Yesterday I had the pleasure of hosting Donna Spencer in our Ask the UXperts chatroom. Donna is the author of our recently released ebook titled A Practical Guide to Information Architecture, and funnily enough, Information Architecture was what we were chatting about.
The premise behind Ask the UXperts is one expert, one hour, all your questions answered. Donna did a stellar job of staying on top of the questions, and the session was an informative one.
A number of resources were mentioned during the hour, and to save you having to sift through the entire transcript to find them, I’ve distilled them into a list.
- A Practical Guide to Information Architecture by Donna Spencer
- Donna in action in our latest forum debate
- Donna’s agency, Maadmob
- An example of a page-level content template
- More content templates
- The UX Mastery Careers forum
- An episode of the Freakanomics podcast, discussing mind-reading amongst other things
If you missed Donna’s session because you didn’t know about it, make sure you join our community to get updates of upcoming sessions. If you’re interested in finding out exactly what went down yesterday, here is the full transcript for your reading pleasure.
So… for those of you that don’t know who Donna is, she is one of the stars of our latest debate http://community.uxmastery.com/forum/news/1620-… | |
She is also the author of our latest ebook call A Practical Guide to IA http://uxmastery.com/practical-ia/ | |
Yay Donna. Just purchased the book and it is very helpful | |
And she runs the digital agency Maadmob http://maadmob.com.au/ | |
I agree, especially the chapter on IA Patterns | |
Great! | |
So Donna, I’ll throw it over to you to introduce the subject, and then we’ll open the floor to questions. Feel free to jump in at any time and I’ll queue the questions if things get busy | |
OK. As a very brief overview, IA is all about organising, structuring and labelling content | |
Often for websites/intranets. Also for business apps, web apps, mobile | |
Over to questions – that’s all the intro we need :) | |
So Donna, to kick things off with… what is the link between IA and UX? | |
Haha. It depends on what your definition of UX is… | |
Give us yours :) | |
You can’t design a good experience without a good information architecture | |
Isn’t “it depends” the answer for everything UX-related? ;) | |
But the person who ‘does the UX’ doesn’t ahve to be the same person who does the IA | |
So the IA is just one of the very important parts of designing an overall good experience | |
Ok, makes sense. I’m going to let someone else ask a question now. | |
Hey Donna. This is a bit of a vague one, but I’m wondering about when you can rely on first-principles in designing an IA (eg, don’t have 500 items all that the same level in a drop-down menu) and when you need to bring in the research. What are the principles you use to make the call? | |
You can rely on first principles for all kinds of basics – like, as you mentioned, not designing enormous lists, not making deep hierarchies. Where you need research/users/people is to refine and make sure it really works for them, for their context, for what they know and for how they use terminology | |
So first principles can get you a long way towards a good outcome, and research will make it really work well | |
I think a lot of times people use research when they should be using first principles, but without those principles, research will get you some of the way as well | |
If you are fairly confident with the domain, content and users, you can draft an IA without research; then test it with users – and that’s how I do most of my work. | |
What are some of the key mistakes that people make when approaching IA? | |
one of the key mistakes people make is paying too much attention to the users, and not paying enough attention to the content. I’ve seen many people go off and do a ton of user research, make the IA around it, then find out that the content doesn’t fit | |
Amen! | |
Second key mistake is to spent far too much time thinking about how they might do the IA (audience, subject etc) instead of just giving it a go | |
Something you’ve come up against Matt? | |
Totally! | |
Typically when a site has a nice homepage but then the inner pages are just big long text dumps. That’s where page-level content templates are super handy. | |
Yuck! | |
And also why you need an interplay between whoever’s doing the design and a content strategist (or a copywriter who understands basic web design). | |
Matt: do you have an example of those templates to share with us? | |
Something that ties those two questions together – sometimes people do the IA to the first and second levels, then don’t go further. | |
Hang on… | |
You need to work out the IA to a quite detailed level as it’s not the first couple of levels that are the hard part – it’s the nitty-gritty detail | |
