Twenty-five years into designing and developing for the web and we still collectively suck at information architecture.
We are taught to be innovative, creative, agile, and iterative, but where and when are we taught how to make complex things clear? In my opinion, the most important thing we can do to make the world a clearer place is teach people how to think critically about structure and language.
We need to teach people that information architecture (IA) decisions are just as important as the look and feel of technology stack choices. We need to teach people the importance of semantics and meaning. We need to teach people to look past the way the rest of the web is structured and consider instead how their corner of the web can be structured to support their own unique intentions.
The web was born to be a democratized building site, and it’s grown into a place that most people visit multiple times per day.
The role of IA is democratizing as well. The tools and resources we use to structure, design, and develop the web are becoming easier to use, and so we need to know the impact that our structural and linguistic choices have on the integrity, efficacy, and accessibility of the places we’re making.
The choices we make about structure and language so things make sense is the essence of IA. It’s a responsibility unevenly distributed across job titles ranging from user experience design, interaction design, content strategy, instructional design, environmental wayfinding, and database architecture. It’s also practiced widely outside the technology and design sector by people like teachers, business owners, policy makers, and others who make things make sense to other people.
Fact: Most people practicing information architecture have never heard the term before. I believe that this is why we aren’t collectively getting better at this important practice.
Without a label, a common nomenclature, IA can seem like an insurmountable mountain to climb. Let’s say you’re working on how to arrange and label the parts of your marketing website, as well as improve the categorization of your online product catalog. To help with these tasks, what do you use as keywords to find your way?
“How to organize a website?”
“What are e-commerce catalog best practices?”
“How to choose categories for product catalogs?”
This is like googling symptoms of a disease you’re suffering from. It is a long, hard, frustrating road to take. Without knowing the words “information architecture,” you are only likely to find the ways other people have already solved specific problems.
Don’t get me wrong, this is a fine first step, but without understanding the conceptual underpinnings of IA, people are more likely to end up propagating patterns they see on the parts of the web they experience. This trend is making too much of the web look and act the same, as if everyone is working from a single floor plan and the entire world is slowly becoming one big suburban subdivision.
In 2013, I was preparing to interview Lou Rosenfeld onstage at World Information Architecture Day in New York City. While doing my homework for the interview, I had the chance to speak with Peter Morville about the rise of IA as a field of practice. He told me that before the term “information architecture” was popularized, people referred to something called “the pain with no name.”
The phraseology of “the pain with no name” is powerful because it properly captures the anxiety involved in making structural and linguistic decisions. It is messy, painful, brain-melting work that takes a commitment to clarity and coherence.
These pains did not die with the birth of web 2.0. Every single person working on the web today has dealt with a situation where the pain with no name has reared its ugly head, leaving disinformation and misinformation in its wake. Consider:
“Our marketing team has a different language than the technology team.”
“Our users don’t understand the language of our business.”
“The way this is labeled or classified is keeping users from finding or understanding it.”
“We have several labels for the same thing and it gets in the way when discussing things.”
These pains persist on every project; disagreements about language and structure often go unresolved due to a lack of clear ownership. Since they’re owned and influenced by everything from business strategy to technical development, it’s hard to fit these conversations onto a Gantt chart or project plan. IA discussions seem to pop up over the course of a project like a game of whack-a-mole.
When I worked on an agency team, it was quite common for copywriters to want responsibility for coming up with the final labels for any navigation system I proposed. They rightly saw these labels as important brand assets. But it was also quite common for us to learn through testing and analytic reports that these branded labels were not performing as expected with users. In meeting after meeting, we struggled and argued over the fact that my proposed labels—while more to the point than theirs—were dry, boring or not “on brand.” Sometimes I won these arguments, but I was usually overpowered by the creative team’s preference for pithy, cute, metaphoric, or irreverent labels that “better matched the brand.”
In the worst incident, the label I proposed made sense to 9 of 10 users in a lab usability test of wireframes. The same content was tested again following development, but was now hidden behind a cute, branded label that made sense to 0 of 10 users. Ultimately, the client was convinced by the creative team that the lab test had biased it in this direction. Once we had a few months of analytics captured from the live site, we saw the problem was, in fact, real. It was the first time I’ve ever seen 0% of users click on a main navigation item.
Seven years later, that label is still on the site and no users have ever clicked on it. The client hasn’t been able to prioritize the budget to fix it since they need to pay for campaign-based work (much of which is ironically hidden behind that cute but confusing label). This was the first time I fully understood how much of my job is to teach others to consider IA and not just listen to my recommendations around it.
I fear that we have become lost in a war of dividing responsibility. Clarity is the victim in these battles. It doesn’t matter who comes up with the label or who decides how something is arranged. What matters is that someone thinks about it and decides a way forward that upholds clarity and intention.
