Nobody Notices Good Information Architecture. That’s the Point.

There is a well-known story that Lief Erickson used to open his Information Architecture course at FH Joanneum. Two young fish are swimming along when an older fish passes and says: “Morning, boys, how’s the water?” The two fish swim on for a bit, then one looks at the other and asks: “What is water?”

Information architecture is the water. It is everywhere, it shapes everything, and most people only notice it when something goes wrong.

I work with a CMS every day. I make decisions about how content is categorised, what navigation items are called, how pages are grouped and where they live. It turns out there is an entire discipline built around exactly these decisions.

What Information Architecture Actually Is

IA is the practice of organising and labelling content so that people can find what they need. It covers how content is grouped, what it is called, how users move through it, and how they know where they are at any given moment. It is the invisible layer underneath every website, app, or digital product that determines whether the experience feels intuitive or frustrating.

It is not design. It is not content in the traditional sense. But it shapes how both are experienced.

Peter Morville and Louis Rosenfeld describe it simply: good IA makes the complex clear. Bad IA makes the simple confusing.

You Are Not the Hero

One of the most useful reframes from the course came from Lief Erickson’s session on storytelling as a tool in IA. The insight sounds simple but changes a lot: the user is the hero, not the brand. The website is not the main character. It is the guide that helps the hero accomplish their goal.

This is something I had already internalised from content creation in practice. When you write for users rather than for the brand, you naturally shift your perspective: what does this person need, what problem are they trying to solve, and how does this content help them get there? What I did not expect was to find the same principle embedded in something as structural as information architecture. The hero framework is not just a storytelling tool. It is a design principle. And seeing it show up across disciplines made it feel less like a concept and more like a rule.

Think about buying a product online. The user arrives with a need. They search, browse, compare, and eventually decide. At every step, the website either helps them move forward or puts an obstacle in their way. A confusing category name. A filter that does not match how they think. A product page that answers the wrong questions. Each of these is a small failure of information architecture, and each one makes the hero’s journey harder.

The job of IA is to clear the path. To make the structure so logical, the labels so obvious, and the navigation so intuitive that the user never has to think about it. They just find what they came for.

A Physical Intuition We Forget to Apply Online

Walk into a supermarket and you immediately know where to go. Not because someone told you, but because the layout communicates it. Vegetables near the entrance. Dairy at the back. Snacks at eye level. You navigate without thinking.

Physical spaces communicate purpose through their layout. Implicit cues tell us where we are and what is expected of us. We understand a library differently from a restaurant, even before we read a single sign.

The same principle applies online. A website that follows established patterns allows users to focus on what they came to do rather than on figuring out how to move around. When those patterns break, even slightly, the cognitive load increases. Users start thinking about the interface instead of the task. And thinking about the interface is exactly what you do not want them doing.

The Words on the Buttons Matter More Than You Think

One of the most underestimated parts of IA is labeling. Labels should quickly take users to the content they need, without taking up too much space on the page or in their brains.

That sounds obvious. In practice, it is surprisingly hard.

Labels are often written from the inside out. A company decides what to call something based on internal terminology, how their departments are structured, what feels natural to the team. But users arrive with their own vocabulary and their own expectations. When the label a company uses does not match the word a user would reach for, the content might as well not exist. It will not be found.

This became even clearer to me when I did an accessibility audit as part of my first project work at FH Joanneum. One of the findings that stuck with me: a link with no readable name is not just a code issue. Someone wrote that label, or didn’t. As I reflected in that article, accessibility is a content quality issue. Good content does not just read well. It works for everyone it is meant for. And that starts with the label.

Working in a CMS, I see this constantly: the internal name for something and the word a customer would use are often not the same thing. And for users relying on assistive technology, a mislabeled link is not just inconvenient. It is a dead end.

Search Is Not a Fallback

When users cannot find something through browsing, they search. Search is often treated as a fallback, the thing users turn to when navigation fails. But good IA treats search as a first-class experience. That means thinking carefully about what content exists, how it is structured, and what terms users actually use when they are looking for something.

During the course, Lief highlighted something that stuck with me: well-designed filters on a search page, combined with a content structure that actually reflects how users think, are what separate useful search from frustrating search. We had integrated an AI-powered search on our website that understands what users mean rather than just matching keywords, and seeing it discussed as an example of good IA in practice made me appreciate the work behind it differently.

IA is not just about where things live. It is about whether they can be found at all.

Navigation Is Not the Goal

Jesse James Garrett put it plainly: users do not visit websites to experience the joy of navigation. They visit to accomplish something. And when the structure puts too many options in front of them at once, the result is often no decision at all. Martin Pfandl describes exactly this in his piece on UX principles: overwhelmed by room options on a booking platform, he postponed the decision entirely. Hick’s Law in action. The more choices, the longer it takes, and sometimes the more likely users are to leave.

Good IA gets out of the way. It creates the conditions for users to find what they need, complete what they came to do, and leave feeling like the experience was easy. The best information architecture is invisible precisely because it works.

Structure Is Content

The decisions I make every day in a CMS, what to call a category, how to group a set of pages, where a piece of content belongs, are information architecture decisions, whether I label them that way or not.

As Nadja Hodzic captures in her piece on lorem ipsum as a content choice: when you build a layout around placeholder text, you have already decided how long the copy will be and how much space the message gets. Structure is not neutral. Every structural decision is a content decision made early, and its consequences arrive later.

The best IA is invisible. It just works. And the fish never has to ask what water is.