Donna Spencer has practiced information architecture at the highest levels for more than two decades. She has taught innumerable workshops and courses, and she wrote an entire book on the subject.
Still, to this day, when she teaches IA, she feels inadequate.
This speaks in no way to her professional knowledge or her skills as a teacher. Donna knows this field as well as anyone and speaks eloquently about it. Rather, it gets at the “jigsaw mess” of the practice of information architecture and the constraints of modern educational environments.
We talked about:
- her work at MakerX and its alignment with her natural maker tendencies
- her definition of information architecture, and the importance of foundational structure work
- how IA work has changed over the years, in particular the shift from manual to automated processes
- the challenges of teaching structural concepts to modern UX practitioners, who tend to focus on journeys and flows
- the importance of metadata in digital product design
- how to get action-oriented designers to consider the structural side of design, and accounting for these different mindsets in your IA work
- her approach to teaching IA
- the importance of understanding how people operate, especially differences in communication styles
- how she teaches her students to model content
- how her academic background in economics prepared her to be a modeler
Donna Spencer: Information Architecture Educator and Author – Episode 112
Transcript
Larry:
Hi, everyone. Welcome to episode number 112 of the Content Strategy Insights Podcast. I am super-delighted today to have with us. Donna Spencer, Donna is currently a principle product designer at Maker, but she’s been known as an information architect for a long time. I mean, I first discovered her book in … I don’t know, 15 years ago, whenever that was. So anyhow, welcome, Donna. You want to tell the folks a little bit more about what you’re up to these days?
Donna:
Thank you very much. Well, so firstly, I just should disclose that what I’m up to at the moment is painting my new house. So, it’s me, safety glasses and a lot of paint all over me.
Larry:
Nice.
Donna:
Because if anybody’s looking at it going, “Why does she look so weird?” That would be why. Aside from painting an old house, because we’re talking between Christmas and New Year, so this is the time when I’m not thinking much about work. As you just said, I’ve just joined this little new company as a principal product designer. To me, that’s pretty neat because it shows that I’ve actually been able to stay on the designer path and reach quite a senior level without being expected to go into management.
Donna:
So, there actually aren’t a lot of places where you can stay senior and stay being a hands-on designer. So, I’m really glad that I’m still able to do that because management was never my path. I also love that this little company I’ve joined is called MakerX because I always say I am a maker, I’m make things, which is again why I’m painting and renovating this house by myself and otherwise I can sew and weave, and I’m all always making stuff. To me, designing is still all of that.
Larry:
A big part of that, I know I’ve heard you talk about this before, is that you’re a compulsive organizer too, which is a foundational thing for information and architecture practice and for a painting project, I’m going to guess, too.
Larry:
One thing, you’ve been doing this for a while and I think we’ve both seen information architecture, I remember vividly when the first edition of the Polar Bear Book came out and reading it covered to cover and being excited that somebody had articulated what we do, but then it never … I don’t know, it took off and I think it’s always been a thing, but it seems to have been neglected in some of the past years. But anyhow, so I think a couple things, I’d love to talk a little bit about that, but I’d also love to just get your take, your definition of what information architecture is. Because I think a lot of people, because of that neglect, I think there are people who couldn’t define it for us. So yeah, I’d love to get your take on that.
Donna:
Yeah. Look, I still think of it as a thing that is needed to be done, not necessarily by a person with a role of information architect, but something that needs to be done on most projects where the things we are working with, whether that’s unstructured content on long pages in websites, or whether it’s the data that we are going to be serving up via API or consuming via an API, or whether it’s inputs to an algorithm. It’s all about understanding the ideas that we are working with, understanding the content that we’re working with, understanding the relationships within all of that, and then also understanding how people think about the tasks that they need to do or the information they need to find, understanding where they are and putting structures in place so that we can actually deliver what we need to the people we need to deliver it to, in a way that works for them.
Donna:
I said that word structures pretty quickly in that definition, but that’s really the important part of it, is doing that underlying foundational work of setting up our data or our content well, so that we can deliver it to people in a fairly automated way. But when we first started doing IA, a lot of us were making hand-created websites with simple hierarchies where we had a hundred pages and we knew where every page went in a tree. We don’t work like that anymore.
Donna:
We’re working on large websites, large apps where we are doing a lot of delivering of content automatically, whether that’s advertising, or related products, or showing a catalog, you can’t actually deliver that to people unless you have figured how to structure it first. That was a long definition, it’s not a definition, maybe a description.