There is more information swirling around in the world than ever before. There are more channels through which we disseminate content. There has never been such a pressing need for critical thinking about structure to ensure things make sense. Yet, I believe that the pain with no name is experiencing a second coming.
In too many cases, educational programs in design and technology have stopped teaching or even talking about IA. Professionals in the web industry have stopped teaching their clients about its importance. Reasons for this include “navigation is dead,” “the web is bottom up, not top down,” and “search overthrew structure”—but these all frame IA as a pattern or fad that went out with tree controls being used as navigation.
These misconceptions need to be addressed if we are going to deal with the reality of the impending “tsunami of information” approaching our shores. The need for clarity will never go out of style, and neither will the importance of language and structure. We will always need to have semantic and structural arguments to get good work done.
I have worked with too many businesses with inherited “lacksonomies” that emerged from the sense that there’s only one way to organize an e-commerce site, mobile app, or marketing site. We forget that most of the interfaces out there are more experiment than proven pattern. In other words, be careful when copying from others.
Many people believe that a large or popular brand has “probably” tested their architectural decisions, when in reality, that’s often not the case. The truth is that we never know if we’re looking at something being A/B tested or redesigned behind the scenes because it’s not working.
How can we be sure that the patterns we’re copying are well-founded?
The truth is that we can’t. Something that works for Amazon might not work for your business. Something Google did might be a terrible decision when applied to your context. I once had a client who wanted their product to be structured like iTunes, because Apple is so great at design.
Only you can help the world to give this pain a name.
When a structural or linguistic decision is being discussed, call it out as information architecture. Give people the label they’re searching for to describe the pain and anxiety being faced. If there is a semantic argument to be had, have it and make sure those you’re arguing with know the impact of leaving such things unresolved.
Teach others about the ramifications of IA decision-making. Warn your coworkers and clients that IA is not a phase or process that can be set once and forgotten. It’s an ongoing discussion that can be impacted during any stage of the work.
Share your IA struggles with colleagues and peers so our community can grow from collective experiences. If you want a venue for sharing and learning more about the global conversation happening around information architecture, find a World IA Day location near you.
10 Reader Comments
Outside of the web sphere (DMS, AI, etc.) there is another term : an Ontology.
Great article – I’m going through a similar navigation label discussion on one of my projects. It’s slightly comforting to hear that this happens to other people too 🙂
Amen. “The pain with no name” isn’t just a slogan, it’s an example.
This reminds me of what Jesse James Garrett said: “I remain amazed and perplexed at how often people think they can solve an information architecture problem with interaction design.”
Though I do totally agree with your argument as an important—part—of the considerations when “Designing” sites, I do also think there’s some relative lessons to be learned from deliberately sticking to convention. Not that—all—convention is good. And certainly your points about assuming any sites have been A/B tested or intelligently user-tested or measured is no good. Then again, wistfully departing from convention with wildly divergent—be it potentially unique and tailored IA—that is so different from user expectations, could also set your users up for failure. (And that goes for all parts of the process, not just the IA.) I totally agree that so much of the web, is so much of the same same. But in many ways, the web is essentially one kind of thing, that we’ve all learned how to interact with, like a bike, car, or phone. Departing from convention to suite your site’s specific needs sounds fantastic. But the amount of sheer effort, ingenuity, and possibly some factor of good timing, maybe shouldn’t be discounted. The iPod was radically, insanely brilliant. It’s user experience was impeccable. And it was distinctly it’s own thang. (Though I wonder how well it may have done in 1980.) But iPods are rare. We definitely need and want them. But maybe some convention is okay too. Constant balance with infinite tradeoffs like most everything of course. Still, I do really believe in what you’ve written here. Thank you!
Thanks for the great piece.
If IA covers all of structure and language, does that mean you consider content (and content strategy) to be IA?
@Jessica — I would not say I consider content and content strategy to “be” IA — instead the way I see it: IA is a skill set that anyone practicing content strategy or making content is already practicing.
excellent article, thank you Abby. Agree with focusing on the language and structure, not copying and pasting from elsewhere or going along with what has been inherited. As an editor, I would call it trying to do a structural edit based on conventions that can be shown to work with your own audience in combination with what the content needs to say. cheers.
Thanks for sharing information for web designing with us.
A bit late, but such a gem of an article, Abby. As a long-time IA, I am ever amazed at how many people still diminish the value of information architecture. Today’s excuse for ignoring IA is that it’s not agile…Sigh…We will all still fight the good fight. Thanks again.
Got something to say?
We have turned off comments, but you can see what folks had to say before we did so.
More from ALA
Personalization Pyramid: A Framework for Designing with User Data
Mobile-First CSS: Is It Time for a Rethink?
Designers, (Re)define Success First
Breaking Out of the Box
How to Sell UX Research with Two Simple Questions