In a few of my posts, I reference something termed â€œusability studiesâ€ as information that we considered when making a decision. Jensen Harris has recently put a post on usability studies up on his Office UI blog on this subject, so you may want to take a look.
We do a lot of studies on the Excel team too. One recent example was something we called a â€œPivotTable Longitudinal Studyâ€. This basically means that we found around a dozen Excel â€œpower usersâ€ internally at Microsoft (non-development-team folks that work in our finance, operations, and other business groups and who spend time every day working on PivotTables in Excel), and we gave them a pre-beta version of the product and asked them to do their work in Excel 12 for an entire week. During the week, they provided us with spontaneous feedback via email. We also met with them several times during the week, had them fill out surveys on specific areas, and we went and watched some of them perform their work during the day. The following week, we conducted some focus groups to prioritize their feedback on Excel 12. These sort of studies gives us a great chance to see how actual users will react to the changes we have made to features and make adjustments as necessary before we ship beta and then final versions of our software.