In what way Donna? | |
Here’s one I did for a community service provider. Being a word doc, it’s kinda ugly but the critical thing is that this format is something that the client could edit easily. Also, having this template meant that writing a page of content took massively shorter coz I wasn’t reinventing the wheel each time I did a new page.https://www.dropbox.com/s/w10sjt18tnze8xx/Progr… | |
It’s actually pretty easy (if you are doing a hierarchy) of coming up with broad groups for a set of content/users. It’s not much harder to come up with a next level. But when you start working with detailed content, it can be really hard to see the dividing lines between the finer categories – for a big site, that is. IA isn’t actually hard at all for a 20-page site :) | |
For more on content templates see http://alistapart.com/article/content-templates… | |
Many thanks! | |
Donna – what is the average time frame would you say for developing an IA? I know that’s like how long is a piece of string but what about a large site (1000s of pages) and many stakeholders? | |
Totally a piece of string. It’s usually not about the amount of content, but about how hard it is conceptually. Organising an online grocery store is going to take you a lot less time than doing something complex like medical content – even though the volume is entirely different. | |
Ditto–thanks Matt F for the A List Apart link and the word doc example. | |
And selling the idea takes longer if you aren’t certain about it, and don’t have a good rationale for why you’ve done various things | |
No worries guys:-) | |
It also depends how well you know the content – if you’ve been working with it for a while it will take less time than if you’re coming in as a consultant and have never worked with it before. | |
So, I can’t give advice on how long it will take. I never even know, even though I’ve done a ton of projects | |
So would you say that it was an advantage for the person that does the UX to do the IA? | |
Not necessarily. It is much more about individual skills, and particularly how good someone is at managing detail. Someone who is more conceptual is not going to be as good at doing IA (which, really is a very detailed job). But the two things are so intertwined that if they are separate people, they have to work together | |
That’s an interesting idea; that the conceptual person isn’t going to be as good at IA. I always thought the exact opposite. Can you clarify? | |
Hi, I don’t know that we got an answer to “what is the definition of UX”. Could we go back to that maybe? | |
To do IA, you really have to be able to pay attention to detail, and work at quite a detailed level. It’s not just about making top level categories. Often it’s about defining a detailed hierarchy, or a taxonomy for a product set, or a set of topics to make sure related links work. That’s detailed work. Someone who is very good at concept/strategy may not have that brain for detail. Of course, the IA isn’t just detailed work – it needs a good understanding of strategy/concept as well. The perfect person has both, but that’s a hard find. | |
Can you explain the detail part, for a site with 3000 pages would you create a taxonomy down to the individual page level? | |
ahh, makes sense | |
Definition of UX, in a few words :) The user experience is the total of all the experiences a person has with a product/system/service. And what we usually want to do is to set in place everything we can to make sure that the experience is a good one – which needs an understanding of people, product, content etc | |
What are some strategies for defining the IA when you have a lot of the same type of content? | |
Danielle – it does all depend on what you need to do in your project. But a lot of the time, yes, you do need to define where those 3000 pages are going to be in a hierarchy, or how the 3000 pages are going to be tagged. Someone has to do it, and content authors/product owners/BAs aren’t necessarily going to be the best people to do it. | |
Hi Donna, this is Becky from Washington DC, where it is still June! I’m returning to the workforce after 13 years as an at-home parent. I’m a former indexer; and this autumn I’ll be a Wayne State Univ. MLIS grad student (online cohort), specializing in information organization. I’d like to find some freelance projects in information architecture that I could do from home to build IA-relevant skills. What are some tasks that agencies/employers, might find helpful from someone who has to work from home? Tasks that I was thinning of include taxonomic work (including thesaurus construction), and content inventory work/ consolidating user research/persona notes? (This is a question for Donna and anyone else who might have ideas–thanks in advance!) | |
Donna: VERY good point!!! | |
