I guess I'm going to out myself a bit here, but I feel the need to clarify a few things. Full disclosure. I'm a people manager in TI and have been for a long time.
First off, I knew Erica a few years ago when she worked in our local office. She's a wonderful person and I hate that she's had such bad experiences. I don't know the full details, but I do know that one of the orgs she was in went through a lot of stuff early on. These are the typical startup is becoming a big company type of issues. They ended up letting people go, reorging the team, getting new managers, letting others go, etc... It was a major shake-up. I wonder how this impacted her. I know she transferred around the time all of this was happening (as did a lot of people in that team) and she became very outspoken. Her interpretation of what happened seems to indicate that she thought she was targeted. I don't know if this is the case, but I wonder if it was more of a team wide issue at the time than something specific with her. Also, it didn't help that we didn't have a regional HR person at the time, so whatever Erica was going through at the time was probably left unrecognized as she had no one to talk to about it. The company was growing fast and satellite offices were more of an afterthought at the time. Keep in mind the context. Google went from a couple thousand to like 60k employees in just a few years. There were growing pains all over the place and we certainly felt it in the remote offices. Especially when it came things like HR, benefits, etc...
I saw the salary spreadsheet. On the surface it seemed alarming. Especially to people who made assumptions about pay at Google. Here's some context. Again, I don't know any details about what's going on with Erica's department. This is just what I see on my end of things (I'm not in SRE, but am under the same VP).
First off, location is a big factor due to needing to be competitive in a particular region. Secondly, performance impacts both raises and the rate you come in at after promotion. People with good ratings tend to get better base pay. Yearly bonuses get a performance multiplier as well. Managers can see all of these factors (base pay, proposed adjustment for raises and promotions, multipler, etc...) We get insight into this during our salary planning process which happens twice a year for promotions and once a year for non-promo. The constants that are used for base pay are all decided based on a formula and in my experience it's completely fair and consistent. The formula literally takes a base pay for a particular job ladder, level, and region and then adds an adjustment for performance. Bonus is calculated as a flat % against pay + a multiplier for annual performance. There is no bias in this as it's all done via software. As a manager, I can do some small adjustments here and there, but it's usually at most a percentage or maybe two. It would be very difficult for me to directly influence someone's pay by say 10% or so. I don't think managers or bias in performance management is the primary cause for pay dependencies. Slight, yes, but not huge differences.
So how do people end up with such huge pay differences? Well, a few ways.
1) Google's job ladders and pay scales were kind of screwed up for a while. In one case I remember, people were adjusted by over 10% (up) in a given year, because the company realized they had the market rate set incorrectly. Google generally doesn't adjust salaries down, so if you got lucky early on and came in at a high salary, it sticks until something changes it (e.g. promo). There were a lot of people who got "lucky" as job ladders and market rates were refined over time. Keep in mind that Google went through all this during the great recession. There are a lot of factors here. In general, pay varied a lot early on. There are a lot of people in SRE who still have these inconsistencies reflected in their base pay.
2) Hiring negotiations. We all know this is an issue. Some people are very good at this and can get a huge difference of pay coming in. I won't go into details here as I feel it's probably a bit confidential, but I do see big differences in pay due to the negotiation skills of the person getting hired. Note that Google has tried to fix this lately by not negotiating base pay as much, but it still happens and was a bigger issue historically. In general, the higher the level, the more this is an issue.
3) Ladder/job transfers. Again, Google doesn't like to adjust people's pay down, so if someone transfers from one job type to another, they may come in making significantly more than what is typical for the new job. There are some rules in place to prevent this, but they aren't stringent enough to completely eliminate it as extremely tight rules would make transfers nearly impossible. Note that Google encourages transfers across ladders and teams.
4) Tenure (raises). If someone stays in a position for a long time and consistently performs well, but doesn't get promoted, they may see their pay go up more (within the range for their ladder) as they get raises year over year. That said, I don't see this very much as usually someone who performs consistently well will get promoted, but there are some cases like when an individual refuses to put themselves up for promotion or they are at the top of their job ladder and can't get promoted.
And of course, there's a lot of other things that could be possible for people who've been with the company more than 5 or 6 years. In the early post IPO days, there was a lot of chaos. There were jobs without ladders, weird things with contractor conversions, very little consistently across eng teams, etc... It was a bit wild west like. HR was a bit of a mess back then too. Again, since pay doesn't usually get adjusted down, you'll see these inconsistencies and on the surface, they'll look odd. Typically, if you dig a little deeper though, there's a good reason.
