What is Information Architecture?
Looking for a comprehensive overview of information architecture? Look no further. This curated list of articles and resources from the fine folks at Optimal Workshop is a brilliant reference.
Looking for a comprehensive overview of information architecture? Look no further. This curated list of articles and resources from the fine folks at Optimal Workshop is a brilliant reference.
Check out this week’s Every Little Thing podcast. It’s nominally about the Dewey Decimal System, but it’s really about the political nature of organizing things and how the organizing systems we create reflect our biases. (Starts 9-minutes in.)
YADIA: “IA defines spatial relationships and organizational systems, and seeks to establish hierarchies, taxonomies, vocabularies, and schema—resulting in documentation like sitemaps, wireframes, content types, and user flows, and allowing us to design things like navigation and search systems.”
from Sara Wachter-Boettcher. “Content Everywhere”
Yet Another Definition of IA (YADIA):
“The activity of Information Architecture [is] designing an abstract and effective organization of information and then exposing that organization to facilitate navigation and information use.”
-from The Discipline of Organizing, by Robert Glushko
Partly to test this blog’s linkage with micro.blog, I thought I’d mention that I’ve been digging back into Designing the Search Experience for inspiration lately. So much good stuff in there about how to turn information behavior studies into practical design solutions. Two words: search modes!
The following is adapted from my opening remarks at World IA Day Seattle 2016, which took place on Saturday, February 20. The theme for World IA Day this year was: Information Everywhere, Architects Everywhere.
If you've been around the information management world for any length of time, you've probably heard the joke about the old fish and the young fish. The old fish says "Water's fine today". And the young fish says, "What's water?"
I didn't say it was a good joke.
But it is useful as a shorthand for explaining something about what information is. We're like the fish, obviously, and information is all around us. We're swimming in it, but we don't even notice it until we learn to see it.
How much information did you encounter last week? This morning? Since you started reading this? I'll bet you couldn't quantify the amount of information around you on any time scale. The room you're in is information, the street outside, the words you're reading, the clothes we're wearing... every sight, smell, sound, and surface carries information, and we process it all in an instant and without even noticing that we're doing it.
We live in a universe of information. And most of the time we can, like the young fish, just swim in it and go about the business of being. But sometimes, we want to shape and form information into something intentional and meaningful, into a web site, an intranet, an app, a monument, or some other information experience. At those moments, when information is both the medium and the message, we must notice the information all around us and attempt to make it meaningful to ourselves and others. We must apply design. We must practice information architecture.
Now, I imagine a variation of the joke about the fish where in this version the old fish says to the young fish: "I'm a fish." And the young fish says, "What's a fish?"
It's still not a good joke.
But I think we encounter something like this when we try to explain to our friends, family, colleagues, and bosses that we're information architects. When I tell someone I'm an information architect, I get something of a blank stare. For the longest time I tried to figure out how to break through that and come up with a cool way of explaining what I do ("I'm like a ninja, but with information."), but I'm starting to lose hope that I'll come up with the right words.
After all, everyone's something of an information architect. Everyone organizes something: closets, movie collections, garages, files on the computer, kitchens, bookshelves... you name it. We all try to impose some sort of order on the world, to create systems that make sense and keep on making sense, and impart some sort of meaning to others. We're all fish. I mean, we're all architects.
It's just that, for those of us who are crazy enough to voluntarily identify ourselves as "information architects", we're doing more than organizing our spice racks or shoe closets. We are doing the same thing, essentially, except we're attempting to do it at scale. We're trying to impose order on thousands and millions of items of information at a time, for users who may number in millions or billions. And these days we're usually trying to do it within a window the size of an index card.
And there's something so interesting about that to me. It seems like a fraught enterprise: doomed yet noble, and occasionally elegant and beautiful. There is information everywhere. And there are architects everywhere. But the rare breed who call themselves information architects are lucky enough to recognize these things; to understand that this is water, and we are fish.
And to be able to know that is pretty damned cool.
It's really hard to get people to understand why it's worth investing in metadata and taxonomy projects. The benefits aren't immediate and the reasons can seem esoteric. It's only after the work is done that the usefulness of metadata starts to become clear.
Proof of this comes in this interview with a colleague of mine at REI. This is a quote I’m going to pull out at every metadata and taxonomy meeting from now on:
“[Collecting metadata] turned out to be really smart. We didn't realize the repercussions of it when we did it. But the structured way we captured the meta-data and user-generated content (UGC) laid the groundwork for how we use that content." (My emphasis.)
I had nothing to do with the decision to collect metadata in this instance, but I've seen firsthand the powerful unintended benefits of having robust structured content. Perhaps one way to convince others ahead of time that they should invest in proper content markup is to collect more testimonials and stories like these. If you know of any others, let me know in the comments.
