UX Lisbon: Lazy Person’s Guide to a Better World

by May 16, 2010

At the User Experience Lisbon (UXLX) conference in Portugal, Steve Krug detailed how to get more out of usability testing with less work in his presentation: The Lazy Person’s Guide to a Better World. Here's my notes on his presentation:

  • You don't have to do as much as you think to find and fix usability problems. This is not being lazy but about being effective.
  • Usability is the easiest way to make websites better. So more people should do it. But there are not enough usability professionals out there to test all the sites on the Web.
  • Usability testing works well. It uncovers lots of problems quickly. Even after just three people you can find a lot of issues to fix.
  • It takes far less resources to find problems than to fix them so your website is always going to have usability problems. Get over it. Fix the serious problems. If you do that you are doing a good job.
  • It's very easy to fix the easy problems first. So it is easy to get seduced into fixing only the simple issues you find. As a result, the most serious usability problems tend to persist over a long time.
  • Focus ruthlessly on a small number of the most important (serious) problems. It's easy not to do this. Serious problems are there not because you are not aware of them but because they are hard to solve. Or because you think you'll fix them eventually. “In the next redesign, well address this.” But your users will continue to suffer.
  • The process of focusing narrowly starts with usability testing observers. Instruct the observers to write down the three most serious issues after each session. Do a debriefing after a test. Do it over lunch and decide what to addres in the next month. Your objective is to decide which serious problems to fix.
  • Go around the room and gather what people thought were serious observed problems. Then decide which issues were the most serious. It is usually not hard to decide what the most serious issues were as people all saw the same tests.
  • For each problem, figure out what needs to be done to fix it. Once you have allocated the resources you can commit in the next month, stop. Throw the rest of the list away.
  • When fixing problems do the least you can do. Tweak, don't redesign. It is very tempting to think about redesigning. But it takes too long. Find the smallest change you can make to eliminate the problem for the most people.
  • If a tweak did not fix the issue, but you think it is the right solution, make the tweak louder. If that does not work, try a different tweak.
  • Tweaks cost less, they require less work, most people don't like change so redesigns annoy them, resigns require lots of meetings, and more. Tweaks are generally better.