of temporary agreement with them. Or not precisely with them, but with one of the few provocative, intellectually-coherent critiques of the practice of usability: that it encourages an incuriosity, a laziness, a bovine insistence on having everything placed within easy reach.

And in fact, I would be surprised if the incredibly dense, saturated, wildly overcoded Japanese media environment hadn’t somehow conditioned the average citizen to exhibit improved pattern recognition skills, finer knowledge-seeking reflexes.

But then I’d consult with Setsuko-san, and I’d see that all of this is beside the point.

She’s a persona I developed for Japanese consumer sites: a 48-year-old housewife in Niigata Prefecture, thoroughly preoccupied with the debts incurred by her idiot son while away at college, and her very significant doubts about the wisdom of her only daughter dating an American Marine.

Setsuko-san is my benchmark in these situations, and while she does not in any strict sense actually exist, I still take her feelings quite seriously. (You know those really annoying, sort of offensive born-again bumperstickers, the ones that say “My boss is a Jewish carpenter”? Well, same thing. My boss is a forty-eight-year-old housewife from Niigata.)

Setsuko-san, like most of the Japanese people I know, has a busy life, one all but defined by its dense and complicated web of mutual obligations and responsibilities. She’s trying to hold a family together, trying to put some money away, trying not to worry too much about what happens during those long hours her husband is away. Are you going to tell me she has the time or patience to wrestle with the vagaries of a bloated car insurance site trickling down a dialup connection?

I didn’t think so.

So what did I learn from all this? What can we draw out of all of this and apply to our experiences architecting informational spaces for non-Western cultures?

I learned about the limits of my own insight, and to be a little more careful about what I consider to be universals. As it turns out, there is a grain of truth in the idea that every audience is different and that this difference demands a variant construction of usability.

You really can’t get too far out ahead of your users, not if you want them to accept, understand, and feel comfortable using what you’ve built. So, for example, if someone on your team has specified pseudo-technical labels like “i-Appli” and “L-Mode,” and the user base doesn’t seem to have a problem parsing them, let it go—even if you strain against it with most of the fibers of your being.

But I also learned to respect my instincts: to trust the understanding I had built up through years of listening to the actual human beings who, for better or worse, use the things we build. This gave me the strength to press ahead when I was told “you can’t hit a home run every time”—meaning, don’t ask too much of us, both us-the-client, and, almost as frequently, us-the-account-management-team.

Between these two lessons I’ve come to believe that you can safely let go of 90 percent of everything you think you know about IA, as long as you never once lose sight of your user. As long as your dedication to this person and their needs remains unswerving, as long as you’re truly organizing information to maximize their understanding, their ease of use, you will not go wrong, whether you’re working in SoMa, Sydney or Shibuya.