I've been a Google watcher for as long as there has been a Google. 

My view has been almost entirely from the outside. I always have a model of what it's like on the inside, but there are obvious contradictions between my belief and reality. In the early days, I thought Google really grokked the web. Contradicting that is their almost complete obliviousness to blogging (or so it seemed). That users were turning into creators at an every-increasing pace didn't seem to penetrate.  

These days, like a lot of others, I see Google as a Microsoft that's rooted in servers instead of clients. Like Microsoft, they tend to over-reach in products, seem unable to start small and bootstrap their way to bigness in any product category. Unlike Microsoft, Google often ships their mistakes, where Microsoft killed them before they reached the market (thinking of Blackbird, Cairo, Hailstorm). Maybe that's because in Microsoft's day shipping had real physical costs. You had to fill a distribution pipe, train retailers and support people. With Google they just invite Scoble in for a demo (figuratively) and the rest is taken care of by the press and bloggers. They just have to keep the servers running (no small feat, of course). 

But I still never understood the process that led to failures like Buzz and Wave. That is, until I read this post by Douwe Osinga, who recently departed Google. He seems to have kept his perspective outside as well as inside during his tenure.  

Highly recommended reading, esp Google thinks Big and Google has a Way.