Information Architecture means different things to different people. Chris and Carla attempt to untangle this confusing term whilst freezing their buttocks in Liverpool Street station.
The Design Untangled Podcast
Episode: DU008 – The Grand Design of Information
Host: Chris Mears and Carla Lindarte
Length: 18:14 minutes
January 22, 2018
[00:17] Chris: Welcome to Design Untangled with me, Chris Mears, and for once sitting next to me in real life is Carla Lindarte. Hello.
[00:25] Carla: Hello! How exciting is this?
[00:27] Chris: Yeah. So to set the scene, we’re sat on a very cold seat in Liverpool Street Station in London.
[00:35] Carla: At lunchtime.
[00:36] Chris: Yeah. At lunch time. So lots of hungry commuters wandering around. So hopefully I’ve worked the mic out properly this time and we won’t have to trash this as usual.
[00:47] Carla: I hope so, because it’s very cold in here. I don’t want to do this again. Alright, so today we wanted to do something a little bit different, be face to face and think about something that I’ve actually been talking about a lot recently, which is information architecture and navigation. So Chris, tell me what do you think are the differences between information architecture and navigation?
[01:13] Chris: Well, navigation is just basically the way you help people get through your information architecture, I think. So in terms of what information architecture is, it’s kind of how you structure what you’ve got in whatever you are designing. So navigation is one route into that, but it could also include kind of how you get to it through search and other methods like that as well.
[01:39] Carla: Yeah, exactly. So a lot of people use those two terms like they were the same thing, but they’re actually very different. Obviously, the information architecture informs the UI, the user interface, which is the navigation, the way you actually deliver that. However, they’re not the same thing. There is someone, I can’t remember his name but he talks about that navigation is the tip of the iceberg, and the information architecture is the whole island, and so as a UX designer, you really need to think about how the content is structured, whether it’s a task-based navigation or a content-based navigation persona…. Oh, sorry… Task-based information architecture or content-based information architecture, because there are websites where people are going with a particular job in mind. So they’re more like task-driven, whereas there are websites where – if we are talking about only websites – they are just content driven, which is like the company trying to push different content models for people. So you need to think about those two things and how to balance them in the structure of the overall site.
[02:47] Chris: Yeah. So that’s not to say that you can only use one type of those navigations per website. You may, depending on what type of personas you’ve got, have one form of navigation for the people trying to achieve a task and another for the ones just kind of looking around for certain types of content. So it’s a couple of different techniques we can talk about for how you come up with it.
[03:11] Carla: I mean there is a big difference between what are the users trying to achieve with your website or product, versus what do you want to publish as a business or what do I want to have to say, and I think as you said, you have to find the balance between the push and the pull, because those two things are different things together. But the main fact for us as the UX designer is to make sure it is user-centered, and for us to be able to make it user-centered, there are some techniques, so you could start by identifying the initial hypothesis or what you think those primary tasks are.
[03:55] Chris: Well, I guess even the starting point for that could be once you map target, you use the journey, right? That’s your first kind of clue as to what people are trying to achieve.
[04:04] Carla: Yeah, exactly. And from there you can pull out of that, what are the primary tasks that my users coming into the site? So card sorting is a way to include the customers or users into the definition of information architecture. You could either do open card sorting which is getting the users, once you come up with the same primary tasks or content types, you get the users to group them in the way that they think is logical, and they can also create a label or a name for that particular group of content. But it is very open and you just do it in more like a face-to-face exercise. It would be much better because as a UX designer, you can listen to what people are thinking and you can understand a bit more about the mental models, although there are online ways to do that as well.
[04:56] Chris: Closed card sorts. So that is primarily a way of testing, I guess, your initial design of the navigation sites to make sure that your structure maps with where people would put things in that structure, so it’s typically the top level categories. You would give people cards which represent the different types of content and then ask them where they would put that. Again, it’s useful to do that face-to-face so if they can’t find a place for it, that’s a clue that maybe your labels are a bit wrong and/or they’ve got a different way of thinking about things. So it’s useful to just probe their minds a little bit about that, and get them to actually write down what category they would put that in.
[05:45] Carla: Yes, exactly. I mean you could go really scientific about this card sorting stuff and in the past, I’ve run projects where you use some Optimal Sort, which is that online tool that you can run card sorting. It’s good to have some quantitative way of how people think about grouping and task and content. However it is very much more valuable in my opinion, to do face-to-face, even if it’s with a smaller group of users, the face-to-face, the conversations where they are actually going through the groupings and they ask questions to each other and it’s really insightful because that not only helps with your information architecture but also, what’s the most important content that you want to have on the site and all the solutions. So I think you could do both, depending on how much time you have and your budget, but I would recommend really doing the face-to-face stuff.
