How Much Should You Pay Your Engineers to Ensure They Stick Around? • Hacker News Board

Re: How Much Should You Pay Your Engineers to Ensure They Stick Around?

#121

Post by nieksand » Thu, Jun 18, 2020, 4:27 PM UTC

ed312 wrote: post #69

vbtemp wrote: post #62

I just want a straight answer from someone who knows more than me. How come every time I switch jobs I get a 30-40% raise, but at any one company I never get more than a 1-3% raise each year, no matter the growth of productivity and responsibility. There's obviously some kind of prisoner's dilemma-like iterated game that reaches that Nash equilibrium, and I'm trying to figure out what it is exactly. I would have love…

I can tell you as a manager the salary bands are generally set by HR for the whole company. I was told point blank at a previous employer that we payed the 70th percentile. I asked how they knew what that was - apparently there are companies that aggregate this data (legally?) and sell it back to HR orgs for $xx,xxx+ a year. HR for some reason wouldn't deviate from this band for top talent (discrimination maybe?)

Compensation data from companies like Radford comes with three challenges

1. Your internal leveling scheme for SDEs needs to have been sorted out. External titles like "senior" are pretty meaningless, so one reasonable solution is mapping SDEs to an internal numerical tier so you can differentiate between "senior" and actual "senior".

2. Radford data comes with a trillion job titles. You need to identify the relevant ones and how your internal tiers map to them.

3. In markets where compensation is exploding (Seattle a few years ago), compensation data is a trailing indicator of what market rates you need to pay for new hires. You get fresher, but noisier, signals from candidates who reject your offers and from your personal network.

Re: How Much Should You Pay Your Engineers to Ensure They Stick Around?

#122

Post by yibg » Thu, Jun 18, 2020, 4:28 PM UTC

vbtemp wrote: post #62

I just want a straight answer from someone who knows more than me. How come every time I switch jobs I get a 30-40% raise, but at any one company I never get more than a 1-3% raise each year, no matter the growth of productivity and responsibility. There's obviously some kind of prisoner's dilemma-like iterated game that reaches that Nash equilibrium, and I'm trying to figure out what it is exactly. I would have love…

A lot of companies have budgets allocated to salary increases, so raises are capped. Why that’s the case I have no idea.

It’s also strange to me that companies aren’t willing to give a nice raise to an engineer, say from 150k to 180k but are completely fine with hiring a replacement at 180k.

Re: How Much Should You Pay Your Engineers to Ensure They Stick Around?

#123

Post by InfiniteRand » Thu, Jun 18, 2020, 4:28 PM UTC

vbtemp wrote: post #62

I just want a straight answer from someone who knows more than me. How come every time I switch jobs I get a 30-40% raise, but at any one company I never get more than a 1-3% raise each year, no matter the growth of productivity and responsibility. There's obviously some kind of prisoner's dilemma-like iterated game that reaches that Nash equilibrium, and I'm trying to figure out what it is exactly. I would have love…

My experience (I have been involved in salary/bonus/raise decisions) is usually salary is set based on the necessary amount to secure the engineer, which usually means at least their previous salary + compensation for taking the risk of a new environment with new people. I should note that I've mostly worked with small companies where these decisions are somewhat ad-hoc and non-standardized.

Raises and bonuses are based on how much can keep the engineer motivated and prevent them from looking for new work (I don't want to sound too cynical here, I think when it comes to measuring how much someone deserves in a raise/bonus is essentially how much it takes to keep the engineer providing this level of work, more or less) (based on management's assessment, and often constrained by general salary policies).

That's how I approach these discussions in broad terms, although keeping an engineer motivated and preventing them from looking for new work are not simple things to analyze

Re: How Much Should You Pay Your Engineers to Ensure They Stick Around?

#124

Post by thanatropism » Thu, Jun 18, 2020, 4:28 PM UTC

Kinrany wrote: post #43

thanatropism wrote: post #8

An interesting question is whether paying enough (imagining an unlimited budget) won't end up selecting for coders who care more about the money than the work. A further question is whether these coders are actually better for the org -- the Henry Ford efficiency wage-type incentive may keep them at consistent pace and quality, whereas rockstars with physics PhD may have parallel agendas to make the project more fun…

Consistent quality requires accurate measurement. If you can't accurately measure quality and there's only external motivation, you get Goodhart's law.

But Goodhart's law applies to anything that's measured at all (it originates from the world monetary policy), even if it can be measured vey accurately.

https://en.wikipedia.org/wiki/Goodhart%27s_law

OTOH maybe quality exists as a qualitative concept but can't be measured -- then you just do pass/fail (like colleges have dropped letter-grades during the Zoomdemic).

What I'm saying is that "shallower" coders, who are motivated by the extra pay, may be more willing to align their concept of quality with the interests of the organization, while self-motivated coders will have their idiosyncratic or fashionable ideas.

When shopping around for a team to make a web app for my startup -- no technical requirements, we were feeling the market for price first -- we got a $$$ proposal for using PHP and Laravel, and a $ proposal for using AngularJS. I regret that decision; it cost us at least a year of time-to-market.

