Making decisions with data
You’ve heard it a lot haven’t you. Data-driven decisions. Essentially, taking the guesswork out of decisions and backing all your choices on data that you have.
That’s all well and good if you’re a Google Analytics analyst (?!), but how do user experience designers gather data to make key decisions?
Prototype and test in-person
The first go-to exercise is to test a prototype, or to usability test an existing website or app.
This can be done in a lab (a room with recruited participants — costly and logistically difficult), or using a guerrilla method — by creating a session and inviting people to test the usability. Some people call this “running a kiosk”.
The critical thing to do here is observe. In my experience, for the first usability tests you do of your product, you just want to put the thing in front of the user and let them do what they want to do. Ask them to vocalise their observations and record. You’re not just testing the design at this stage — you’re testing the concept, or if you’re lean, the MVP (minimal viable product).
The great thing about this is, you don’t need expensive tools. You’re just watching someone use your app. You’re using your eyes! By just doing this, you will learn a whole lot about how easy your product is to use and understand some rather obvious usability problems.
Testing internally — just no
In the last two roles I’ve been in, the organisation has found it sufficient to usability test in-house. I mean — in the office in which you work. Usually, I’d say “any testing is better than no testing”. However, in this case, the results can be pretty misleading and biased. You’re testing with expert users, close to the product. It’s a lazy approach, and a box-ticking exercise. Organise a trip to your clients and get them to use it. They’ll appreciate the openness of your approach and will probably use the face time to give you further feedback on existing products. Winner winner chicken dinner!
For some though, finding people, getting them to sit and test something takes time, effort and realistically, some organisations just won’t be open to it. They’ll ask you “why are we showing a half-baked app to our valuable customers?”. In that case, you’ve got more work to do in educating the business on the value of user experience! I’ll write another post will go into that.
Remote testing can be a good next step. For quick and dirty testing, submit your site to peek.usertesting.com. You receive a video of someone using your site and vocalising their feedback within 24 hours. Sure, you can’t pick the demographic and sure, you’re not able to ask them specific questions, but again, this is a quick and dirty method of giving you evidence about the usability of your product — and better than testing in your office.
Usertesting.com have a larger plan that increases the testing amount from three per month, so you can conduct more customised tests. Why they don’t create a package where they send you a video each week for a low-subscription I don’t know!
Note: I’ve found that userbrain offer this exact feature — w00t!
Can it be social?
The best thing about Peek, and UserTesting.com at large, is that they have a load of willing users on standby. Not many companies can match that.
But I do wonder whether there’s a way of user testing on social media. I see people doing business on Twitter — making connections, sharing details, even pitching in DMs. How can usability tests get more social?
It’s all about incentivising people. Once they are incentivised (probably best through money!) it may be way more possible. You can share surveys on Twitter, but they rarely capture enough users — because they’re not incentivised.
When I recently started working with Wiggle, I asked my followers on Twitter if they had any feedback about the site. The feedback I received was incredibly valuable. Just asking people what they think (especially with an existing service) is a step in the right direction. Keep the channels open!
If you’re keen about the numbers, you can always dive into tools. Here’s a few you can use:
- UserTesting.com / Whatusersdo: Conduct usability tests
- CrazyEgg: Will give you a heat map of where users click — handy for understanding what CTAs are working and which aren’t
- Google Analytics: Will tell about user journeys. Maybe they’re going straight from your homepage to your Pricing page — what does this tell you?
- Optimizely: Do A-B tests to iterate a web page.
- InVision: Prototype designs from Sketch or Photoshop (and also use their new Boards feature for Style Guides).
I’ll go into tools in another post, but these should get you started
Don’t forget the human touch
But no matter how many tools you use, I would still say you should use human contact as much as possible. Somethings won’t be captured online. We’re observing the user experience remember! Just sitting next to someone and watching how they use a website or app can tell you so much — it’s hard to replicate that.
Though, when time and budget is tight, remote testing online is still better than nothing — so you’ve got plenty of ways of capturing user’s feedback.
Don’t just do once
Finally, don’t make the mistake of many companies who think UX is a do-once thing. Plan regular usability testing.
- Collect user feedback (e.g. what do you think of this feature?). Use affinity diagrams to group and understand themes, and then write hypothesises to test out.
- Conduct usability testing. Have a regular time in the month booked, gather the feedback, affinity diagram and write hypothesises. If you’re short of time and access, use remote testing.
- Use tools. Look at Google Analytics and install a couple of services onto your site to measure usability. One app to test out, if you want to use the HEART metric system, is Temper.io — it’s well worth a look!
Please send me some feedback — what more do you want to know, or is there something here that I didn’t describe well?