Hacker News Reader

Comment by username90 | Hacker News Reader

I think many people overestimate how large the teams who are behind most of Google's products. You have a few large ones like search, ads or cloud, but many of the smaller ones are just a few dozens who over the years dwindle down to just a handful as people realize it is a dead end project. At some point the team is not longer big enough to maintain the project and it gets shut down.

View on HN

Sounds like a management/culture issue. That's not an excuse to not cycle in new staff and people move around.

Yup. Can't get that promotion if you are on the wrong product/team, that's what I heard at least.

It's still a reason not to use their products, if you know the staff cares more about the promo than the product. Why would I want to use something (or build a business on something) that would disappear soon thereafter.

Managers are just employees as well who also pick projects that will benefit their careers. Why work on a small thing with a few million users where you will never get promoted since you are already at the top, when you could just move to the next growth ship where lots of people are hired so moving up the ranks as a manager happens naturally? To me it seems like this culture seems to go all the way to the top. It is not like Sundar got promoted so quickly due to his work on an obscure product.

> Why work on a small thing with a few million users where you will never get promoted

That's the management problem right there. You should be able to get promotions for maintaining/improving a small product. The fact that you can't is what causes this endless shutdown mentality.

If the project is a dead end, and nobody knows how to change that, it sounds like a good idea to let it die.

Why support a dying product? Pivoting is a key part of startup success. Why can't large companies recognize a shift in user needs (after a decade) and decide that something is no longer valuable?

Ultimately, you really are never going to see innovation if the cost of launching a product is "staff the product literally forever, no matter if it is losing money or no longer aligned with priorities".

This seems like an argument in reverse.

Also, this makes me wonder: does every project have at least one security engineer until it gets shut down?

Well, the thing is that most products at Google were not started by upper management, so they have no real support from above. If the grassroots disappear the project dies since there is nobody planting new seeds.

> Also, this makes me wonder: does every project have at least one security engineer until it gets shut down?

Usually a specific project doesn't have dedicated security engineers but instead the engineers have to get approval from security engineers for their relevant designs.

This is probably different for teams dedicated to security like those that write fuzzers and the such.

Reply on news.ycombinator.com