The Product Design Process

Organizing Top Tasks

Paul Boag

Paul Boag

The Product Design Process

Check out a free preview of the full The Product Design Process course

The "Organizing Top Tasks" Lesson is part of the full, The Product Design Process course featured in this preview video. Here's what you'd learn in this lesson:

Paul discusses the importance of organizing information architecture and navigation in applications. He explains how to identify top tasks and organize them using open and closed card sorting.


Transcript from the "Organizing Top Tasks" Lesson

>> You've cleaned your results, all the tasks that you want a user to do. We know which ones are the most important to users, the top third or so. You'll be able to see the sudden drop-off. So now we need to organize around these, right? In our application, and actually in our applications that can be quite tricky.

I find that a lot of applications don't put as much attention into their information architecture and navigation as more traditional websites do. And so you have to be very careful and consider how you structure and how you place information. So a classic example is, in a lot of apps, the settings section is where functionality goes to die, it just gets filled up with all kinds of crap.

So there's lot of thought that needs to go into that. And a lot of thought needs to be done around things like your top-level navigation in particular, those immediately available options that are available. And that's because if people start off right, they get that first click correct, then they're much more likely to be able to navigate to the feature or piece of functionality they need.

So spend a bit of time making sure that those top tasks are instantly in people's faces and visible. So there are all kinds of ways of working on information architecture and I'll go into this in a lot more detail in the course on research and testing so I will skim over it now.

But certainly look at things like using Open Card Sorting and Closed Card Sorting when it comes to organizing and structuring the information on your website. You also consider Tree Testing as well. There are some great tools out there that make it really easy if you've never seen Tree Testing before then, sorry, Card Sorting before is basically you just ask users to take those top tasks, those main things people want to do and group them in ways that make sense to them, right?

So instead of just organizing all your features and functionality in ways that make sense to you, we get users to do that and they kind of group it together in a way that resonates with them. And you can, there's some great tools out there, I tend to use one called UX metrics to do this kind of work cuz it enables you to do card sorting remotely.

I would encourage you to checkout my course that covers this kind of thing in a lot more detail. But yeah, you can go in and you can easily create a new card sort and you can basically give it a name, you give it all the tasks that you wanna add into it, that people are organizing.

And then they organize them in whatever way makes sense to them, it only takes a few minutes to set up, it's super easy to do. And then you'll get a lot of results back. And you can basically review where people have assigned cards what they work, what they grouped them in, where have they placed them, and that very broadly.

And this is a massive oversimplification can be used as the basis for your top-level navigation. Again, go check out the other course to get into this in a lot more detail. So, and actually increasingly I'm using Chat GPT to help me do look at these results and organize them.

Because you'll find a lot of, obviously if people can organize things into any buckets that make sense to them everybody's gonna word those buckets in slightly different ways. One person might call it settings, another one might call it admin, or slightly different wording and different stuff like that.

So Chat GPT is really good at helping you find conceptually similar groups and combining those together rather than you having to go manually through and find things that are really similar. It can also total up any votes from combined groups, which means that you can still see which groups are the most popular.

And you can even give it constraints like only allow me to only have seven sections as a maximum or something like that. And yeah, you can assign different functions then automatically to different groups, all using Chat GPT. I'll talk this all through in a lot more detail in the other course.

But it's that basis of identify your top tasks, organize things around those top tasks basically makes a huge difference. Just a little bit of advice when you're creating your apps and organizing things fewer options are better. We can only hold about four things in our short-term memory, which sucks but that's humans for you.

And that's why your credit card has 16 digit numbers split into four groups of four. So as anytime you have more than four sections, you're making people think harder and people don't like thinking. If you've got to have more than four options, maybe consider grouping those different options together.

So, for example, it would be common to have your kinda main navigation, which would be a four primary sections in your app. But also have secondary navigation that has things like contact, there are settings, admin, that kind of thing in it. So you're kinda grouping things. Keep your label short whenever possible, one or two words maximum.

Be very careful to ensure that there's no ambiguity in the wordings being used. So classic examples of this is the use of jargon, where people don't know what the word is talking about, but also even real words sometimes. The example I give on the user research and testing course is I worked once on a university website, trying to attract undergraduates to attend the university.

And we did some usability testing where we asked them to go and find certain course information for, and how to apply and things like that. And as we ran the test, on several of the tests, we found that these undergraduates looking for course information ended up in the alumni section, right?

And after it happened, I think the second or third time I said to the person, why are you in alumni, right? And the answer I got back was I didn't know what the word alumni meant. So I thought maybe what I'm looking for might be in there. So actually words can be, you can send people off on completely the wrong route, so avoid ambiguity.