In my opinion, IA is not something that needs to be sold. IA is already inherent to whatever someone is working on or has in place. If you are making something, you will be tackling the IA within and around it. With or without me you will “do IA.”
I guess in sales speak we could say “IA is included for free in all projects” — because a system without an information architecture does not exist. Rather than selling information architecture, I find that I do have to “explain” what it is and why it matters so that it can be worked on and improved upon (not ignored or inherited which is all too often the case)
Whether you're interested in "selling" IA or not, the fact is you'll probably have to explain what you do to others. Probably multiple times. Per day. You could do worse than have a few of Abby's scripts memorized.
Last week I attended the IA Summit in Baltimore. It was my first IA Summit and I'm very glad I decided to go. I met so many intelligent, thoughtful, and passionate practitioners (and academics) of information architecture, and I found myself inspired and challenged to raise my game and do better work.
The following are some of my rough notes from the five days of the conference. While this is short on narrative, I hope there are enough nuggets of wisdom and links to explore that you'll find something new to think about.
Some of the major concepts that kept cropping up for me:
"Un esprit nouveau souffle aujourd'hui" (A new spirit breathes [lives] today)
Andrea Resmini introduced the day by putting the current information ecology in perspective for us. We have moved from a world where computing was done in a specific time and place to a world where computing is ever-present. "Cyberspace is not a place you go to but rather a layer tightly integrated into the world around us."
For example, the average citizen in the EU spends 29.5 hours computing each week. The average citizen in the UK spends 39 hours computing each week.
The day was structured around groups of Ignite-style talks, with discussion and an exercise after each.
David Fiorito: The Cultural Dimensions Of Information Architecture.
Culture is a learned and shared way of living; we are creating culture through IA
David Peter Simon: Complexity Mapping
We need to develop device-agnostic content.
We think a lot about how content will be presented (different forms like tablets, watches, google glasses) but we don't think so much about how information will be presented on those new forms.
Simon Norris
"Man is a being in search of meaning" -Plato
Take a meaning-first approach to IA. Start with meaning
Matt Nish-Lapidus: Design For The Network: The Practice Of 21st Century Design And IA
The Romans built road networks to spread their cultural software. We build networks to spread a different type of cultural software.
Lev Manovich talks about the Cultural Interface: "…we are no longer interfacing to a computer but to culture encoded in digital form …" (Lev Manovich, The Language of New Media (2001))
We need a new design practice for the networked world, one that embraces humans, technology, craft, and interface.
Culture is moving faster and faster, moving up in the pace layer diagram
Was Le Corbusier an IA? He was part of our history, one of the people whose shoulders we stand on, but his work must be understood in the context of his time. Calling him an IA wouldn't be fair.
Does the IA create meaning by designing which resource is linked, or is there no way to create meaning because there are so many hyperlinks that users create their own meaning?
Christina Wodtke suggested developing a "Poetics of information architecture"
(Note: I missed a speaker here, but I think it was Duane Degler)
Sally Burford: The Practitioners of Web IA Reviewed some research into The practice of IA at small and large organizations. IA needs to develop an identity, to take a strong stance
Jason Hobbes and Terrence Fenn: The IA of Meaning-Making
Defining the problem defines the solution; this is the problem/solution ecology
The IA design deliverable is resolving the problem/solution ecology
What's missing around IA in the academic community?
There is no written history of IA. (Some suggested a book by Earl Morrow (?), but others thought even that book was out of date).
It's hard to have a career in academia when the term IA is not understood.
Dan Klyn: Dutch Uncles, Ducks, And Decorated Sheds
Meaning should be the center of our reframe of IA
Don't bake meaning into structure; keep them separate
Jorge Arango: "Good Fit" In The Design Of Information Spaces
Jorge developed themes from Alexander's Notes On The Synthesis Of Form.
The designer has to create a mental picture of the actual world. In really complex problems, formal representations of the mental model need to be created.
Andrew Hinton: A Model for Information Environments
IA has always been architecture of a dimension of a shared reality.
We should continue working on reframing IA by thinking of it as a design problem.
IA is about solving a problem.
Don't reinvent, link!
Shared model + shared language + shared understanding = consistent user experience
Structured content refers to information or content that has been broken down and classified using metadata.
Knowledge rejects rigid structure.
When we use the same language to describe the same things, we can build a web of knowledge across various services.
Structured content breaks down information into things and the relationships between them. A content model maps our subject domain, not our website structure or content inventory. Assertions distinguish between real-world things and the documents that refer to them.
Experts map the world, users mark points of interest.
Modeling begins with research, talking to experts and users to understand and articulate the subject. Ubiquitous language describes the granular terms that will be used by everyone on the project. Boundary objects show where your subject model can connect to neighboring models.
Content is hard.
Everything starts with good content. Content is not the stuff we pour into nicely designed containers. It's not the stuff we chase up from the client at the last moment. It is not a placeholder. Content is the whole damn point.