Larry:
No, but I think that it’s a great example of why it can be hard to define it because it is so … it’s foundational, it cuts across all this stuff. So if you had to say all those things you just said to make sure that it was all in there, but you also alluded in there to the evolution of the practice, that it used to be a hundred-page website, you just figure out how to navigate it, how to label stuff. Boom, you’re good to go. Way different now, can you talk a little bit about how that has evolved, the new activities or new conceptualizations of things that are necessary because of that evolution?
Donna:
Yeah. I had an epiphany when I was teaching a class recently and I’d been talking all the time about what I just said there with delivering content automatically, and somebody said, “What do you mean by that?” I realized that’s really that evolution of when we started and we were doing a lot of manual work and manual hierarchies and just setting up, even to thousands of pages or tens of thousands of pages of content, but in a kind of manual way. Now, as I said, really what we are doing is trying to figure out how, who work with much larger volumes of information and how to make that happen using mathematics and algorithms and artificial intelligence and machine learning and things and stuff. You can imagine any large content or catalog site, you design a template for something like a content page or an index page and then you write rules around what content is shown here and how. How do the filters work? What is in the filters? What are those categories?
Donna:
So that evolution, from doing things quite manually, to now doing this in a really automated way, I think has … I mean, it’s a different kind of thing that we are working on, but it also is, it’s a different set of skills and there’s a third lens on it where most user experience professionals who are newish to the field have been taught in a really, quite different way, been taught about user needs and they’ve been taught about screens and flows and doing a journey map and understanding how a person gets from one end to the other of a thing. But they haven’t been taught about that structural … how do you make this thing work? How do you get the content out of the bucket and onto the page? So, we’ve got a big gap there in the way teaching has been happening for UX folks, has never helped them understand that there is this real data lens that they need to have.
Larry:
The way you just said that makes me think that … how do you fill that gap, I guess, for the new folks, the folks who’ve come into the discipline because … I think that’s something that, I come out of publishing and I’ve watched … when I reflect on my career, I’m like, I can’t identify the day, but over the last 20 years, I’ve shifted from identifying as a publisher to being a UX designer. But people who came in just as UX designers from the get-go, how do you tease out the need for the underlying structure to inform those flows and screens and all that stuff?
Donna:
It’s really, really hard. What I see a lot on projects and when people are either coming to a class or when I’m talking to focus on a new project is that they don’t know what they haven’t done and they don’t know what’s missing and they’ve done the flows and they’ve done the screens and then some business analyst or developer’s saying, “Yeah, but … ” and they’re like, “Oh.”
Donna:
I was working with a project a couple of years ago where they’re doing something fairly complex where they would try to deliver really good recommendations to people for a particular product. They had a really good, big idea and I’m like, “But where’s that data going to come from? Where are the classifications going to come from? How are we going to make sure that the way that we do those recommendations doesn’t continue to reinforce structural bias?”
Donna:
And they’re like, “Oh no, well, we’ll just paint the categories. We’ll just make them.” I didn’t want to be like the grumpy old lady going, “You can’t just make them. These things don’t just make themselves.” “Look, oh no, it’s all right. We’ll come up with the list. We’ll come up with the things.” I’m like, “You can’t just come up with them, this is … ” but they couldn’t understand how complex it was going to be. I was like, “This is not going to work. This is actually probably impossible.” Really good, big idea. So it falls apart and then what ends up happening is mediocre products get delivered or products that reinforce structural bias because nobody has understood the nature of the categories and algorithms and the way that we bundle up stuff to deliver it automatically.
Larry:
You’re making me think now about the evolution of metadata strategy. It used to be that you just categorized documents, so you could spit them back out, but now you have to categorize at a different level so that you can put it together in a useful way. Is that a correct way to summarize that?
Donna:
Yeah, yeah. So on all projects, everybody has a different kind of lens of what this might be. I’ve never actually worked somewhere where anybody called it metadata strategy, but yes, absolutely.
Larry:
Right, because like you said at some point a product manager or somebody looks at it and goes, “Well, how are you going to do this?” And they’re like, “Well, they’ll just go from this screen to this screen, or they’ll … ” but it’s like, “Well, and how exactly?”
Donna:
How exactly, exactly what is in these categories. Exactly what is the metadata?
Larry:
So, you’re reminding me a little bit of Brad Frost’s Atomic Design, or approaches like that, where you have to articulate what it is you’re working, I guess maybe is that sort of what’s going on here that you have to … The designers are thinking flows, they’re thinking in verbs and action and action- And what you’re working with is things, nouns, objects. Is that the disconnect, do you think?
