The heart of what seems to be going on:
The main theme that seems to be emerging: An elimination of Google’s more centralized functional structure–where Rosenberg was one of several manager kingpins–to one in which the individual business units and their engineers, such as its most independent Android division, rule more autonomously.
Certainly, a fully centralized and functional structure can have serious drawbacks in a company as large as Google has become. With a lack of business accountability, pet features can crowd out high priorities, decision making becomes slow and responsibility for line-of-business success is defuse. Having lived inside a large uber-functional company myself, it is not hard to see some downsides.
However, the other thread here is the primacy of engineers over managers. Engineer supremacy is a fashionable meme at present, not least because FaceBook (a 1000 person company) is practicing it to considerable effect. I find it appealing, too – often, you get the best results by making the people creating the product responsible for getting it right.
Yet at Google, even in the existing culture, engineers can and do get to initiate a great many project under Google’s “20 percent time” rule, and have great influence more generally, too. There are problems with the approach:
- It produces a too many initiatives, none of which may get the momentum and support they really need to fully succeed; it is weak at direction setting and prioritization
- It has a tendency to produce incremental and/or derivative ideas, which may have their place, but is weaker at pursuing something strategically or radically (see this discussion of Google Apps, for instance)
The solution is real, specific, strategic understanding at the senior levels, albeit open to the ideas bubbling up throughout the organization. Even as a company gets big, it still needs to choose its key bets with great care – there can be only a few of them – and pursue them with powerful focus. A business unit organization may help, but it can’t solve the problem if the problem is a lack of strategic direction or strategic prioritization.