"What we can learn from the intelligence community" - Lucy Spence at Lightning UX

 by Martin Belam, 9 February 2011

Last week I was at an evening of Lightning UX organised by Lee McIvor, a very welcome addition to the events calendar on the London UX scene. I've been blogging my notes from the evening, and over the last few days have covered talks by Harry Brignull, Cennydd Bowles, Boon Chew, Jonathan Kahn and Tyler Tate. This post is about Lucy Spence's talk on puzzles and mysteries.

"What we can learn from the intelligence community" - Lucy Spence

Lucy Spence from LoveFilm gave an appropriately movie-themed set of slides, which I think identified a fundamental flaw with some of the ways that we allow raw usability to impact on the broader experience of digital products. She was pointing out that there is a difference between a puzzle and a mystery.

A puzzle has a binary solution.

A mystery may contain puzzles in it, but it is a broader set of intrigue.

Lucy feels that often the usability testing we do tries to solve puzzles in a binary fashion, when actually we are dealing with UX mysteries. She worried that even with masses of data, it is still sometimes hard to tell exactly what is going on with a site and why people are behaving the way they are. She recommended using radical A/B testing to learn before committing. She argued that testing totally different solutions side-by-side allows you to admit that you don't always know what is going to happen.

At one point Lucy said that improving user experience wasn't about playing "whack-a-mole", where you make interface changes that shift the problem around the site, for example with conversion drop-off, rather than addressing the fundamentals of the pricing or shipping conditions or some other business issue that was the real cause of the failure.

What I liked about Lucy's talk was that it played to some of my own concerns with relying solely on usability testing without looking at user experience and business propositions as a whole. You can quite often observe in the lab, particularly when testing with prototypes or mock-ups, users fixate on details to do with the use of dummy content, and the findings from that session will be based on a flawed understanding of the true product offering. Or you can observe a repeated drop-off from conversion, that turns out to be nothing to do with the colour of the submit buttons, and everything to do with the pricing structure or the shipping policies.

Lucy has published her own sketchnote of the rest of the evening, and has also made her slides available:

Keep up to date on my new blog