- To many inside, the company feels like it lacks direction. It feels like the leadership doesn't know where to take Google and this pervades pretty much everything;
- There is a lot of discontent at the the amount of money Google throws at executives. I mean Sundar got paid ~$200m LAST YEAR. Levandowski got $120m+. Back in the day, Patrick Pachette was parising the benefits of scrappiness, which saved $40m in travel costs that year. Incidentally, his was paid ~$40m extra that year;
- It feels like there are constant reorgs going on. This plays into the directionless point above but is also symptomatic of politics run amok and middle managers creating work for themselves and covering their asses. I mean a decision can never be judged good or bad if it lasts at most 6 months before another reorg.
- Some units were deliberately keep apart from Google. Most notably Youtube, Android and X. This exacerbates a cultural drift such that Google is really becoming autonomous and separate business units;
- Internal mobility is harder than it used to be and also less advisable for your career, as a general rule.
- It takes too long to do pretty much anything, particularly anything UI related. This is in part because UI work is looked down upon by the engineering leadership and the tech stacks are like 10 years behind the rest of the world.
- Sundar came up through Chrome and still seemed to me see everything through a Chrome colored lens. Chrome in particular had (has?) a bad rep for hazing and their retention numbers for women engineers in particularly aren't great.
- Google promotion culture seems to value individual technical ability above all else. I can see how this breeds the views of the likes of Damore. Worse, it seems like a lot of PAs tolerate someone being an ass if they're a high performer, which is an unfortunate reward loop.
- This all said, there are a lot of talented and great people and projects at Google. At the same time you could probably get rid of half the company because they don't really seem to have anything to do.
- Left to their own devices teams will create work for themselves. This is most apparent in Maps, which definitely went through a period of new work and features making the experience decidedly worse by, say, unifying data pipelines because consistency to some is an absolute good.
- The effects of Vic's disastrous reign can still be felt and Larry needs to take the ultimate blame for that.
- There seems to be little or no regard for the harm in burning user trust. This comes in many forms but includes announcing projects and then cancelling or abandoning them.
At least in my corner of Google the direction seems pretty clear. We are trying make our storage stack cheaper and more efficient so we can pass the savings to internal (other PA's) and external (Google Cloud Platform) customers.
The big thing I see happening at Google is that as Google has grown, it has become more and more like a "big company". So that means different PA's (product areas) will have different experiences, and different cultures. I wouldn't call them autonomous and separate business units, but there is a trend in that direction. And honestly, I think that is a Good thing. It means the company is scaling.
It does mean that the company is not going to have "one" direction. The last time we had something like that was one Vic pushed everything into Social, and it's clear that was a disaster. So Android is going to have its core direction, and You Tube is going to have its core direction, and Technical Infrastructure/Cloud is going to have its core direction..... and I'm OK with that.
Finally, as far as "half the company don't really have anything to do" --- that certainly isn't my experience at TI. We have way too many things we would like to do, and we're hiring. :-)
P.S. Google is not a dysfunctional big company. Eight years ago I left IBM, and so I know what dysfunctional big companies are like. I've also worked at VA Linux systems, so I also have a very good idea what dysfunctional small companies/startups are like. As far as I'm concerned, Google's is a place where I'm quite happy. Is it perfect? Of course not; but it's better than all of my other previous employment experiences.
Google has always had a single product area: grow and defend advertising revenue. Every major initiative can be cast in that light. Consider the "non profit" multi-billion dollar efforts.
Chrome? Prevents Microsoft or Mozilla from strangling them at the browser.
Android? Stops Apple from strangling them at the phone.
Fiber? Was intended to stop ISPs strangling them at the premises. A rare misfire which they probably wish they'd pushed harder.
The first consequential alternative revenue stream to appear in Google's lifetime is GCP. In Google's position I'd spend just about anything to get a double-digit marketshare. I suspect they have the best technology out of the three, but don't enjoy first-mover (Amazon) or sales channels & bundling power (Microsoft).
Disclosure: I work for Pivotal, we work closely with Google. But also with Microsoft and a lot of other partners too.
I don’t know if you were here long enough to experience the range of promotion processes, but only the lowest levels of tech promotions are based on individual contributions and ability. To reach “senior” grade, the lowest career grade at Google, requires significant demonstration of leadership. To get above “senior” based on individual contribution alone is virtually impossible. But to rise higher based on leadership alone is quite doable.
I would say technical ability is the least important of the three dimensions of Google’s promotion process for SWEs.
>Google promotion culture seems to value individual technical ability above all else.
If I was at a company that promoted people on anything else but their merit and skills, I'd leave immediately because of misalignment of values. If hard work and skills are not valued, the company itself is doomed.
When people say technical ability they are often drawing a distinction between overall ability to engineer (which includes many technical and non-technical things) and ability in some super specific area (C++ template metaprogramming wizard).
It's a forest for the trees sort of thing. In my limited experience with former googlers and microsoft people, they came from environments that were so far removed from real customers and product feedback that their only measure of success was if they could impress their peers by using the very latest features of C++ move semantics, as opposed to architecting/building a scalable system that actually solved a customer problem.
Literally I know of a startup that basically failed because the ex-microsoft guy spent most of this time building C++ $WHATEVER to get better performance on our i/o bound app and the ex-googler spent most of his time reorganizing the repo. They never got around to delivering software that they could iterate on.
[edit]
Both of those places are big companies, if you feel the need to respond with how you work at company X and your team is great at getting things done, you really don't need to bother. We get it, companies with tends of thousands of employees probably have many different subcultures.
Superficially what you say might be right but the truth is far more insidious than that.