Donna:
That is a disconnect, and of course it’s like every project is a different style. So, we’re being really abstract here and generalizing, but yeah, a flow is about going from this screen to this screen, to this screen and doing these things, but we are talking about the things and in order to get the things onto the screen, often we need to have figured out how the things relate. So, you go this screen to this screen and you’ve made some selections in the first one, what’s going to be shown on the second one, for example? Or what things are we showing there based on what we know about the person and what the situation is and what the content actually … how it can work. So yeah, that’s a really good summary of one way of thinking about things about flows and steps. I’m always thinking about the nouns and the relationship.
Larry:
Yeah. That’s something, again, I’m thinking out loud here, but one of the big trends, I think we’re all seeing is this, that a lot of this has to do with decoupling, decoupling of content from documents, I think is the thing that’s driving a lot of what we’re talking about. Is that an accurate way to think about it? And what you’re talking about, assembling the right thing in that flow is sort of a recoupling in the moment is that …
Donna:
Yeah, it recouples it in the moment and you write the rules around it to say, “Show these things at this point, based on all of these other things that might have happened.” But I think the interesting thing here is that the skillset to do user research and understand your customer’s journey now and understand a future customer journey and set up those flows, that is a particular set of skills and a way of thinking about the world that really probably isn’t the same set of skills and the same style of brain as setting up the structure of the data so that it can turn up in the right place. Definitely some people can do both. I can definitely do both. I don’t do the, make it look amazing bit, but I can definitely do the flows and the structure about thinking structurally and thinking about content metadata is really different. It really isn’t the same as doing research and doing flows. So all the time we don’t have those skills on a project, the work isn’t happening well.
Larry:
Right. Well that goes back to something you said, right when you’re in introducing your idea of IA, that it’s a skillset. It’s not necessarily a job title, so it can happen in any number of places, I guess. Well, and this gets into … I want to make sure we talk about your role as an educator and maybe this is a good place to talk about that because you’ve done a ton of workshops and courses and classes. You’ve done everything over the years, written a book, multiple books, you know how to impart this information as well as anyone. How do you help a nonspecialist cultivate IA skill?
Donna:
Oh, I still don’t do it well. Well, I mean, maybe people think that I do it well and they’re happy when they leave, but I … maybe I shouldn’t admit this out loud. Whenever I’m teaching IA, I’m feeling inadequate the whole time. I’m feeling like I’m not explaining well, and this is … why am I saying this is so hard and how am I not getting it through? But then I pull myself out of it. I’m like, actually, I’m trying to describe a lot of things in a way that people can apply them to their really specific projects. So yeah, like as I said, in the old days, I taught people how to come up with hierarchies and categories. And now I’m having to teach them these quite abstract, foundational, fundamental ways, ways that humans think and way that content models is how I teach IA.
Donna:
But I just never actually have time to do it well. I often end up doing like one and two day workshops and it’s not … I mean, that’s how education works and professionals often don’t get time to really immerse and there’s no complaining there, but it means I have to land as much fundamentals and basics in people’s heads and hope that they can make sense with the things I give them later on to a project.
Larry:
Yeah and the way you just said, it’s that combination of … Well, it’s that combination of the human behavior and I guess … Well, I want to go back because your book, I mean you were doing this years ago because your book, the first chapter is sort of an overview, but the second chapter or part is about the people part of it. It’s sort of like the journey stuff is people stuff and then the structure stuff is content and data in there. Do people have trouble stitching those together? That sounds like what’s going on here, or part of it anyway, right?
Donna:
Yeah, well to understand the people stuff, you not only have to understand what people need to achieve in a task, we need to understand fundamentally how people think. So, when I teach IA, the core thing that I’ve taught every single class for probably close to 20 years is I always start with teaching category theory. I’m like, “I am not apologizing for teaching you about Aristotle and then teaching you why Aristotle was wrong, but why it has leaked into our whole way of thinking about the world because Aristotelian philosophy believes that there is a correct way of structuring information that’s out in the world and you go get it and you apply it. And actually categories form in our brains based on our experiences, and our experiences are formed by our communities and our language and everything we’ve done. So, the way that categories, and categories and concepts and thought, the way we bundle stuff in our heads, is really varied across different cultures and communities.
Donna:
So we’re trying to look for coherence in a particular user group that we’re working for in a particular way, some way that we can bridge a gap for them of understanding where they are now and where they need to be based on how they think and what they think about. It’s not hard, hard, but it needs work and it needs an understanding that people think about things in really different ways. I think one of the little examples of this in traveling to the US a lot, and traveling to Europe, and traveling to South America is that idea of politeness and the way that people interact with each other, you can see that’s really different culturally. The way that you say, “Yes, I’ll … ” if you offer somebody something in some cultures they’re like, “Sure, fine.” And then some they’re, “Oh no, I can’t possibly.”