[06:35] Chris: Yeah. There’s quite a popular Jakob Nielsen thing around card sorting that to be statistically significant, if you care about that sort of thing or your stakeholders do, I think it’s five sets of five users right before your trade off starts dropping off in terms of what you learn. So yeah, if you can do 25 people, great. If not, as Carla says, then just being able to interact and speak to people can be very valuable as well.
[07:05]Carla: Yeah. I’ve also done , in the past I’ve actually, if you’re working in an office or somewhere when you have other people coming in and out, I’ve actually left the different cards on the corridor and then sent an email to the people saying, did you want to help us group this content and stuff? So if you don’t have the time to do it face-to-face, or you don’t have the money to recruit users, you can still do it with people around you. Some research is better than none, so doing that is very valuable. And I think more and more, especially with time frames, I’ve found the tree deck, which is reversed card sorting, it’s how they call it, tree testing or reverse card sorting is actually very good because in a very short time frame, you can test all your initial hypotheses. So let’s say you only had a little bit of time to do some research and card sorting with users, you then have the opportunity to use a tool like Optimal Sort to test your initial hypothesis on the structure of the site and obviously you need to focus on where the more problematic areas are and coming up with the right tasks to get people to find the content where you place it, but yeah, it’s a very, very powerful tool because I’ve run projects for app development as well as website development where we could test two, three different iterations of the IA that we’re looking at and the labels as well, and then you can get insights very quickly and see which one is performing best than the other.
[08:40]Chris: Yeah, so in terms of thinking about how IA applies to search as well, the other thing you have to think of is not everyone is going to navigate through your site or whatever for a menu or drop down. They may well search and want to refine that search, so I guess how IA proliferates itself in search results is through filters. So that’s another aspect of kind of grouping information and content that you have to understand because some people will prefer to search rather than browse. So don’t just be constrained about thinking about your mega drop dropdowns. You need to think about the different routes in people will have.
[09:18] Carla: Yes, exactly and different levels as well. I mean mega drop downs, I’m not a fan of mega menus if I’m honest because I think sometimes, especially when you don’t have a very clear governance process in place, they become a massive monster that is very, very hard to maintain and there’s so many options. The more options to give to a user, the harder it is for them to make a decision. So I understand why, for example, retail companies and stuff like that have them, but at the same time I think you could come up with another way to people discovering your content that is not necessarily these massive mega menus. However, I know they are used across multiple sites. If you think about website design, you could simplify it
[10:02] Chris: Another kind of way the IA sort of serves as you might not necessarily think about is keywords as well. So again, it’s kind of search methods that people might use to find content. But you have to think about do you have the right keywords in your content to enable people to search for and find what they’re looking for? So it does dribble down to even your content design and what you’re producing on the actual page or whatever it is.
[10:30] Carla: That was really, really important. I was having a conversation with a content strategist. I might get her to do an interview for us, but she was saying that even in order to inform information architecture, you really need to look outside of your website. So you have to look at SEO. You have to look at social media listening, all these tools that allow you to understand what people are thinking about your product, what they are actually talking about, and those keywords as you say will inform the taxonomy of the site, with obviously information architecture and taxonomy are so related because the groupings and the labels and the content structure that you use for your site is a reflection of your taxonomy. So that is also something that as UXers, we might not get too involved in taxonomy, but it’s really important that we understand that concept as well and work with content strategists and obviously tech people to see how you deliver a search.
[11:28] Chris: Yeah, and the term itself, used not so much these days, but you do still see it occasionally as it used to be, it used to be what UX designers called themselves, but now I think if someone rocks up calling themselves an information architect, you would expect them to be kind of very deep in this information structuring side of things, rather than necessarily doing wire frames and stuff like that. So I think it’s kind of changed from being a generalist term for UX before UX was really a thing in the consciousness to now. It’s actually a very specialist kind of role.
[12:02] Carla: Yes, that is true. Also just made me remember that very recently I’ve been interviewing a lot of UX designers, junior UX designers and they all want to drop the word card sorting into their interviews, and the reason why I’m saying this is, Oh yes you can do card sorting and I think if you haven’t done it before, it is better if you really researched about it cause it’s not as simple as it could be, especially when we are doing website redesigns, you didn’t need to look at where to focus. You can fix old problems of very deep information architecture with one card sorting exercise. You could actually be doing more and more depending on the different areas. So if you think about banking for example, think about mortgages is very different to current accounts, very different to everything else. So when you talk about card sorting you can say, I haven’t used it before but this is what I think it is, but a lot of people use it as well. Oh yes, you do card sorting, but it’s not as simple as it sounds.
