Insider tip on Google: to be promoted you must demonstrate that your product had a sufficiently complicated problem, and you came up with a solution to it.
You need not demonstrate that the problem is unique or original, merely that the solution resulted from individual work and demonstrates technical insight.
So the company rewards reimplementation, and it's easiest to prove individual work and insight if your guys built the whole product.
Hence Hangouts died in favour of Allo and Duo, and Allo and Duo will die in favour of the next identikit products, because it could no longer attract sufficient internal talent — the relatively complicated bits were already done. So in terms of your career, it's a dead-end team.
The search engine will always be there because they can interchange the internal products that go into it without anybody knowing the difference. Some other things, like GMail and Maps, are big enough that they should survive just because killing them is almost unimaginable (but, again, expect their internals to be reimplemented needlessly for personal enrichment). The Android team, thankfully, is still mostly outside of the normal Google mismanagement.
For anything that never makes it all the way to tier one, don't get too used to it.
Biting the hand that feeds IT
