Press for the agency

Adventures in User Experience

This article is courtesy of .net Magazine (download the original).

 

Over the past eight years, our studio (www.smashlab.com) has built many websites for many different clients. But the lessons we’ve learned building our own community have reminded us that we still have a lot to master.

 

About a year ago, we had an idea for a website that would help connect like-minded users. Based on a simple paradigm, MakeFive (www.makefive.com) enables people to create, share and vote for top-five lists. We started planning, had a working prototype within five months and in November 2007 we went live. We’d planned on an iterative process, and that’s just what we got. In the months since the launch, we’ve reworked the site and I’d like to share some of the things we’ve learned.

 

Simple and consistent
No matter what you build, you’ll probably plan-in more features than you need. We really wanted to do things right, so we’d think about issues, contingencies and extra features in great detail. Take away all but the necessary ones – you can always add more as the need presents itself.

 

Another issue we faced initially was that, with a site requiring extensive user input, many informational cues were needed. So, to keep the design from becoming a complete mess, we had to establish basic rules and try not to break them.

 

For instance, we determined heading styles from the very beginning, which sounds like a no-brainer, but it’s easier said than done. How much space between a button and the bottom of a dialogue box? How do you deal with a field within a field within a field? It gets tricky quickly. Remember that simple, consistent treatments reduce the likelihood of the development team peeing on your chair when you’re not looking.

Speak their language
A core feature of our site was a drag-and-drop function for creating lists. We spent weeks planning and building it, but it was an utter dud. I’m sure it would have worked in a desktop environment but it just didn’t translate to the browser. As such we tossed it for something much more ‘Web 1.0’. It works just fine and people immediately understand how to use it.

 

Understanding both what the user wants and why they want it is key. Designers are like parents: we love our babies no matter how buffoonish they may be. We found that our site being “kind of neat” wasn’t nearly enough to entice people to use it. But when we added a simple reward structure and authority system, everything changed and people would log in for hours. Remember, the fact that you love it doesn’t mean anyone else will.

 

Similarly, in order to make our site more usable, we found ourselves making text much more literal. For example, ‘Community Response’ was quickly changed to, ‘The Top Five’, which was faster to scan and comprehend. Along the way we also learned to make the buttons bigger.

Nothing is sacred
Good designers will sacrifice their favourite elements for the good of the project. We found that certain conventions didn’t jive with users. It’s been a process, but nothing is sacred – I’d dress the thing up in spandex now if it helped.

 

Just before launch, my wife gave the site a test run. It drove me bonkers – I kept thinking, “You’re using it wrong!” But everyone interprets navigation differently and I had to learn to accommodate that. It takes willpower, but I now sit silently and try to learn from how others engage in the site.

 

Accepting imperfection is just as important as seeking it. Problems with MakeFive are brought to our attention daily. These are documented and addressed, but it takes time. If we wanted everything perfect, it never would have gone live. If you decide to launch your own community-driven site, expect a wild ride. There can be some real lows, but nothing beats the rush of seeing thousands of users logged into your service.

 

Eric Karjaluoto is creative director at smashLAB (www.smashlab.com). His thoughts on design can be found at www.ideasonideas.com.