8 ms·

I tried Joel's salary scale method, and the results were... mixed.

107 points by blasdel 18y ago

antipax 18y ago
Another example of the dangers of following Joel's advice on software managment. If you think your company is like his... it's probably not. I'm sure it works amazingly at his company, but his is almost always the exception to the rule.
- df07 18y ago
  While this may be true, I didn't really get that from the article. It sounded like the system was largely a success, except that they set the maximum level at their highest-ranked people and stopped. Those people got frustrated because they had no way to move upwards. At Fog Creek, there are several levels that no developers have achieved yet. I'm fairly certain that if someone did achieve those levels they'd rearrange the system to create a few more levels, so that everyone has something to reach for and some idea of how to continue to improve themselves.
  - JacobAldridge 18y ago
    I combine that with the quote "I'm a company starter, not really an empire builder". The author couldn't see the bigger opportunity for the business, and the staff starting banging their heads on his limited vision. If he'd had a much bigger vision, he probably could have (as you indicate Fog Creek has done) included future levels for growth.
- pwim 18y ago
  This comment reminds me of http://news.ycombinator.com/item?id=538466 http://news.ycombinator.com/item?id=538466
- jedc 18y ago
  Organizations are complex systems. Cherry-picking parts of one system (whether it's from Joel, 37signals, or anyone else) is essentially useless outside of the original context. It's why culture is so important to get right from the start; once "baked in" it's very difficult to change.
jballanc 18y ago
Great essay that points out, I think, the strength of Joel's system. The author cites the primary weakness of the system as being that it doesn't translate to non-programmers. He also (correctly, as far as I know) points out that Joel has the advantage of only really having to worry about programmers. But then, I think that's the whole point of Joel's system. Sales people are paid on commission because it's a direct reflection on their skills. Likewise, I think that assistants and tech support staff probably should be allowed to negotiate since negotiation is a key skill to their position. Where Joel's system works is that a programmer's primary skill set does not involve selling and does not require negotiation/interpersonal skills (at least, not as a primary concern). I think the only problem the author really had (aside from the apparent disinterest in seeing the system through) was one of messaging.
Ennis 18y ago
I used to work at Avery's startup NITI! I've been wondering whatever happened to the team since they sold to IBM. Thanks so much for this. What a delightful coincidence!
barredo 18y ago
So, instead of employees crying about their salaries they are crying about their level. Humans are this way. We are a jealous and want-it-all-even-if-we-dont-deserve. It's not a problem of salaries or level, it's a human problem. Said that, I say that Joel's method seems a better to me.
- shiro 18y ago
  Did you read the entire article? The author says jealousy wasn't really a problem after all. The bigger problem were to keep motivation of the highest ranked people, and the mismatch between this system and the rest of the company. I found this very interesting case study.
  - barredo 18y ago
    I did read the article. Just because the author says 'jealousy' wasn't the main problem I will not think otherwise. My opinion is that it was. Only masked.
    - swombat 18y ago
      Umm... that doesn't even begin to make sense... Are you sure you read the same article?
    - kragen 18y ago
      Is there some inconsistency in the article that leads you to believe that it is inaccurate, do you have some other source for information about what happened at NITI, or are you just deliberately ignoring any information that tends to contradict what you already happen to believe?
  - chris11 18y ago
    I wonder if the ranking had anything to do with the lack of motivation. Joel system had managers sitting down with people and talking about their work and what level they deserved. So the employees basically knew what they were doing and what they needed to do to improve. Whereas in this case the ranking was just related to the respect of your peers. Also, once you become extremely competent in something your skill becomes mostly unconscious. So basically these top employees were told that for reasons they did not fully understand, they deserved to be pay more than everyone else. So while the salary was simple, it probably came across as extremely subjective to those people.
Tichy 18y ago
Joel also writes about measuring distorting results in some of his articles. Just thinking: if a scale is public, wouldn't people scramble to just match the criteria of the scale? For example, supposed somebody is considered "expert" if he has written a book about the subject. So wouldn't people try to somehow write books about things, which would take away a lot of their time and energy that they might otherwise have invested in their job directly?
- shiro 18y ago
  It may be a double-edged sword. You can take advantage of that human nature. Encouraging employees to write books on the relevant fields may benefit the company---it can be a nice PR of the company's technology level, it can attract good people to join the company, and if the company's competitive field is a niche, more books may expand the customer base.
mhp 18y ago
I'd like to know a little more about the end... he says they switched back to private salaries and then a bit later half of the team left and half got merged into another company. Uh, that sounds like a bad result (but may have been totally orthogonal to the salary change). And he's right. This works really well for programmers who like things to be algorithmic. It doesn't necessarily make much sense to the non programmers. That doesn't mean it's wrong, just "different" for them. We do audition style interviews... meaning when you come to interview for a programmer position you write code all day. If you come to interview for a sales position, you will be selling all day during your interviews. If you come to interview as an office manager, you will be doing those tasks during your interview. We don't sit and have long conversations about where you see yourself in five years. I can tell you that we've never had a single programmer that thought this interview style was strange. We had two office manager interviewees walk out during the middle of the interview. That style of interview was just strange to them and freaked them out.
herdrick 18y ago
"Your problem is finding ways to help your best people stay happy and achieve levels you haven't invented yet. That problem is the one that defeated me, but I'd rather have that one than any other." Wow - I couldn't disagree more. That's the absolute worst problem you can have. Anything that encourages your best programmers to leave is poison.
- gjm11 18y ago
  For sure, anything that encourages your best programmers to leave is Really Bad. My interpretation of the quoted sentences isn't "Having my best people leave is the best problem I could have" but "Having people good enough that I can't make a meaningful performance scale that they aren't at the top of is the best problem I could have". Which isn't obviously wrong, though presumably it's not because of implementing Spolsky's scheme that he had such good people.
gambling8nt 18y ago
The problem presented--how do the people at the top advance--has a simple solution. Have them select people to specifically mentor (via a mutual selection process) and pay them bonuses based on how much those people improve. You motivate the best and supply them with concrete goals by asking them to try to teach--which ultimately sets them up to be better managers down the road.
- Oxryly 18y ago
  While its a good idea to incorporate mentoring, tying your salary to someone else's performance does not sound like a good idea. I think the key is to leave several levels at the top empty, open to advancement.

