[bsvalley]

Other than the 20 percent free time, compensation and offices, the engineering culture at Google doesn’t seem to have anything special based on this well written pdf. It is in fact very similar to other large tech companies. Just in case you were still wondering what really attracts great engineers.

[hknd]

I think it attracts a lot of great engineers that they can (mostly) work on something which is used by billions of users. You can basically improve the life of 1B ppl by writing code.

Also: having a FAANG company in your CV makes it super easy join any other tech company.

[bsvalley]

Reality is that most of the engineers at Google don't work on google search nor google map. They work on "smaller" projects that don't necessarily reach billions of users. It doesn't change the fact that everything has to be engineered in order to work for a large amount of users, true, but do you really get that reach? nope unless you're in a very hot and selective team at G.

Also, your statement is valid for companies like Facebook, Uber, Apple, Amazon, etc. It is not a valid argument anymore in my opinion. But I totally get your point.

[jogjayr]

Google has 7 products with 1+ billion users.[1]

1. https://www.popsci.com/google-has-7-products-with-1-billion-...

[bsvalley]

Now let's look at the ration billion-user-products versus the total number of products. That'll give you the ration of employees having a huge impact:

https://en.wikipedia.org/wiki/List_of_Google_products

And we're not even looking at the number of active users per product versus the total amount of gmail ID's enabled for all google products by default (e.g. Google+)

[jogjayr]

> Now let's look at the ration billion-user-products versus the total number of products. That'll give you the ration of employees having a huge impact

That logic doesn't make any sense. No company allocates staff equally to all their products. More popular products (or higher revenue-generating products) are always better-staffed than less popular products.

[bsvalley]

You can't say "always". This doesn't make sense either because it's on a case by case basis. You could have a simple product used by billions of people, which doesn't require many devs, or you could have a complicated infrastructure no one knows about that gets billions of requests from 10 other products per day. This could be handled by an army of engineers.

Is your hot product in a maintenance mode? If so, just a few devs can handle it.

Then you have companies with groups like retail, legal, hardware, etc. They require tones of software engineers to build internal tools. They definitely don't reach billions of users and you see a lot of these teams. What I found out while working at some of the Faang's is that the hottest teams are usually very lean. You'd be surprised how one single rock star engineer can handle. When you start having +10k engineers in your company, only a minority of folks will end up working on the hot stuff.

[yifanl]

I'm not sure what you could do to improve engineering culture beyond making your office a little bit nicer without starting to take away from life outside the office.

[bsvalley]

You can innovate in many different areas. For example code reviews, CI/CD pipelines, sharing knowledge across the organization. It seems that Google as a pretty solid engineering culture but it doesn't really innovate much like a lot of other large tech companies. Standards are meant to be broken.

[joshuamorton]

I mean, google was the innovator for many of these practices. They originated Gerrit and no code review tool comes close to critique, although GitHub and co are catching up.

A lot of the innovation (kythe, grok) isn't mentioned here, but is necessary.