There are so many ways you can test usability and perform usability studies. I guess there is no final solution to how you should test your system: you probably need many tools and many types of studies but last week, we tested a study outside the lab environment.
A good thing about AB-testing and tools which record users while using the system is that you don’t test stuff in a lab environment. Normally, when you interview users you set up a meeting, invite them to a silent room and turn off the phones. The user can then test the system without being disturbed from the outside. But this is not how he will use the system. Then, the phones will be ringing and the deadline has been missed. This is when you see the tripple clicks.
With automated systems, you can see the more realistic usage but then you miss out on the face to face communication in the study. You cannot ask for details and sense the feelings from the participants.
So, in an effort to combine these types of studies, we performed a usability study in the environment in which our users reside. In a home, with kids running around. This will not be the sole test but will be one part of our usability studies.
But does the situation matter? Of course it does! As Dan Ariely found in this study, we make different choices and hold different values depending on the situation.