[13:05] Chris: It’s about kind of fixing your efforts, so the example you just used, it might be a yearlong project to sort out the IA for the mortgages section of the banking site.
[13:15] Carla: Yes exactly.
[13:15] Chris: So you can’t solve all the problems all the time or so you need to show that you’re aware of that and how UIs can prioritize stuff and where you think is most important to fix things.
[13:27] Carla: Yeah. Also something that we need to bear in mind as well is that you don’t necessarily have to have all the answers. What I mean with that is that you could have a lot of hypothesis and still build the site, but if you build it in a way that is flexible to either change the labelling or change the structure, I think it’s still okay to go. For example, in banking, there’s always this conversation about do we go needs first. So for example, do we talk about home buying, or do we go product first, which is do we talk about mortgages? And there’s not a right or wrong answer, because it depends on the mental model of someone coming into your site. So you would potentially try and test, and try to A/B test if you can, proper A/B testing with multi-variant tests, and look at what’s performing best, because maybe people with a task in mind looking for mortgages and they can’t find them, but maybe someone who’s just looking for advice on how to buy a home. So what I’m trying to say is you don’t have to have all of the answers, especially if you’re just designing the site. You have to have clear hypothesis of what you want to learn with both approaches
[14:37] Chris: Yeah, and in terms of how you come up with those approaches, right off the top of my head, if you’ve got a site with decent traffic, one thing you can look at is the search results page and where people are just abandoning after running a search and not clicking through to anything. That’s maybe a little clue that something’s not quite right there in terms how things are labelled and structured.
[14:5] Carla: Yeah, that is true. Very recently with one of our clients, there was a lot of talk about whether people think about remortgaging or they think about switching to a new deal, so we had to do a lot of research to actually understand how people think about that. So the bank will think about it’s just a switch, one deal to another. So from a bank perspective, you think that the people are just remortgaging, whether they remortgage with the same bank or with another bank. So just doing some qualitative research as well is really important, like just in-depth interviews with users to see how they think about things, especially at the beginning of your redesign of information architecture is really useful. So you can use those insights to inform the way you structure the site or the section that you’re working on.
[15:51] Chris: Yeah, we’ve been going through a remortgage recently and it’s important to keep your thinking wide because we just used a mortgage broker, which is obviously another different type of user that may need a completely different IA structure and different kinds of information that they need to find. So yeah, as a ways understand your users, understand their needs, understand what they’re trying to accomplish and understand what they’re trying to look for.
[16:16] Carla: Yeah. Good. Do you have anything else to say about the wonderful world of information architecture?
[16:20] Chris: Probably, but my ass is getting freezing on this chair. They’re like these horrible metal chairs and they designed stop homeless people sitting on them for too long, I think.
[16:33] Carla: Yeah. We have to find a better place to do this next time.
[16:37] Chris: Well hopefully the recording is useful. So I’ll let you do the plugs this time.
[16:41] Carla: Oh, I can’t remember all of that. You are the only one.
[16:45] Chris: Twitter, website, email.
[16:45] Carla: I can’t remember them.
[16:47] Chris: Personal LinkedIn messages to you.
[16:49] Carla: Yeah, just send me a personal LinkedIn message. No, I don’t remember all the different things.
[16:58] Chris: All right, fine. So we are at @designuntangled on Twitter, individually you are @carlalindarte. I’m less easy to remember @Chris_mears_ux. We’ve got a website called designuntangled.co.uk You can email us at email@example.com Send Carla personal LinkedIn messages by searching for her, and that’s all of it, isn’t it?
[17:25] Carla: That’s all it is, yeah.
[17:25] Chris: That’s all our stuff.
[17:27] Carla: So see you next time. Please give us any feedback or topics to talk about. If you want to learn more about card sorting, for example, we can do one just about card sorting, but yeah, if you have any questions, just send us a note.
[17:40] Chris: If you’ve got any ideas of different locations which are really noisy, we can record these in maybe like a car wash or something. It will be good for the next one.
[17:48] Carla: Probably really good, actually. A car wash would be very, very interesting. Maybe in the summer though. Yeah. All right. Thank you
[17:56] Chris: Cheers
[17:56] Carla: Bye.