In some design Utopia, everything would be tested. An unseen army of usability specialists would verify everything and free our minds from worrying about unforeseen outcomes. Our users would be empowered, our messages would hit home, our harps would be perfectly tuned, and our gins and tonic perfectly mixed (lime, not lemon, thank you).
Until that day, we’re stuck with the real world, in which designers sometimes have to trust instinct and speculation rather than proof. We try to insure ourselves against getting things totally wrong; to wit, an arsenal of design fundamentals. Mapping. Affordance. Redundancy. Theories with impressive German names. We’ve spent time reading the books, listening to others and forming our own intuitive principles from practical observation. This helps us sleep at night. We’re professionals, dammit.
In a previous job, I ran a new graphical element on our site. The theory was watertight: better visual weight, higher legibility, stronger typography, aesthetically harmonious. A no-brainer. Yet it failed. Horribly. After causing a noticeable dip on conversion (most of e-commerce UX design is predicated on increasing sales), we quickly conceded defeat and rolled it back.
It was embarrassing, needless to say. Whilst not staking my career on this design, I’d taken the time to argue its merits, explaining to the powers that be that although some of them disliked it visually, it was the right solution for many reasons.
This story, to me, explains a significant problem with designing by metrics. You get rapid feedback on whether an approach works, but none whatsoever on why. Sure enough, when we reverted to the ‘weaker’ version, normal service was resumed. To this day, I haven’t a clue about the cause; yes, I could have run some usability tests but for a lone image it would have been pure self-placating overkill.
So, when the design gods forsake us, where do we turn? Obviously, we need rapid feedback. A poorly-received design can be measured in many ways – in this case a simple conversion metric, but in other instances it might surface as reduced clickthroughs, backchannel mutterings, failed usability tests or customer complaints. Remaining alert will depend greatly on our relationship with other parts of the business and the market itself. We also deserve the occasional reminder that user experience design is a subjective matter. Theory is valuable and useful, but the outside world has an uncanny habit of regularly throwing us off our ivory towers. Perhaps this is no bad thing.
That’s why, on reflection, I’m glad this episode happened, despite the hurt pride. Doing what I do wouldn’t be nearly as fun if things worked every time. Who wants to work in a field reducible to process, heuristics and how-tos?