Modern PM research techniques

In a comment in the previous post there was a mention of maybe the PM role will go away with the advent of new technologies that make it easier to drive data directly to the developers. In particular there was a mention of Watson -- which is our online crash data -- which has been a miraculous help to developers (an amazing innovation, created almost entirely by development at the early stages).

My opinion is that the more data there is the more there is going to be a need to step back and to really understand what is valuable and what is not. The next evolution of data along the lines of Watson has been the Office Customer Experience Improvement Program (sometimes called SQM internally) which was pioneered by the MSN team and is used by many groups at Microsoft (and has been used in Office previously).

In his BayCHI talk last week, Jensen described some of the data used by PM to help to develop the new UI. In many ways the UI work in Office "12" shows how the joint work across all the disciplines mentioned can yield a success that would not be possible with only one perspective. One thing to keep in mind is the scale of this design--no product does as many things for as many people in as many different ways as Microsoft Office, so the decision to evolve the experience and the chosen designs are a big deal to us and our customers.

Rashmi Sinha at BayCHI posted some kind words about the whole interaction we are having with customers, which I think dovetails with the comments previously.

Jensen's slides are also posted from that talk. From the slides you can see that the data helped us to better understand how the product is used by a large number of people, but the design was still a collaborative effort that followed the model I outlined. And of course we're still in beta so we have been going through a significant amount of feedback and iteration. Also as a reminder, Jensen's manager, Julie Larson-Green, has a video that shows the UI in action.

--Steven