Also, options that are too similar to one another can be a big problem. So for example sometimes you go on a website where it has sectors and industries, well, hang on a minute. What's the difference between those two, and so you're left unsure of which to go into look for case studies or whatever it is you're after.

And also don't throw too many clicks. So there's this kind of thing that goes round that users don't like to click. And that the research doesn't really support that anymore. It certainly was a thing for a long, long time back in the the pre-broadband days where if you clicked, you had to wait ten minutes.

But these days users are pretty happy clicking their way around, and so as long as they feel that they're going in the right direction, you don't have a huge problem. So once you've got your top-level navigation in place, built around your top tasks, there'll be a whole load of other tasks that people wanna do.

And you're like, well, how do they fit into all of this as well? So what I then tend to do is take those secondary tasks that somebody might wanna do on the website and see whether they fit in with the navigation I've just made up, right? Based on the top task analysis.

So this is where I tend to do a closed card sort. And a closed card sort is very similar to an open card sort except that the sections that I created are already there. And all people have got to do is put those into the different buckets, whichever bucket they feel is most appropriate.

But notice I've added an extra bucket called I'm not sure. So that if people drop in stuff into the not sure section, they've got somewhere to put it if they don't know what to do with it, right? So this way you can now find out whether all those tiny tasks or those little secondary things fit into your navigation or not.

Again, it's really, really simple to set up. You click Create a Closed Card, so you add all the groups of your top-level navigation in. So that people can organize into that and then you just add all of your tasks in and people can just drag them around as they want to.

But then what you'll end up with is you'll be able to see where all the cards have been assigned, which groups are in and what you'll do is you'll get an agreement score where it tells you how many people agree that it should be in that section. So this card for example, at the top there have life insurance, right?

Has been put in benefits, 227 times 99% of people putting it in benefits, if you get that kind of result that's golden, that's just wonderful, right? So basically, most of the time, you're just assigning things to the top group, right? The group where it fits. But sometimes you'll get low agreement scores, right?

So this one's got an agreement score of 40%. And what's particularly interesting about this one is 92 times it's been put in working at UF and 87 times it's been put into contacts. So basically, both of those sections are pretty much equal, so they could be an either of those, right?

So now what do you do? You're kinda stuck as to where to put it. Well, it's not that big a deal, you just cross a link from one to the other, issue solved. But also you'll get some stuff like this one that's been put into not sure, right?

That's a bit worrying. Most people have put it into not sure, that's worrying, right? No, not really. Because if you look at it, 94 people have put it into not sure, 88 have put it into working at UF, which means that if there was no not sure category, which there isn't on a live website, right?

Most people would have settled for go looking in working in UF. So you could probably in that case just put it into working in UF and you'll be fine, yeah.
>> Is that percentage just for the first line?
>> Yeah, right, so it's saying 41% have put it in, I'm not sure, out of all the votes.

>> It's curious that they don't now the percentages for the other ones.
>> It's mainly the top one you care about. I think it's a great example of you probably, if you put percentages for all of them it would have just been a bit overwhelming. I actually I quite liked the fact that it doesn't makes it cleaner personally but, so basically, assign cards in the not sure to whatever the next highest winner is in most cases.

I mean, none of this is an exact science when you're working on information architecture. A lot of the time, you just have to make a judgment call about what to do, what feels right. But push comes to shove, my general rule of thumb is, if something has got an agreement score of over about 50% I don't panic too much.

And most not sure options are added to the second most common group. And also you can do a thing called tree testing just to double check any of this afterwards later. Sometimes if you've got an app that's particularly complicated, you might have even lower sections involved in it.

So basically my rule of thumb is if there's more than about what have I done there? Yeah, if there's more than about seven-ish tasks within any particular section, I'll probably look at dividing it into sub-sections. So that you're not overwhelming people with a big long list of tasks they're doing at any one time, and there's a couple of options for dealing with that.

You could do more card sorting if you wanted to but most of the time I just make an educated guess and I won't read this whole thing but I even get Chat GPT to help me with that. And I ask it to access an information architecture and make some suggestions for me.

You'll get the slide deck, you can pump it in yourself and see what happens. But it's pretty good Chat GPT is pretty good at organizing things that are conceptually similar and labeling those things is good with language. It's a large language model. So I don't always use it as it is but it's useful.

Learn Straight from the Experts Who Shape the Modern Web

  • In-depth Courses
  • Industry Leading Experts
  • Learning Paths
  • Live Interactive Workshops
Get Unlimited Access Now