For one your prospects are tied to how well your project does. Sounds reasonable right? But what this results in is many teams battling to "win" with politics often playing a large role in who is the victor, rewrites of existing code essentially to prove your technical prowess and make your promotion case and little to no valuation of team value rather than individual value, namely that teams are (or can be) greater than their sum of parts.
The last point is a key one and an area where the likes of Damore go wrong. The best team isn't one simply that's the sum of the people who individually had the highest "merit".
Put another way the cynical view is your shit never breaks no one notices but if it breaks in a big way and you fix it, well that's impact. Likewise adding a feature isn't nearly as much impact as rewriting the whole thing from scratch.
A better way to put this is, in general, refactoring code to make it healthier, or improving product excellence by fixing some long running bugs, does not get the same visibility as launching something.
However, to be fair, I haven't been at many companies where "cleaned up code base with large scale refactor and improved documentation" gets you noticed. It's just not as sexy as a customer facing launch.
As someone who has worked in a place with a similar attitude, I think I understand what he meant.
The stuff I saw:
1. People will not do easy yet impactful tasks. Everyone will try to solve a challenging problem, even if its effect on the bottom line is negligible. Folks who solve simple problems that bring in a lot of money are less likely to be promoted. At times, an easier solution was actively campaigned against by someone whose brilliant pet project would be impacted.
2. Highly valuing individual technical ability often ends up with devaluing team cohesion. In the place I worked, the people were very capable and smart. But the teams were not that effective. If something went wrong, people were less likely to see how existing policies and the organizational structure led to these problems, and instead focused more on which individuals were to blame. Everything - success or failure - was at the individual level.
So I do not think he meant that a company should not value individual technical ability. But if they value it more than effective team management, etc - I would not work there. Been there, done that.
This just means management sucked at recognizing real talent. Real talent includes knowing how to pick low-hanging fruit, knowing how to make the challenging problem look simple, knowing how to work with a team.
> Patrick Pachette was parising the benefits of scrappiness, which saved $40m in travel costs that year. Incidentally, his was paid ~$40m extra that year;
Where does that number come from? I see the NYT reporting that he got a "golden parachute" of that amount, but they give no source. The WSJ says he sold 55M in stock he owned, which is quite different from an exit payout.
Well you could be right. His stock comp for 2013 was $40m. That was the scrappy year.
The point is that instead of cutting the travel and entertainment budget for the whole company it would have been a lot more effective to simply show Pichette the door.
40m in fiscal 2014. This also contains the answer you previously wanted:
“Upon Patrick and Omid’s respective departures, all of their outstanding equity was cancelled and we made cash payments equal to the value of their unvested biennial equity grants, prorated for the time between the grant date and the cancellation date. The payments equaled $56.2 million to Patrick and $16.3 million to Omid.“
Vesting schedules are for peons. For executives, unvested shares are converted to cash when you quit! Amazing!
Great. The most concerning issue - the question of trusting all your data with a for-profit entity which doesn't get directly paid for the services it provides - gets a token mention as the very last point ("just in case, let me throw in a casual mention about this somewhere at the bottom"). I am fairly sure that announcing projects and then cancelling them, on a scale of 1-10 for burning user trust, will get a score of 0. In comparison, here are some higher scores for issues which are actually 'burning concerns':
8 - the kind of tracking data which was presented in the Waymo case
9 - the kind of data mining which happens when you combine the most popular email service + highly popular browser + most popular website analytics tool + most popular mobile OS
10 - the efforts to get into providing 'free' internet just in case a few bits and bytes escape into the ether, and attempts to acquire companies which may be collecting/assembling harder to reach datasets
And then the rest of the folks here wonder, "Do people inside Google actually spend any time thinking about whether they might be burning their user's trust?" Based on your response, I would say that it gets about the same level of token acknowledgment inside.
- To many inside, the company feels like it lacks direction. It feels like the leadership doesn't know where to take Google and this pervades pretty much everything;
- There is a lot of discontent at the the amount of money Google throws at executives. I mean Sundar got paid ~$200m LAST YEAR. Levandowski got $120m+. Back in the day, Patrick Pachette was parising the benefits of scrappiness, which saved $40m in travel costs that year. Incidentally, his was paid ~$40m extra that year;
- It feels like there are constant reorgs going on. This plays into the directionless point above but is also symptomatic of politics run amok and middle managers creating work for themselves and covering their asses. I mean a decision can never be judged good or bad if it lasts at most 6 months before another reorg.
- Some units were deliberately keep apart from Google. Most notably Youtube, Android and X. This exacerbates a cultural drift such that Google is really becoming autonomous and separate business units;
- Internal mobility is harder than it used to be and also less advisable for your career, as a general rule.
- It takes too long to do pretty much anything, particularly anything UI related. This is in part because UI work is looked down upon by the engineering leadership and the tech stacks are like 10 years behind the rest of the world.
- Sundar came up through Chrome and still seemed to me see everything through a Chrome colored lens. Chrome in particular had (has?) a bad rep for hazing and their retention numbers for women engineers in particularly aren't great.
- Google promotion culture seems to value individual technical ability above all else. I can see how this breeds the views of the likes of Damore. Worse, it seems like a lot of PAs tolerate someone being an ass if they're a high performer, which is an unfortunate reward loop.
- This all said, there are a lot of talented and great people and projects at Google. At the same time you could probably get rid of half the company because they don't really seem to have anything to do.
- Left to their own devices teams will create work for themselves. This is most apparent in Maps, which definitely went through a period of new work and features making the experience decidedly worse by, say, unifying data pipelines because consistency to some is an absolute good.
- The effects of Vic's disastrous reign can still be felt and Larry needs to take the ultimate blame for that.
- There seems to be little or no regard for the harm in burning user trust. This comes in many forms but includes announcing projects and then cancelling or abandoning them.