"Do what you do best, and link to the rest." -Jeff Jarvis
The link forces specialization and specialization demands quality
Other parts of your organization may be sitting on a goldmine of content or business data. Find it and exploit it.
The BBC's principles:
We often walk right past the next big thing.
Blog post: Bears, Bats, and Bees
We are in a "model crisis"
We need a Google for your room. A personal search service for the things that are important to you.
We live in both the digital and physical environments simultaneously.
We need to look at how we comprehend environments generally in order to understand how we understand digital environments.
Pace layers of information environments:
Information in three modes
Traditional cognitive theory assumes the brain is like a computer
Embodied cognition says that the brain uses the body and the environment to understand
Some key ideas from James J. Gibson’s theories
Semantic information changes how we experience environment
"Language is a form of cognitive scaffolding." - Andy Clark, Supersizing the Mind
The body and the ecological information can override the semantic information, as in a door that looks like a push door but has a sign that says pull.
Digital information enables pervasive semantic place-making
A design presentation for the American Society for Immortalization Science and Technology
We need a better way to define what good IA is. And we need better ways to communicate about IA.
In architecture, simple forms can be recombined in different forms to create new forms.
Forms are made up of and can contain other forms
Nodes and links are our form and space
Nodes are our basic building blocks, and also the walls and buildings themselves
Nodes can contain and are made up of other nodes
Link: the relationship between nodes
There is a Form/Space Hierarchy (most complex at the top:)
Similarly, there is a Node/Link Hierarchy (most complex at the top:)
We need to apply different approaches to different levels of the node/link hierarchy
IA is focused on the structural integrity of meaning.
Taxonomy is rhetorical, place-making, and tied to embodiment
Mobile taxonomy is device independent
Mobile taxonomy is articulated
Mobile taxonomy is flexible
Taxonomies are goal-driven. They are rhetorical
Rhetoric - the means by which we inform, persuade, of motivate particular audiences in specific situations
Architecture is rhetoric for spaces
Taxonomy - a method of arrangement conceived to create a particular kind of understanding
Taxonomy is
This was a rich and thought-provoking talk, but difficult to summarize. The major themes were:
There is a new spirit. It’s not towards order but towards disorder and multiplicity
A spirit of context, place and meaning
A spirit of sense-making, … and creating new places for humans to work and play
Peter Morville and Lisa Welchman
Big governance vs, local governance
There’s a lot of stuff happening at the local [organizational] level, but few people if anyone who are looking at the overall picture.
There are usually political barriers to solving information problems on sites. They’re not IA problems.
The governance of your web:
Your “stable environment” equals clear sponsorship, goals, and accountability.
Your “good genes” are goals are policy-driven, standards-based framework
These things will allow your web to grow and still be recognizable as your web. That’s all it is.
Enterprise web governance ensures the proper stewardship of an organizational web presence.
Stewardship, not ownership.
Abstraction - the intentional filtering of information to focus on core goals
We must carefully choose the model through which we hold conversations. Abstractions help stakeholders discover.
Metadata enables consistency, context, and interoperability.
A cross channel experience requires sending information across channels.
“Messages have no meaning” - Andrea Resmini
… But every message has a structure
Standards:
Ontology is things and their relationships. When you have a concept of the space and you express it in a way that can be interchanged, you have an ontology.
Scott leads The Noun Project
Types of written languages:
Pictographic. Ideographic. Syllabic. Phonetic.
Christian Crumlish, Bryce Glass, Donna Lichaw
The panelists discussed moving out of IA and UX and into product roles.
Donna - "I’m doing more IA and UX now as a product manager than I ever did when I was a UX designer"
Christian - my titles have changed, but I’ve been basically doing the same thing: defining the damn thing of the product, connecting people and ideas.
Donna - A product manager is like a producer: they conceive the vision and get people moving in the same direction
Donna - In New York, no one has any idea what a product manager is
Bryce - "We may be confused, but we have a community of confusion"
Christian- Third wave IA - or Resmini-Hinton-Arango IA - the high level stuff; what is this, what does it do, what is going on in the user’s brain
Christian - There’s a clearer career path for the people who are in charge of product. There’s more headroom.
Xerox PARC identified a disease: biggerism
Solving the big problem seems more interesting, more exciting, more challenging
Small data problems: a large personal library, a desktop full of icons, a gmail inbox
Big data is about stuff that’s too big for anyone to understand
Small data is about:
… And yet we focus on the app, not the coordination of the data
Organizations have big data problems
People have small data problems
Mathematicians have a saying: Problems worthy of attack prove their worth by fighting back.
We have small data problems worth working on. We are hitting the edges of search. Where’s the difference between searching and filtering? We have faceted search and faceted navigation. Which is it?
