> you can request a transfer to any team that will take you
If you're trying to transfer because your manager is giving you crap, you'll find it harder than you think to get the transfer approved. In the meantime, your manager can sandbag your quarterly performance rating and taint your compensation and chances for promotion for a long time. A bad manager at Google is still a bad manager.
> Performance reviews are peer-driven.
This is partly true. Promotions are decided by committee with input from your peers. But without strong performance ratings from your manager and appropriate projects to work on, you won't get promoted no matter what your peers say. The committees are very reluctant to override the manager.
At Google, the not-so-nice managers are the ones who use low ratings to decrease their reports' mobility. You can actually see this on a Perf graph. 3.0, 3.0, 3.0, 3.0, 3.0, 3.5, 3.5. The step discontinuity is because the number itself isn't reported to employee. The bucket is. (3.0 to 3.4 is a huge range, from the 3rd to 65th percentile of the company, but any number in this stretch is reported as "Meets Expectations".) The 3.5 jump happens because the employee gets sick of "Meets Expectations" and the half-step semi-promotion is used to perk him up a little bit. But again, "Exceeds Expectations" is a huge range (3.5 to 3.9) and the numbers are still put in the low end of the range. 3.5 will very rarely get a person promoted.
What the numbers are supposed to represent is a probability that the employee would succeed in the next higher level. 3.x is an x out of 10 chance that a level-N employee would be successful at level N + 1. 2.9 and below is cause to start the PIP/termination process, with a recommendation of timeframe (2.9 = 60 days, 2.8 = 45, 2.7 = 30); 4.1 to 5.0 represents various degrees of consideration for various (sometimes very lucrative) special awards. At 5.0, you might see an EMG award that would impress even bankers.
The ones who overuse 3.0 and 3.5 are the not-so-nice managers, not the evil ones. The evil managers are the ones who use PIPs to block transfers. Those are the 4.1-to-PIP cases where the target has to get a VP override in order to make the move.
In the meantime, your manager can sandbag your quarterly performance rating and taint your compensation and chances for promotion for a long time. A bad manager at Google is still a bad manager.
Bingo, and Google has plenty of those.
Google is a pretty great place if you have a good manager, not because the executives know what they're doing, but because you're getting paid to hang out with some brilliant, creative, curious engineers. A lot of them are really good people as well. So if you have a good manager, you can do very well at Google, but like any company, it's bad if you have a bad manager.
Promotions are decided by committee with input from your peers. But without strong performance ratings from your manager and appropriate projects to work on, you won't get promoted no matter what your peers say. The committees are very reluctant to override the manager.
Exactly. It used to be an OR-gate between manager and peer feedback. Now it's an AND-gate.
As a recovering academic, I felt a twinge of horror when I read this. When I didn't love the work I pushed on mainly out of guilt, and it was more than I could carry after a while. Maybe I should have tried resentment.
True for Real Googlers. Real Googler status used to be informally conferred at Senior SWE but now it's somewhere between Senior SWE and Staff SWE. Real Googlers can pull a Yegge (quit their project in public). Peasants get PIP'd by their managers if they are seen talking to TLs from other teams (transfer risk) in the microkitchen.
stcredzero mentions a laundry list of soul-sapping bad behaviour and in previous jobs I've suffered through (or quit over) probably all of them at one time or another.
There are some subtle yet fundamentally important differences in how Google (from my experience) approaches software development that end up addressing a lot of these.
1. Most (if not all) engineering managers are... engineers. I've seen internal job postings for engineering directors (to put that in perspective, an eng director might have 30-100 engineers under him or her typically with some engineering managers in between although some particularly senior engineers may report directly to their director) that call for "deep knowledge of C++". Some eng directors still submit CLs. This is something Google takes very seriously;
2. As an engineer, I don't have to work for you (where "you" is any particular manager or director). This is incredibly important because it puts strict limits on how much crap you can be given because (assuming you're in good standing) you can request a transfer to any team that will take you (within limits and your existing manager can delay you but can't block you);
3. Internally, Google is extremely open with what's going on in the company. There are a handful of things off-limit but in almost all cases you can view the data and code and announcements for almost every project in Google;
4. The level of meetings and overall bullshit has been, in my experience, incredibly low compared to previous employers;
5. Performance reviews are peer-driven. This system is not perfect and you'll have critics who point to thinks like you need senior people around you to get promoted. Some of these claims have merit but the system while not perfect is still (IMHO) very good; and
6. Gratitude. As Steve Yegge [1] puts it:
> You can't help but want to do your absolute best for Google; you feel like you owe it to them for taking such incredibly good care of you.
I find many of these alleviate burnout by way of mitigating resentment. I'm not sure I fully agree that burnout is solely caused by resentment however. I've known people who have been well-rewarded and respected for what they've done but they just reach a point where they need a break and have to do something else. I'd characterize this as burnout but perhaps definitions vary.
Lastly, I'll touch on a point someone else mentioned: as time goes on my tolerance for bullshit and time-wasting goes down. No I'm not going to your daily standup ("mini performance review") because it's a waste of my time. No I'm not going to work on your shitty project that's made shitty technology choices for political rather than technical reasons. I don't have the time nor inclination to indulge you in this.
[1]: http://steve-yegge.blogspot.com/2006/09/good-agile-bad-agile...