>> Is product success a useful metric for whether or not Google cancel services?

Yes. Unsuccessful products get cancelled.

The successful services get to stay but the customers get hit with "insane price hikes":

https://www.geoawesomeness.com/developers-up-in-arms-over-go...

Except for Reader, Inbox, Cloud Print, the Nest API, Code, among many others I'm sure. It doesn't seem to matter whether they're successful or popular.

I was a big reader person, but google is victim of its own success. These products were barely above a few million users. At google scale,1% is more than 15m person, and usually 1% is not justifiable

Yes. But also no. RIP Reader.

Probably didn't help that there wasn't an easy path to monetization for Reader.

Huh? If Google didn't want to charge a fee for use (which doesn't seem to be how Google works), then they could easily have inserted their own adds into the RSS feeds. Or had an ad bar in the Reader interface.

It seems like Google of all places would have been able to figure out how to support that with ads.

I just don't think that they wanted to. It didn't fit into how they thought people would be interacting with the Internet in the future, and I can't say they were all that wrong...

Too small of a product for google scale. See my other comment in this thread.

Usually yes.