Problems are framed by how we think about the work we do. [I would say they’re also framed by the tools we have. -sm]
We need to make significant positive progress on a hard problem. We need to not be distracted by biggerism.
When we go in to a project with a personal methodology…
We need ambiguity tolerance
Discovery is not a phase, It is an ongoing activity
Problem solving is squiggly
Classic personas are the users we want to come to the site. We give them happy paths.
At the other end of the scale are anti-personas, people we want to discourage. (People who intend harm, people who are doing things with the service we don’t want, people who are too young)
Negative personas are not the primary audience, but may still come and be disappointed. Are we giving them what they need?
We’re not designing for everyone. But we need to acknowledge and handle pain points. In order to increase the number of satisfied customers.
Does it matter? Yes
You can’t control who comes to the site, so you have to try just to not annoy them.
Design for success or failure.
So I'm a wee bit late posting this, but I was working on a poster presentation for this year's IA Summit when I came across the notes I took for my presentation at InfoCamp Seattle 2012. It's now posted here for internet posterity.
InfoCamp intro to VRM (PDF)
I thought this might be interesting for two reasons. First, I think the central message of the presentation is still relevant: information management professionals should be at the forefront of thinking about how people identify, store, and manage their own data in a VRM-enabled world. The PDF roughs out the main ideas of VRM and suggests how information management professionals might contribute to this emerging movement.
(Update 3-22-13: Holy smokes! I didn't realize that Tracy Wolfe, the Mod Librarian, posted a nice writeup of this presentation back in October. I really need to start paying attention to this whole internet thing. I hear it's really taking off.)
Secondly, (and a bit more prosaically) I drafted these notes in Scapple, an awfully interesting brainstorming program under development at Literature and Latte (the fine folks who make Scrivener). Scapple basically lets you enter free-form bits of text on a canvas of whatever size you choose. You can move the bits around to regroup them, and you can drag one text bit over another to relate the two with a dotted line. It was truly one of the best brainstorming experiences I've had on a computer. I ended up using the Scapple document as both my presenter notes and my presentation slide.
If you're like me and you're really not into the linear process of outlining, but you still want to do something like outlining on a computer, I encourage you to give Scapple a try. It's free for a limited time while in Beta. But watch out: you might get hooked.
Drew Olanoff suggests that Twitter could extend the hashtag idea to make "bits and pieces of data clickable." I'm not a fan of Twitter these days, but I love the idea of incorporating data structure into online content through simple affordances. Ah, so that’s Twitter’s strategy: A “clickable” world - The Next Web:
"By structuring data, Twitter could make its network of information easier to navigate and discover upon. It would also help the company structure its API so that third-party developers wouldn’t have to dig through every single tweet for particular information. I can click or tap around Wikipedia for hours, since everything is linked by its editors. Twitter could engage users in the same way."
This is why the future is in metadata.
Phil Windley gets us thinking about what a "smart bike" might do:
Imagine the bike being connected to its manufacturer, the bike store that sold it, and its owner. From its earliest point in being, the bike would be able to keep track of data about itself, things like its specifications, when it was made, and even the provenance of the materials used in its manufacture. The bike would keep track of inventory data like when it was delivered to the bike shop, who assembled it, its price, and when it was bought and by who. And all this would be possible with a personal cloud for a bike.
While Phil's vision could be accomplished without an on-board bike computer, it's hard to imagine a truly useful fire-and-forget system without one, or without a connected infrastructure at the important points of presence: the manufacturer, the store, the bike shop, etc. But this sort of automated administration would be useful for all sorts of things, and I don't see any technical hurdles standing in the way. It's just a matter of building the supporting schemas and software.
(This post was originally published on The Machine That Goes Ping on 5/26/12)
I ran across this definition of information architecture at IBM's developerWorks site and thought I'd share it here:
Design patterns for information architecture with DITA map domains
Information architecture Information architecture can be summarized as the design discipline that organizes information and its navigation so an audience can acquire knowledge easily and efficiently. For instance, the information architecture of a Web site often provides a hierarchy of Web pages for drilling down from general to detailed information, different types of Web pages for different purposes such as news and documentation, and so on. An information architecture is subliminal when it works well. The lack of information architecture is glaring when it works poorly. The user cannot find information or, even worse, cannot recognize or assimilate information when by chance it is encountered. You probably have experience with Web sites that are poorly organized or uneven in their approach, so that conventions learned in one part of the Web site have no application elsewhere. Extracting knowledge from such information resources is exhausting, and users quickly abandon the effort and seek the information elsewhere. The same issues apply with equal force to other online information systems, such as help systems. The organization and navigation of the information has a dramatic impact on the user's ability to acquire knowledge.
I just came across some notes I took at this year's orientation for Mid-Career MSIMers. "The Question" came up, as it inevitably does in these kinds of gatherings: What is information management? Here are a few snippets I pulled from the answers back in September.
Originally posted at MSIM 2011