Alright, so back to the spreadsheet. Ah... the infamous spreadsheet. This just seems like it has bad news written all over it. Here's how this thing came to be. An email went out and basically said "hey, put your pay into this sheet." Now, please note that the person who sent it had been pretty vocal previously about bias, unfairness, etc. I'm pretty sure there might be some selection bias in this methodology... Also, Google spreadsheets tracks revision history, so nothing was anonymous. When I looked it over, I didn't really see anything alarming to be honest. There certainly inconsistencies, but they seemed to be linked to the issues I mentioned above. The biggest factor seemed to be location. Note that this spreadsheet was shared PUBLICALLY for a while. The whole way this was done wasn't that great. I don't take issue with the concept, but I personally found the way this was conducted a bit unprofessional. An anonymized survey would have been much better.
Finally, let's talk about bonuses (outside of annual bonus). There are ways for people to give bonuses to others at Google for deeds they deem to be outside of normal work duties. Typically these are things like assisting with something that's not part of one's job ladder, helping with volunteer projects, or having significant impact on a large project. The amount of review these bonuses receive depends on the dollar amount. Peer bonuses are a relatively small fixed amount. Anyone can award them to anyone else and they only require manager approval. By design there is little guidance and oversight. How these get awarded is EXTREMELY inconsistent, but so far the company seems to be somewhat ok with this given that they are a very small amount. My policy (and the general guidance) is that they should be for small efforts that are outside the scope of the individual's job ladder. These are like 1-2 day things here and there. Not large project contributions and not for things that someone is consistently doing. If the individual is performing something consistently, then we are expected to reflect that in the persons performance reviews so that it gets factored into their yearly bonus and annual pay increase. Hopefully the reason for this makes sense (hint: it's better for the individual).
Finally, I also don't think I would have approved peer bonuses for what was done with this spreadsheet. Not because of the fact that these individuals collected people's pay, but because their methods were very poor. I'm pretty sure all it did was anger or confuse people. Why not collect the data anonymously, rope in an analytics team, do an analysis, control for bias etc.. and then present at a CFR or even all hands? I find it hard to believe that anyone would push back on this at Google.
Sounds like a "typical" 10k+ employees enterprise. What puzzles me, I've read in the past (like 5, or ?10? years ago) stories what someone got hired from the street, got promoted, got 1M$ options, got 5M$ base withing 3 years and so on. Again all the story above sounds all that is over and google is just another mature big enterprise with typical "games", am I correct?
Thank you for taking the time to write this contribution. For someone who is in the process of trying to join Google, it's extremely illuminating. One thing I've been wondering is, since the org structure is way flatter than most big orgs, is it at all possible for a candidate to tell what level (or hell, even the ladder) a given position is listed at?
No. Google doesn't really work like that. I as a manager can interview candidates and screen for a certain level, but the req does not assume a level. Level gets decided by committee when the interview packet (feedback, resume, etc..) is reviewed at the end of the process. I can recommend a level, but they don't have to accept it. I've had candidates get bumped up and down. If I want a specific level and the candidate doesn't make it, I won't extend an offer.
First off, I knew Erica a few years ago when she worked in our local office. She's a wonderful person and I hate that she's had such bad experiences. I don't know the full details, but I do know that one of the orgs she was in went through a lot of stuff early on. These are the typical startup is becoming a big company type of issues. They ended up letting people go, reorging the team, getting new managers, letting others go, etc... It was a major shake-up. I wonder how this impacted her. I know she transferred around the time all of this was happening (as did a lot of people in that team) and she became very outspoken. Her interpretation of what happened seems to indicate that she thought she was targeted. I don't know if this is the case, but I wonder if it was more of a team wide issue at the time than something specific with her. Also, it didn't help that we didn't have a regional HR person at the time, so whatever Erica was going through at the time was probably left unrecognized as she had no one to talk to about it. The company was growing fast and satellite offices were more of an afterthought at the time. Keep in mind the context. Google went from a couple thousand to like 60k employees in just a few years. There were growing pains all over the place and we certainly felt it in the remote offices. Especially when it came things like HR, benefits, etc...
I saw the salary spreadsheet. On the surface it seemed alarming. Especially to people who made assumptions about pay at Google. Here's some context. Again, I don't know any details about what's going on with Erica's department. This is just what I see on my end of things (I'm not in SRE, but am under the same VP).
First off, location is a big factor due to needing to be competitive in a particular region. Secondly, performance impacts both raises and the rate you come in at after promotion. People with good ratings tend to get better base pay. Yearly bonuses get a performance multiplier as well. Managers can see all of these factors (base pay, proposed adjustment for raises and promotions, multipler, etc...) We get insight into this during our salary planning process which happens twice a year for promotions and once a year for non-promo. The constants that are used for base pay are all decided based on a formula and in my experience it's completely fair and consistent. The formula literally takes a base pay for a particular job ladder, level, and region and then adds an adjustment for performance. Bonus is calculated as a flat % against pay + a multiplier for annual performance. There is no bias in this as it's all done via software. As a manager, I can do some small adjustments here and there, but it's usually at most a percentage or maybe two. It would be very difficult for me to directly influence someone's pay by say 10% or so. I don't think managers or bias in performance management is the primary cause for pay dependencies. Slight, yes, but not huge differences.
