The Art and Science of Information Architecture

In the context of a discussion around information architecture, (IA), definitions are fairly plentiful and even synonymous terms exist, like information design, that, depending on the context, may not mean IA in the strict sense at all.

There’s a certain irony that this lack of clarity and ambiguity exists around this topic since in a good IA there is clarity in terminology and the organization of the content and no ambiguity between labels. So let’s establish what we mean when we talk about IA.

Defining Information Architecture

There are certain definitions from authoritative bodies/authors that have made their way into Wikipedia, but that in themselves leave room for interpretation:

The Structural Design of Shared Information Environments.

This definition suggests that there is overlap at the least, and complete integration at the most, between IA and interaction design; how we design the structure is the focus here.  Information environments indicate that what is being structured includes content, so how we structure that content is part of IA.

The Art and Science of Organizing and Labelling Web Sites, Intranets, Online Communities and software to Support Usability and Findability.

This definition doesn’t mention design explicitly; the focus here is on organizing the content so that users can find what they are looking for, and once they do find it, they can use it effectively and efficiently.

These organizational and labelling activities are cited as art and science.  Part of the art derives from the fact that there are many ways to solve IA challenges and judging which is best is a bit of an art; but, part of the art also alludes to the fact that human beings are the users for whom we’re creating this IA, so understanding their needs and figuring out the best IA to suit most of our users most of the time is also part art.

The science involved here largely includes cognitive science and information science; understanding how our specific user personas think about and organize the body of content we have to work with and then applying best practices in terms of classifying that content and controlling the vocabulary we use to refer to it.

More Than Content Knowledge

The broadness of these definitions reflects the interdisciplinary nature of the practice of IA and its complexity.  They also suggest that creating a good IA requires more than a thorough understanding of the content we are going to be organizing.

First and foremost, we need to understand, through user research, who is going to be using this content and what are the mental models of those people – how do they think about this content.  We can’t divine this insight or move ahead entirely on assumptions – we must engage with real users to find this out.

In parallel, we can start to understand the body of content we are going to be dealing with.  A content audit provides a superficial, but highly beneficial, pass over the content so we can see what “types” of content exist and what it is about.  A content inventory takes much longer than an audit, but enables us to dive deeply into all the content and discover what’s really there, what’s ROT (redundant, out of date, trivial) and what relationships currently exist within the content.

Once we know who our users are, and understand the content better either through an audit or an inventory, we can conduct a card sorting activity. Card sorting will enable us to better understand how real users organize the content.

After these three things are done: user research, content audit/inventory and card sorting, then the fun IA work starts. We use the research as inputs and start to classify the content into groupings based on a schema that is informed by the inputs and artfully structured by us, the Information Architects.