I'm not on MSR directly, but I am part of a future research team at Microsoft. Internally things are interesting - you have the freedom to do pretty much whatev

I'm not on MSR directly, but I am part of a future research team at Microsoft. Internally things are interesting - you have the freedom to do pretty much whatever you want, but you have to justify it's usefulness one way or another. Microsoft is made up of thousands of tiny little teams that are mostly autonomous - each team has to prove that there's a reason for it to exist though. With research, especially things that will only come to fruition ten or twenty years out, it's very hard to directly measure it's impact. One of the things we end up doing is alternating between future and near reaching projects so that we can justify our existence with projects that are tangible in the present, but also contribute to future goals as well. In a sense we are product focused as well, except our target market isn't consumers, but rather people up on high within microsoft. We have to sell our research to them one way or another, and it's often the research teams that have the best advertising sense about them that end up doing the best.