So how do people end up with such huge pay differences? Well, a few ways.
1) Google's job ladders and pay scales were kind of screwed up for a while. In one case I remember, people were adjusted by over 10% (up) in a given year, because the company realized they had the market rate set incorrectly. Google generally doesn't adjust salaries down, so if you got lucky early on and came in at a high salary, it sticks until something changes it (e.g. promo). There were a lot of people who got "lucky" as job ladders and market rates were refined over time. Keep in mind that Google went through all this during the great recession. There are a lot of factors here. In general, pay varied a lot early on. There are a lot of people in SRE who still have these inconsistencies reflected in their base pay.
2) Hiring negotiations. We all know this is an issue. Some people are very good at this and can get a huge difference of pay coming in. I won't go into details here as I feel it's probably a bit confidential, but I do see big differences in pay due to the negotiation skills of the person getting hired. Note that Google has tried to fix this lately by not negotiating base pay as much, but it still happens and was a bigger issue historically. In general, the higher the level, the more this is an issue.
3) Ladder/job transfers. Again, Google doesn't like to adjust people's pay down, so if someone transfers from one job type to another, they may come in making significantly more than what is typical for the new job. There are some rules in place to prevent this, but they aren't stringent enough to completely eliminate it as extremely tight rules would make transfers nearly impossible. Note that Google encourages transfers across ladders and teams.
4) Tenure (raises). If someone stays in a position for a long time and consistently performs well, but doesn't get promoted, they may see their pay go up more (within the range for their ladder) as they get raises year over year. That said, I don't see this very much as usually someone who performs consistently well will get promoted, but there are some cases like when an individual refuses to put themselves up for promotion or they are at the top of their job ladder and can't get promoted.
And of course, there's a lot of other things that could be possible for people who've been with the company more than 5 or 6 years. In the early post IPO days, there was a lot of chaos. There were jobs without ladders, weird things with contractor conversions, very little consistently across eng teams, etc... It was a bit wild west like. HR was a bit of a mess back then too. Again, since pay doesn't usually get adjusted down, you'll see these inconsistencies and on the surface, they'll look odd. Typically, if you dig a little deeper though, there's a good reason.
Alright, so back to the spreadsheet. Ah... the infamous spreadsheet. This just seems like it has bad news written all over it. Here's how this thing came to be. An email went out and basically said "hey, put your pay into this sheet." Now, please note that the person who sent it had been pretty vocal previously about bias, unfairness, etc. I'm pretty sure there might be some selection bias in this methodology... Also, Google spreadsheets tracks revision history, so nothing was anonymous. When I looked it over, I didn't really see anything alarming to be honest. There certainly inconsistencies, but they seemed to be linked to the issues I mentioned above. The biggest factor seemed to be location. Note that this spreadsheet was shared PUBLICALLY for a while. The whole way this was done wasn't that great. I don't take issue with the concept, but I personally found the way this was conducted a bit unprofessional. An anonymized survey would have been much better.
Finally, let's talk about bonuses (outside of annual bonus). There are ways for people to give bonuses to others at Google for deeds they deem to be outside of normal work duties. Typically these are things like assisting with something that's not part of one's job ladder, helping with volunteer projects, or having significant impact on a large project. The amount of review these bonuses receive depends on the dollar amount. Peer bonuses are a relatively small fixed amount. Anyone can award them to anyone else and they only require manager approval. By design there is little guidance and oversight. How these get awarded is EXTREMELY inconsistent, but so far the company seems to be somewhat ok with this given that they are a very small amount. My policy (and the general guidance) is that they should be for small efforts that are outside the scope of the individual's job ladder. These are like 1-2 day things here and there. Not large project contributions and not for things that someone is consistently doing. If the individual is performing something consistently, then we are expected to reflect that in the persons performance reviews so that it gets factored into their yearly bonus and annual pay increase. Hopefully the reason for this makes sense (hint: it's better for the individual).
Finally, I also don't think I would have approved peer bonuses for what was done with this spreadsheet. Not because of the fact that these individuals collected people's pay, but because their methods were very poor. I'm pretty sure all it did was anger or confuse people. Why not collect the data anonymously, rope in an analytics team, do an analysis, control for bias etc.. and then present at a CFR or even all hands? I find it hard to believe that anyone would push back on this at Google.