Donna:
And you do that whole routine. So you think about that idea and then apply that to everything that we deal with, you can see that we think really differently, it’s embedded deep in our brains and then we’re trying to bridge, some kind of getting people through a task or helping them learn with our content. So you really have to understand all those peopley things and then figure out how can apply that well to the content that you’re working with and that just needs work. It isn’t just a, “I’m going to do some navigation design today. I’m going to make up some half-assed categories.”
Larry:
Well, no, there’s so much in what you just said and it’s like, I’m going to forgive you for thinking you’re not a good teacher when you’re teaching IA, just what you just ran through there in the last two minutes is human psychology and human behavior and the history of philosophy and there’s a lot going on here, but I want to thank you for a couple things. One, you gave me a great click bait thing for when I promote this episode, that “Aristotle was wrong” and we’ll lead with that, so that’s great. But also when you were talking about just those ways of organizing things like, that gets at both, again, some of that research stuff around grasping mental models and trying to get inside your user’s head, but then it also, but that equally applies on the organization and structuring end when you’re doing your information architecture, how have you seen … because personally, I’m studying ontology engineering and ontology practice right now and really seeing benefit in that, whether I ever actually do that, just for my information architecture practice, but there’s other ways, there’s other approaches to organizing the work of figuring out how people think, does that make sense? How do you do that?
Donna:
Oh, I probably do it different on every project, I think. I don’t know that I have a repeatable rigorous way of doing it. I just talked about all that stuff and you can tell that’s all in my head, so because I have that all in my head and I’m good at understanding content and structure, I can put things together. I can just go from that jigsaw mess of things I know about people and things I know about content, to making a leap, to coming up with something, but that’s because I’ve been doing it for 20 years and again, that’s why sometimes I find it hard and I feel inadequate teaching because it’s hard to go backwards and break it down into a process.
Larry:
Right, I mean, that’s how I learn, is I teach, because then you have to do what you just said, you have to break it down and make it learnable. But I guess if you could, when you reflect on all … and also what you’re saying about, like when you were talking about just the human variety across the globe, and I’m like, “Holy crap, that’s a whole other episode,” just about localization and information architecture and all that. But ultimately that’s about the basics of communication, like knowing your audience or your users in a UX practice and then addressing it to them. So thinking of, if you can, the most generic persona you can conjure up in the moment, what would be the top two or three take-homes you would hope that someone would leave your courses with, about the practice of information architecture?
Donna:
I think that the couple of take-homes are that we think in different ways. If you can at least understand that actually thinking deeply is different across cultures and communities, that’s amazing. That’s amazing to go, “Oh, right. We actually don’t think even similarly. Good, I need to know that and then I can apply it.” So, there’s just learning, that is good. I think the other real takeaway when I’m trying to teach is to help … and you and I talked before about talking about content modeling.
Donna:
The other takeaway is to help people understand how to go about trying to understand and model or represent the thing that they’re working. So, I always get people to do an activity where … I usually say, “You are working for a games store,” whether that’s an online game store or a face-to-face physical game store, it doesn’t matter. Think about all of the kind of things that they sell. So, here game stores will sell board games and puzzles and Dungeons & Dragons, books, and figurines. Then if you add computer games in, then there’s games and consoles and stuff. So, you think about all the things that they might sell and then I get them to start drawing a model of that. I encourage them to think about the relationships between things and I usually prime them with some examples like, “Not just board games and puzzles, but what do you do about brands like Marvel?”
Donna:
So, you might have a Marvel board game, you might have a Marvel puzzle, you might have a Marvel digital game, you might have figurines, you might have something else. Or how do you go about thinking about other aspects of that? The thing that works really well in a workshop setting there is people start with things that are the nouns, the boardgames and puzzles and then they start thinking about what are those things about and how do we represent how they would relate to each other?
Donna:
Then they tie themselves in knots, trying to figure out what the hell they’re doing anyway and why did I ask them to do this? Which is exactly what they need to … that’s the process they need to go through, to go, “Oh, hang on. What are we doing here and why?” When they walk away with that going, “What are we doing here? Why? Okay, now I need to think about my project and what do I really need to do? What do I need to understand? What are the content objects we’re working with and what relationships do we need to know about?” Because it’s easy to make a … no, it’s not easy. It’s easy to accidentally try to make a model of everything you know, and to blow it out. So you pull it back and say, “For my project, for the things that we’re trying to achieve here, how much do we need to work with?” So if people walk out our workshop with those two things and enough of like, “Okay, cool. That was interesting. I have a feeling that I’ll try that on my next project.” I’m good.
