Post-mortem on a UI “input” screen

Giff Constable Aprizi, UI/UX

Today I am working on some UI/UX changes to our product, and thought I would quickly talk through a UI example (soon to die under my axe) from our beta. Some context: in the spirit of getting a beta up as quickly as possible, our first user experience (FUE) was pretty crude. After registration (also soon to disappear), the first thing everyone saw was this screen:


These three questions were basically what I asked our participants during our “wizard of oz” alpha. While not ideal from a user-experience perspective, they had the distinct advantage letting us learn about the interests of our potential customers in an open-ended way.

The screen was simple enough that most people made it to the next screen, but it had quite a few problems as well.

First, it required the user to type. Useful, as I said, from an MVP “learn from users” approach, but in general I recommend having the user click rather than type.

Second, it made the user stop and think. Not everyone had a quick answer to the first question, “what are you shopping for now?” Key point: try not to put the user in a place where they can get stuck. Instead, make things flow and get them to the real action as quickly and fluidly as possible.

Third, and this was the biggie, this input screen put people in a “search” mindset rather than a “browse” mindset. By typing in items, it felt like a search query. It led to the assumption that we would have a very deep and very broad database, because otherwise we would not let the user type *anything* in, right?

If they typed in “pleated grey flannel pants”, and nothing showed up because we’re not trying to solve the “find pleated grey flannel pants” problem*, the user would be disappointed. Master-of-the-Obvious tip: disappointment is bad. Key point: when you are designing your interface, make sure you think through the expectations your UX might create, because your design exists in a context of other products which have come before.

Other minor notes:
About 50% of the people saw and appreciated the suggestions on the right, because it gave them more context for the questions. The other 50% tuned out the right column because that’s typically ad space.

Lastly, know that even on a relatively simple screen, people are going to miss what seem like obvious UI elements. In this screen, while it might seem logical that the user would type something into each text box and either hit <Enter> or the “Add” button, in some cases the user never saw the Add button at all.

I don’t know if all this is interesting or merely obvious to everyone, but there you go!

* So what problem are we solving? We want to make it less of a chore to discover wonderful products/designers/boutiques in the ‘long tail’ of e-commerce. This direction was one of the critical decisions from our beta.