107 points by blasdel 18y ago

Another example of the dangers of following Joel's advice on software managment. If you think your company is like his... it's probably not. I'm sure it works amazingly at his company, but his is almost always the exception to the rule.

df07 18y ago
While this may be true, I didn't really get that from the article. It sounded like the system was largely a success, except that they set the maximum level at their highest-ranked people and stopped. Those people got frustrated because they had no way to move upwards. At Fog Creek, there are several levels that no developers have achieved yet. I'm fairly certain that if someone did achieve those levels they'd rearrange the system to create a few more levels, so that everyone has something to reach for and some idea of how to continue to improve themselves.
- JacobAldridge 18y ago
  I combine that with the quote "I'm a company starter, not really an empire builder". The author couldn't see the bigger opportunity for the business, and the staff starting banging their heads on his limited vision. If he'd had a much bigger vision, he probably could have (as you indicate Fog Creek has done) included future levels for growth.
pwim 18y ago
This comment reminds me of http://news.ycombinator.com/item?id=538466 http://news.ycombinator.com/item?id=538466
jedc 18y ago
Organizations are complex systems. Cherry-picking parts of one system (whether it's from Joel, 37signals, or anyone else) is essentially useless outside of the original context. It's why culture is so important to get right from the start; once "baked in" it's very difficult to change.

jballanc 18y ago

Great essay that points out, I think, the strength of Joel's system. The author cites the primary weakness of the system as being that it doesn't translate to non-programmers. He also (correctly, as far as I know) points out that Joel has the advantage of only really having to worry about programmers. But then, I think that's the whole point of Joel's system. Sales people are paid on commission because it's a direct reflection on their skills. Likewise, I think that assistants and tech support staff probably should be allowed to negotiate since negotiation is a key skill to their position. Where Joel's system works is that a programmer's primary skill set does not involve selling and does not require negotiation/interpersonal skills (at least, not as a primary concern). I think the only problem the author really had (aside from the apparent disinterest in seeing the system through) was one of messaging.