Julia – usually, if you have a lot of the same type of content, you are going to be looking at figuring out the topics, or what the content is ‘about’ so you can help people narrow down just to the subject/topic they are most interested in. | |
Thanks – we do have subjects/topics to work with | |
Becky – that’s a really hard one for me to answer. My experience is that anything to do with jobs, getting work, starting in a career, is that it is very different country to country. USA is very different to Australia :) So I’m not much help with that | |
Understood, thanks! | |
Did I catch everything? Any follow-up questions to my short answers :) ? | |
Becky Wolsk: Have you considered posting your question on our forums? We have a large number of members from the US. | |
That would be the appropriate forum | |
Thanks Hawk, that is an excellent idea! I should have thought of that. Oh–and thanks for the link to the proper place! | |
Pleasure :) | |
I suspect the answer will be test :) but what do you think of audience based high level categories versus any other way? | |
I have a better answer than that :)… | |
Audience-based categories are very, very hard to get to work well. In order for them to work, the user needs to be able to look at them and say ‘yes, today, I am that person’ (which isn’t always easy – I have tons of examples where it’s really hard to figure out who you are today). And the content needs to split nicely across the audience groups – which it often doesn’t (cont) | |
Often an audience scheme seems like it will be the obvious answer, especially for organisations who are arranged around services for different groups (personal, business, government etc). But a lot of the time the content that is being provided is relevant to more than one audience. If you find this, it’s an indication that an audience scheme isn’t right for you (…cont) | |
So the best way to resolve that is to create a draft audience scheme and a draft of another type of scheme (subject/topic) and then test both. In almost all projects where I’ve done this, a subject/topic scheme works out best | |
Becky W: Not a direct answer, but a lot of orgs are supplementing their IA work with online tools such as Optimal Workshop. Applying your experience/intelligence to help them get to the essence of the data insights could be an activity that would work from home. You could also put up ads on elance.com etc for the kind of work you’re mentioning. It might be hard to get some traction initially if just working from home, so at least consider getting your foot in the door somewhere one or two days a month/fortnight. | |
Question: I tend to do mainly IAs for smaller sites (so less than 50 pages), which means that my average client doesn’t even know what information architecture is. I sell it to them in terms of ‘making information easier to find’ and ‘creating a smooth journey through the site’, but any other thoughts from you guys how you package it up for people who don’t have existing literacy would be fantastic. And selling it to web designers as well who often have a straight design background. | |
Sometimes the best approach is to have the main approach as a subject/topic scheme, with a couple of special pages for particular audiences to find the things that are just for them | |
Great answer thanks. I thought this was the case (particularly with content overlapping audiences) but for some reason different people keep suggesting it. | |
Thanks Alex, that sounds like a useful approach! | |
Danielle – yes, it really does seem obvious, so always comes up. Which is why I actually have an answer for that one :) | |
Becky W: try approaching some local web design agencies. They’ll often be interested in people who can provide surge capacity – especially if it’s in something like generating taxonomy that they may not be confident in. | |
Donna Spencer: Any tips on what to do when the sheer volume of content is such that it would be impossible to audit everything? I’m thinking, big corporates or universities, for instance, who often have multiple “intranets” and thousands of sprawling pages? | |
Matt – that’s a good enough approach for folks who don’t know. Another way is to find some way of explaining that the real users don’t know the stuff that the organisation is providing (which is why they are there) and one of the reasons to do some special work (research and testing) is to make sure you’re presenting it to people in a way that makes sense for them. (…cont) | |
Matt F: will do–I am near Northern Virginia, which has many local agencies, especially near Tysons Corner. Thanks! | |
I often tell people about the principle that I’ve forgotten the name of – where once you know something it’s really hard not to know it. That’s a good reminder that the users may not have any idea of what the ‘stuff’ is about and organising it the way the business knows it may not help | |
Becky Wolsk: Good luck. :) | |
Donna Spencer: The Baader-Meinhof Phenomenon | |