Larry:
Great. I hope this helps people because I see a lot of need for this. One thing, you’ve mentioned the word model several times, and it occurs to me when I reflect on this conversation that I think a lot about content modeling and domain modeling but also customer journey mapping or user journey maps, that’s modeling a journey, or modeling … it’s like everything we’re doing, we’re mapping mental models to understand people, or empathy mapping is trying to map the feeling part of the brain. So do you see modeling as a core skill for information architects?
Donna:
Yep. So, yeah, I actually have a degree in economics and one of the things that we just learned really early on was about making models of the world. That’s what economics does. I was doing a talk a couple of years ago, and I realized finally that other people don’t know what models are. I’m like, “Oh, okay.” What we’re trying to do here is go from a messy complexity of the world and represent it in some way so that we can use that to make decisions, or to talk with other people, or help others understand what we are working with and so that we can use that to actually build our projects.
Donna:
So being able to model, whether that’s content modeling, or domain modeling like you said, or journey maps are models, because you’re trying to figure out how to represent something that’s actually pretty complex and that probably isn’t linear and that everybody does differently in a way that you can communicate and find places in there to do various things. So, these are all models. I think it’s a pretty fundamental skill and I reckon it’s probably not taught like that. It’s probably taught, like you should do … I teach UX and in the courses that I teach, which I don’t write the content for, it’s basically like go and do a journey map, but nobody actually … they don’t talk about what is happening underneath there.
Larry:
Right, I guess that’s what got me thinking about it is that … Anyhow, I was just thinking out loud there, but it seems like that taking these skills that we apply and just applying them to each part of that, each of the things you’ve talked about today. I can’t believe this …
Donna:
Well, I was talking to somebody the other day, somebody in one of my classes just about different thinking styles. Somebody said to me that she had just been diagnosed with ADHD. We were talking about just different thinking styles and some people can’t make models. To some, doing that and obviously I’m not try to draw a bridge between ADHD and making models, we were just having a general conversation about different styles of thinking of what works. Some people can’t do it. They like, “Okay, but what are we doing here?” And some people are really good at taking something abstract and turning it into a representation.
Donna:
So, if it’s not your skill, then that’s okay. Find somebody who is good at doing making models and frameworks and people who are good at making process and understanding how to abstract a process can be good at this as well.
Larry:
Great. That’s perfect. Hey, I can’t believe it. We’re coming up close to time already, it always goes so quick. But hey, I want to make sure, if there’s anything last, anything that’s come up in the conversation that you want to make sure you follow up on, or just anything that’s on your mind that you want to share with the folks. I want to make sure you have a chance.
Donna:
. . . all the things that are on my mind revolve around paint and plumbing. So it’s worth me mentioning that the book you mentioned, the non-card sorting one, it’s called A Practical Guide To Information Architecture and it is on my website for free because my distributor decided not to sell it anymore and gave me no notice. And I’m like, “What am I going to do with this?” It’s old, it’s still good. You can just have it for free. Sometimes people ask me if they can throw some money at me. I’m like, “I have no way of doing that.” Go buy one of my other books at the same time. So you can have a free book and go buy another book, buy my one on presenting design or something as a way to pop some money in my bank. But yeah, it’s worth people knowing that book is there just for having. I wrote it just before we really started working on mobile, so it’s got some huge gaps, but it’s still got lots of good fundamentals and they can learn about why Aristotle was wrong.
Larry:
I flipped through it again, in the last couple weeks and it’s still really good. It holds up really well. And it’s worth way more than free. So, I’ll include links to the other books to make sure that other people have a chance to support you as well. Hey, one last thing, Donna, what’s the best way for folks to follow you or stay in touch on social media or online anywhere?
Donna:
Probably LinkedIn is best. So I use Twitter, but just not a lot anymore. I’m fairly active on LinkedIn. I’m there a lot. I do get a lot of request connections. So the thing I do ask people is please say to me, “Hey, I heard you on Larry’s podcast and I’d love to connect,” because that means I like actually connect. The other 300 in my inbox, I’ll do it one day. You can find me easily on LinkedIn, just as Donna Spencer. It’ll be LinkedIn, whatever, whatever, Donna Spencer. I can send you the …
Larry:
I’ll put the link in the show notes as well.
Donna:
Well, the other places I’m Maad Donna with two As, like on Twitter and other places.
Larry:
Oh right, yeah. M-A-A-D Donna.
Donna:
M-A-A-D.
Larry:
Okay, yep. I’ll make sure that’s link too as well. Well, thanks so much, Donna. I really enjoyed this conversation, super fun.
Donna:
Thank you very much. Really good to talk to you as well.