Tools for Usability Testing

[00:00:00]
>> Then of course, you're gonna probably produce some form of design comp or mockup or something like that. And we've actually already test briefly, sorry, we've already touched briefly on the kinds of tests that you can do at this stage. So I talked about the five second test, where you show them the design for five seconds, then you hide it and you ask them some questions.

[00:00:23]
You might ask them which was the question I said previously of what items did you see on the page? And that gives you an idea of whether you got the visual hierarchy right. You could also make note of the order in which people have written stuff in, because the things they write first are the things that had the biggest impact on them.

[00:00:42]
So that can give you an idea of whether you've got the emphasis of the page right. Or another one often asked is, please describe in your own words what the website you've just seen offers. And that's often a very depressing test to read because you think you've been really clear and said exactly what your website is about, and then they think it's about something entirely different.

[00:01:03]
Something as fundamental as what do you do often can be poorly communicated on websites. So five second test, really great little test to run. As I said, you can do that on a tool like UsabilityHub. Then there's first click tests, which I mentioned earlier, so you ask people to click on a graphic to complete a particular task.

[00:01:25]
So where would you click to sign up for our newsletter? And are people clicking in the right place? And you'll get a heat map like this sort of show where people have clicked. And you can very quickly get an idea of whether people are getting confused by two areas that are named very similar or might have similar content in.

[00:01:45]
So this is really great where a client insists on going with, the classic one is you have a series of products, right? And each product has got its own fancy name, the Robo 3000 and the Navigator Pro and things like that. And the client is going with our product names, we must have them, right?

[00:02:07]
And so you end up with these content blocks for each product, okay? And you know nobody's gonna have a clue what any of those things are. So what you do is you run a quick test, the first click test that says, click on the, you want to find out more about a product that enables you to do XYZ, where would you click?

[00:02:29]
And of course, the product names don't describe that. And so people don't know where to click, they click in the wrong place, and you've got evidence then to say Mr. Client, Mr. Stakeholder, you are a dumbass. But you don't say it like that because you're nice. So first click test can be really useful.

[00:02:44]
And then there's what I call the semantic differential survey, which often can be called a brand keyword survey. Which is where you show the design to people and you say, on a scale of one to five, how much do you agree with the statement this design is friendly or trustworthy, etc.?

[00:03:02]
Or you show them the competition and your design together, and you go select the design above that you consider most friendly. So you can start to build up a picture then of what's friendly, what's not, what's working and what isn't from the aesthetics point of view. And all of that, as I said, you can do with a tool called usabilityhub.com.

[00:03:22]
I think it's quite a pricey tool for what it is, but they have a free option that I think is a really good free account. So you can get away with doing most of what I've talked about using their free account. And then I talked about eye tracking, didn't I?

[00:03:39]
And I talked about potentially, I like to think of eye tracking as like, or particularly eye tracking simulation software, as like a spell checker for designers. How do you know people are gonna look where you want them to? It's not 100% accurate all the time, but it will give you a pretty good idea.

[00:03:56]
So you can use Attention Insight, which is the website on the right, for that kinda simulation software, where you can just upload a design and it'll create heat map based on where it thinks people will look. I use that all the time, just using their free account, to sanity check for me.

[00:04:13]
It's also really good for showing to clients when they complain that maybe the call to action isn't visible enough or that my logo needs to be bigger or whatever. You've got evidence to prove that people are seeing that stuff. If you wanna do real eye tracking, believe it or not, that is now possible without any special equipment.

[00:04:32]
All you need is a laptop with a webcam. And there's this amazing software called RealEye that will help you with that. But maybe we're getting a bit carried away talking about eye tracking studies in this one. But hopefully I'm showing you relatively lightweight quick ways. Uploading a graphic to Attention Insight takes like 30 seconds and you get instant sense of where people are gonna look.

[00:04:58]
So after the design, you might go into building a more fancy prototype, where all your pages interlink and it's high fidelity. You might be skipping some of these stages, and that's fine if you are. I'm not saying that everybody needs to do all of this. But if you do prototyping, again, some remote facilitated testing tends to work quite well.

[00:05:19]
The reason I do it facilitated here, in other words, you sit in on the sessions and kinda help people, is with a prototype website, there's obviously bits that work and bits that don't work. So people need a bit of help when they inevitably click on something that's broken or isn't working properly.

[00:05:38]
So you need to be around for that. Well, when you're testing a live site, you can do it unfacilitated cuz everything should work. And then when you get into the actual build, you're actually building the thing, sure, I mean, you should be able to do some unfacilitated remote testing right at the end there.

[00:05:56]
Point people at a website and ask them to complete certain key tasks. Oftentimes I'll do bit more eye tracking stuff at that point because we've got a lot of changes have been made as we go along, but all fairly simple stuff that shouldn't be too time-consuming to do.

[00:06:16]
You don't need to do all of that testing I've outlined there. You can just pick and choose what you've got the time and the budget to do within the project. I would advise if you're gonna do anything, test early because that's when you get the most value from testing.

[00:06:33]
If I could only pick one stage to test, it would be to test design comps. Because not only do you get the benefit of things like first click tests that tell you if you've got your navigation right. You get the five second test which tells you whether you got your visual hierarchy right, and your brand keyword test that tells you whether you've got your aesthetics right.

[00:06:55]
So not only do you get a lot of real valuable data, but it also avoids that endless iteration and back and forth with the client. So I think you get the most bang for your buck at the design stage, but really, you can be testing throughout.