[martius]

The reason is quite simple: why spend engineering headcount on a less successful product?

> Some wouldn't have been viable, sure. Others were probably too ingrained in Google's hardware/software ecosystem to be separated out (although I wonder if nowadays everything Google runs on its cloud offering, which would make it simpler, just change the billing).

Google Cloud is built on top of Google's tech ecosystem, not the other way around.

[jacquesm]

Because things need time and alignment of incentives between creators and consumers and if you interfere in that relationship all the time things will never ever succeed. The last thing any project needs is an investor with a majority interest that fucks up your plans all the time, can take your employees away at will and can axe the project at any time because it doesn't perform according to their metrics.

That's why VCs take a minority stake in start-ups. The trouble usually begins when the founders dilute to the point that they no longer have a majority.