Donna – There’s a brilliant demonstration of this on a recent Freakonomics podcast which mixes Queen, back-masking and pot: http://freakonomics.com/2014/03/13/its-fun-to-s… | |
Matt M – content audits for enormous things like universities. Ive done this a lot. I usually do a high level audit first and then poke around some of the deeper parts of the site to get an idea of what’s around. There will always be some content that’s really important to lots of people, and some that is only of use to a couple of people. I’d take a more detailed inventory of the important stuff (e.g. all the student-focused and subject-focused stuff for a university) and just a shallow list of the obscure | |
Donna, on your point re selling research. You’re probably right, I guess I’ve always been reticent about asking my clients to pony up and pay the extra. Next time this situation comes along, I’ll take a good stiff drink and do my darndest. | |
Thanks Donna. My experience is that the “find some way” is often difficult in large orgs where there is vocab that is ingrained. I guess that’s true of any change, whether it’s to the website or other parts of the business. But any tips on getting buy-in on something like “I know you call it that, but no-one else understands what it is”, or “I know that page is important to you, but nobody has visited it in 3 years” while minimising ruffling of feathers would be welcome :) | |
Matt F – I usually say something flippant like “Well, I can make it up and have no idea whether it’s going to be good; or we can do a little bit of research and then at least we’ll know whether I’m wrong” :) | |
I like it | |
@ Donna, do you take the research back to your user group or to the client? | |
Nice! | |
“I know you call it that, but no-one understands what that is”. One good way is to show search terms – show that people are calling it something else. Ot do some usabilty testing and show that people looked in an entirely different section. But you’ll never win it as an ‘I think’ /’you think’ argument. Someone who knows the terminology and understands it is going to have a lot of trouble understanding that people don’t get it. I see this happen even when I work with a client – I start understanding things, talking the lingo and forget what it was like not to know. Working with a telco last year did that – I started talking about plans and prepaid and forgot that people don’t understand these :) | |
Julia – I take research to whoever needs to learn from it :) | |
Donna Spencer: What are your recommendations about FAQ pages? Are they useful as a way to redirect users to content, or just a symptom of a lazy or difficult IA? | |
Great question @hawk | |
OMG – my favourite topic. FAQs!!!! I think FAQs are almost always ‘we didn’t know what to do with this, so stuck it all in a page’ )…cont) | |
The time when a FAQ is OK is when it really is a set of FAQs – like, for example, a knowledgebase, or for something that people need to know when setting up software – where there really are things that people actually ask. <…cont) | |
But even in those situations, the answers should also be embedded in the regular content in a good place in the IA – and a lot of time people only put the answer in the FAQ. That fails everyone who goes looking in the place where they think the answer should be. | |
So, in most cases, FAQs should be removed and replaced with real content :) (LOL – that one is always contentious) | |
So would a good compromise be to create an FAQ page that acts like a TOC and redirects people to where the content SHOULD be, thereby serving both roles? | |
Maybe, but only if they really are frequently asked/needed questions. It should be short enough to scan too – some of them around are insanely long | |
Cool. Thanks. | |
So we have about 5 minutes left in the session. If anyone is sitting on a question, now would be a great time to ask. | |
And everyone who has written an FAQ now walks out thinking I’m terrible :) I’ve had the funniest reactions to that one in workshops :) | |
You’ve been pretty legendary as far as staying on top of things goes, Donna. | |
Thanks so much Donna & the good people at UXMastery. | |
Love the FAQ response as well. We tend to accept FAQs as drafts to the site and then rewrite them to where they should be :) | |
Before you all head off, make sure you check out Donna’s book | |
A Practical Guilde to Information Architecture | |
Good on you Danielle :) I can see why content authors write them as they are easy, but not a good way of making content findable | |
And if you want to see her in action, check out our latest debate http://community.uxmastery.com/forum/news/1620-… | |
We should have a FAQ debate :) | |
If you’re up for an FAQ debate I’m happy to find you a sparring partner! | |
And our next chat will be in a couple of weeks on Content Strategy | |
I’ll send out an email :) | |
Have a great day/night everyone |
