CM Harrington 18 December 2006 Fantastic article. It’s so important to know what you are testing. It’s easy to be distracted from that. With “low res” mock-ups (paper prototypes, etc), it’s easy to make hundreds of variations for each step of the process… You’re generally not testing their ability to click a form widget, but you are probably testing their ability to navigate a new site based on an established workflow. The main versions of the screens are all you should need to conduct that study. A/B testing is also a good idea, especially if you’re asking your audience to change their workflow.