[netcan]

Another way of looking at this, perhaps, is that business concerns are deterministic. Deterministic is too strong a word, of course... but that's the idea.

15 years ago, the milkiness of potential cash cows was unknown. Search had an ad model that worked really well, as long as you had majority market share. Meanwhile, google had some great wed development chops. They could make a move on email, rss, a web based office suite and sweep the whole market in a short span. It was cheap enough, and each one might turn out to be a major business. Being free made sense. Higher product success rates and who cares about $10m here and there. The goal was that some of these would become multi billion dollar businesses.

I'm sure people remember the "google wipeouts" of startup territories when they released the latest something for free.

Fast forward... the bar for financial success is 25X bigger. Google's main problem is putting capital to work. We're way less naive about financial potential. Most of Google's important competitive dynamics are within oligopolies. There aren't hundreds of startups that threaten android or even youtube at any given time. A lot of the rollovers from Google pre 2010 just don't make a ton of sense.

You end up with these business/product units arbitrarily attached to stuff in the business. It's not obvious what they're for commercially, which makes them hard to manage. A commercially self sustaining unit can just operate like a business. It's easier.

It's the organisational version of technical debt.