The original team quit (they worked for a trusted friend of our "business cofounder") because they weren't really experienced in web dev and their opportunity cost was off the roof - they were crack coders that did all kinds of different projects. Then we decided to hire someone (with good references) as a consultant and manage him ourselves. Dude was apparently capable but couldn't be screamed into using off-the-shelf components rather than rewriting and refactoring constantly. He eventually flaked on us, got too many projects and we weren't fun anymore I guess. Finally we found a small company online. By all appearances the new team hates being around computers but can be managed by Trello and feature branches. Web dev is a bore, but it's a living, and not everyone manages to make a living out of their surfing passion.

Re: How Much Should You Pay Your Engineers to Ensure They Stick Around?

#125

Post by delecti » Thu, Jun 18, 2020, 4:28 PM UTC

throwaway4666 wrote: post #5

exabrial wrote: post #2

It's only about 50% about pay for STEM fields. They have to be engaged, feel like they have a voice or influence in the culture/business decisions, have great equipment to use, and flexible time off.

All of it can be dispensed with if you pay even more

It's still a balance between compensation, interesting challenges, and good work environment. You can provide less of one if you provide more of the others, or less of two if you provide much more of the other.

Re: How Much Should You Pay Your Engineers to Ensure They Stick Around?

#126

Post by flerchin » Thu, Jun 18, 2020, 4:28 PM UTC

You should pay them market rates to replace them (ie, much higher than you currently are). Yes, you should also foster their need for growth and learning, but that's because it's an investment in their productivity.

Investing in engineer's productivity, and paying them what they're worth, turns out that's all it takes.

Re: How Much Should You Pay Your Engineers to Ensure They Stick Around?

#127

Post by sgtnoodle » Thu, Jun 18, 2020, 4:28 PM UTC

mywittyname wrote: post #64

I left a job which paid in the 95+ percentile of my area for one that paid right about the median wage (maybe a little less). My reason: I like to own the software I write. We had an influx of people on our team who began to rewrite sections of stuff I wrote a long time ago for no real reason. Basically, some of them didn't like promise chains and decided to rewrite every one they encountered. Things finally bubbled…

I mainly write embedded software rather than web stuff. I will say that the hardest to read and maintain embedded code I've ever had to work with was structured to use asynchronous chains of callback functions. It can get bad enough that you just can't fix a bug without dousing it with gasoline and lighting it up. Not to say that such patterns can't be used effectively, but maybe your former co-workers had some previous emotional trauma!

Re: How Much Should You Pay Your Engineers to Ensure They Stick Around?

#128

Post by agensaequivocum » Thu, Jun 18, 2020, 4:29 PM UTC

vbtemp wrote: post #62

I just want a straight answer from someone who knows more than me. How come every time I switch jobs I get a 30-40% raise, but at any one company I never get more than a 1-3% raise each year, no matter the growth of productivity and responsibility. There's obviously some kind of prisoner's dilemma-like iterated game that reaches that Nash equilibrium, and I'm trying to figure out what it is exactly. I would have love…

> How come every time I switch jobs I get a 30-40% raise.

I see this frequently. I have a question: how many times have you done this? Don't you hit a limit at some point fairly quickly?

Re: How Much Should You Pay Your Engineers to Ensure They Stick Around?

#129

Post by scld » Thu, Jun 18, 2020, 4:29 PM UTC

vbtemp wrote: post #62

I just want a straight answer from someone who knows more than me. How come every time I switch jobs I get a 30-40% raise, but at any one company I never get more than a 1-3% raise each year, no matter the growth of productivity and responsibility. There's obviously some kind of prisoner's dilemma-like iterated game that reaches that Nash equilibrium, and I'm trying to figure out what it is exactly. I would have love…

I wouldn't be surprised if a small part of the story is just that you only hear about big bumps from colleagues when they leave.

I've gotten 20-30% raises before but I'm not going to go around and tell people that at my current job.

However, if I find a new job and get a big pay bump, I'd be more likely to mention the pay bump (especially if I'm trying to soft-poach them :-) )

Re: How Much Should You Pay Your Engineers to Ensure They Stick Around?

#130

Post by Pfhreak » Thu, Jun 18, 2020, 4:29 PM UTC

Loughla wrote: post #109

Pfhreak wrote: post #97

Earlier quoted context omitted.

Plenty of folks these days are sharing both their starting and current compensations, both internally and externally. I (personally) believe it's super healthy to do.

I agree in theory that it's super healthy to do, but the reality is that people have, what I would argue are, incorrect attitudes about who to blame. I work in a field that posts salary publicly (higher education). It is public knowledge what we all make, with databases devoted to this. When someone is hired at more than a current employee, instead of getting mad at the employer for their compensation plan that takes…

It's a learned behavior. It benefits employers to have employees squabbling with each other over pay rather than uniting and saying, "We'd like to talk with you, together, about pay". For some reason, the US in particular has a hard time with this.