Figuring out what *really* matters in UX
How Pixels, Axioms and A-B Testing Can Show Us the Way
Emily Rawitsch, Director of User Experience Design at Invaluable in Boston, enjoys talking about visual design from the inside looking out, considering the intersection of strategy plus research plus design. She spoke at Madison+ UX 2014 on Saturday, “I’m passionate about promoting the business value of design.” She went on to say that even the smallest changes can be worthwhile mentioning Google who recently tweaked their logo by moving the “l” down one pixel and one pixel to the left.
“A big theme of the conference seemed to be that the little things matter,” noted Pamela Pavliscak, founder of Change Sciences and fellow Madison+ UX 2014 speaker. “Emily’s talk was beautifully done and really resonated with me. Bottom line, everything matters and UXers are the ones who really care about that.”
“People form an impression of your website in the blink of an eye,” Rawitsch says, “50 milliseconds. A good first impression will apply to the rest of the website regardless of how it actually functions.”
Eric Dahl, one of the founders of Midwest UX, echoed Emily’s ideas in his talk on UX Axioms, a discussion he continues under the hashtag #uxaxioms. First, Eric defined UX as a holistic orientation looking at how what we make impacts people, business, and the world. “It’s all about people, it’s not about the object,” he says. “This is about not fetishizing the stuff we make. It’s a letting go of your own ego.”
Dahl ask UX professionals to focus on the experience of what they create, not on its function. To illustrate this idea he asked audience members to get out something to write on and take 20 seconds to design a vase. Then, he asked everyone to design a better way for people to enjoy flowers in their home. Someone drew a window, someone a planter, one a hanging basket. Eric wants to see this kind of broader thinking as a matter of course in the UX world.
His advice resonated with attendees, particularly Philip Likens who spoke on Friday about UX for wearable tech. Of Eric’s talk, he says, “There were several moments that I mentally bookmarked and said to myself, ‘I need to explore this topic in more depth.’”
Dahl challenged his fellow UXers to reframe the problem to open up new solutions. “UX is a synthesis of so many different fields and it is constantly changing under our feet. We aren’t experts,” he says. “We are experimenters.”
Experimenters, of course, need hard data. Corey Losenegger from Etsy stepped up to talk about how UX professionals can use A/B testing to get needed information about how small changes to a site can lead to huge behavior shifts on the part of users. “The theme of my talk,” he begins, “is about how what we think our users might like isn’t always what they like.”
Imagine that you have a theory, something you think will make your product better, Losenegger explains. You’re going to put a collection of users in a bucket for the experiment and the rest will be your control. Losenegger then highlighted several interesting scenarios in which testing went haywire. “The interesting thing about things going wrong in testing is it prevents us from making huge mistakes when it matters.”
And there are many types of failures. Misrepresentation failures are things that go wrong with the running of the test that caused errors. This category includes uneven bucketing, different page weights causing event loss, unsupported browser fails, and bot traffic fail. “Bots do weird things when you let them loose on your site,” says Losenegger. On Etsy, for example, a bot that tests that the site is available and responding searches repeatedly for owl merchandise. This leads to owls winding up in the search bar as a trending topic, so more people started searching for owls.
“Our intuition is very often wrong,” he says. “The power of A/B testing is getting a real look at how users interact with your feature before you ship it to everyone.”
Madison+ UX attendees ultimately learned that as they decide what really matters and what’s going to work best for their users, they have many valuable tools at their disposal, from the power of their own design intuition, to the wisdom of others in the field, to the concrete data delivered by testing.
Rawitsch encourages everyone involved in UX to think big picture, and to play nice with developers. After all, she reminds us, “done is better than perfect.”