Ennis 18y ago

I used to work at Avery's startup NITI! I've been wondering whatever happened to the team since they sold to IBM. Thanks so much for this. What a delightful coincidence!

barredo 18y ago

So, instead of employees crying about their salaries they are crying about their level. Humans are this way. We are a jealous and want-it-all-even-if-we-dont-deserve. It's not a problem of salaries or level, it's a human problem. Said that, I say that Joel's method seems a better to me.

shiro 18y ago
Did you read the entire article? The author says jealousy wasn't really a problem after all. The bigger problem were to keep motivation of the highest ranked people, and the mismatch between this system and the rest of the company. I found this very interesting case study.
- barredo 18y ago
  I did read the article. Just because the author says 'jealousy' wasn't the main problem I will not think otherwise. My opinion is that it was. Only masked.
  - swombat 18y ago
    Umm... that doesn't even begin to make sense... Are you sure you read the same article?
  - kragen 18y ago
    Is there some inconsistency in the article that leads you to believe that it is inaccurate, do you have some other source for information about what happened at NITI, or are you just deliberately ignoring any information that tends to contradict what you already happen to believe?
- chris11 18y ago
  I wonder if the ranking had anything to do with the lack of motivation. Joel system had managers sitting down with people and talking about their work and what level they deserved. So the employees basically knew what they were doing and what they needed to do to improve. Whereas in this case the ranking was just related to the respect of your peers. Also, once you become extremely competent in something your skill becomes mostly unconscious. So basically these top employees were told that for reasons they did not fully understand, they deserved to be pay more than everyone else. So while the salary was simple, it probably came across as extremely subjective to those people.

Tichy 18y ago

Joel also writes about measuring distorting results in some of his articles. Just thinking: if a scale is public, wouldn't people scramble to just match the criteria of the scale? For example, supposed somebody is considered "expert" if he has written a book about the subject. So wouldn't people try to somehow write books about things, which would take away a lot of their time and energy that they might otherwise have invested in their job directly?

shiro 18y ago
It may be a double-edged sword. You can take advantage of that human nature. Encouraging employees to write books on the relevant fields may benefit the company---it can be a nice PR of the company's technology level, it can attract good people to join the company, and if the company's competitive field is a niche, more books may expand the customer base.

mhp 18y ago

I'd like to know a little more about the end... he says they switched back to private salaries and then a bit later half of the team left and half got merged into another company. Uh, that sounds like a bad result (but may have been totally orthogonal to the salary change). And he's right. This works really well for programmers who like things to be algorithmic. It doesn't necessarily make much sense to the non programmers. That doesn't mean it's wrong, just "different" for them. We do audition style interviews... meaning when you come to interview for a programmer position you write code all day. If you come to interview for a sales position, you will be selling all day during your interviews. If you come to interview as an office manager, you will be doing those tasks during your interview. We don't sit and have long conversations about where you see yourself in five years. I can tell you that we've never had a single programmer that thought this interview style was strange. We had two office manager interviewees walk out during the middle of the interview. That style of interview was just strange to them and freaked them out.

herdrick 18y ago

"Your problem is finding ways to help your best people stay happy and achieve levels you haven't invented yet. That problem is the one that defeated me, but I'd rather have that one than any other." Wow - I couldn't disagree more. That's the absolute worst problem you can have. Anything that encourages your best programmers to leave is poison.

gjm11 18y ago
For sure, anything that encourages your best programmers to leave is Really Bad. My interpretation of the quoted sentences isn't "Having my best people leave is the best problem I could have" but "Having people good enough that I can't make a meaningful performance scale that they aren't at the top of is the best problem I could have". Which isn't obviously wrong, though presumably it's not because of implementing Spolsky's scheme that he had such good people.

gambling8nt 18y ago

The problem presented--how do the people at the top advance--has a simple solution. Have them select people to specifically mentor (via a mutual selection process) and pay them bonuses based on how much those people improve. You motivate the best and supply them with concrete goals by asking them to try to teach--which ultimately sets them up to be better managers down the road.

Oxryly 18y ago
While its a good idea to incorporate mentoring, tying your salary to someone else's performance does not sound like a good idea. I think the key is to leave several levels at the top empty, open to advancement.