"You must like dealing with people to be great at management."

This is more accurate:

"You must like talking with people to be great at management."

When I first started managing, I tried to send people a lot of email. I like writing, so email comes naturally. One thing I learned, rather quickly, is that most people don't like writing. Even well educated people don't like writing. Most people find it easier to talk than to write. Most people would rather talk an issue through than write about it. The more difficult the issue, the more that people want to talk.

The next line should have been the sub-headline:

"Imagine you spend a full day in back-to-back 1:1s talking to people. Does that sound awful or awesome?"

This bit is very true:

"Unfortunately at many places, you’re roadblocked in advancing your career unless you become a manager. This sadly incentivizes the wrong outcomes, in which people who don’t love dealing with people become bad managers, and both they and their team suffers."

I wish this was more true:

"In more technical roles, we are fortunate that many companies support separate but equal career tracks, where at a certain level of seniority you can choose to go deep into the craft of your discipline or go into people management."

I have mostly worked with startups during the last 16 years, and in all of them the top managers were paid more than the top engineers. The only "exceptions" were the founders who weren't paying themselves. But obviously, their path to renumeration is very different than it is for those of us getting paid a wage.

Of course the problem is as soon as you have more than a few direct reports you will run out of time in the day and you will start to build up management debt. I haven’t solved that problem :)

Just no. Just GOD NO. Want to be a good manager? Get the hell out of my way. You're not there for my benefit (i didn't hire you, and odds are, I probably don't think you should have a job at all, let alone one that pays more than me) so don't pretend you are. Its insulting to both of us. I'm there because you've hired me, given me money, to do some labor that you can't/won't do. If I want you, I will come to you, and it will usually be because some genius decided that I don't have the authority to fix a problem or do something. If you come to me, its because you're telling me to spend my labour on something else.

If you're going to make yourself even vaguely useful, be my human shield to keep the rest of the useless people (read: management/executives/marketers/HR) and distractions away from people actually doing work.

I appreciate this is all context sensitive and my response is filled with hyperbole, but you'll also hopefully understand that i'm particularly frustrated with managers/employers in our society wanting to be my mother/psychiatrist/friend? If I wanted to be your friend or meet with you, do you think people would have to pay me money to be in your presence?

Its not a very Hacker-esque movie, but there's a brilliant line from the English patient:

"I once traveled with a terrific guide, who was taking me to Faya. He didn't speak for nine hours. At the end of it he pointed at the horizon and said - Faya! That was a good day."

I am not sure where I suggested that a manager is supposed to be there for the employees benefit - the role of the manager is to maximize the performance of the employee and team. I also didn’t suggest their role is to mother or being a friend either (although being a friend can help).

The major job of a manager is to ensure translation of business objectives into output that is directed towards the end goal. You might think you understand exactly what your job is, but it is exceptionally common for people to not really know what the company is working towards and exactly what management wants them to do, or what is important (to the company). This is not a failing of anyone, but just a consequence of incomplete information.

Vice versa it is also very common with hands off (or too busy managers) to not know what the blocks or distractions are that are impeding progress. The whole situation becomes even more fraught when the business objectives are fluid and changing.

The only way to achieve the business objectives is to spend time making sure the information flows happen and problems are solved. I wish there was another way as it is incredibly time consuming.

Easier said than done, because of course there are bad managers out there. But even in difficult situations it can sometimes help to try "manage up" - again, the value-building and the mission is bigger than any one person.

This is exactly the phenomenon I am referring to in my comment (see sister comment on this thread level). Often no one know what management wants them to do including management themselves. The end goal is clear but how to best get there is anyones guess.

Surely it's the job of the team to work out how to get there? As the article states, managers are/should be facilitators, not necessarily the experts.

Even as another lowly peon I absolutely disagree with this. My boss/manager is absolutely there for my benefit. As much as I'm there "to do some labor that (he) can't/won't do" he's there to do work that I can't/won't do, primarily to be my "human shield", but also doing all the legwork needed to make sure that interesting jobs (and thus money) keep flowing my way. If I didn't see any benefit in working for a manager, I'd be working for myself.

For example, marketing at a previous company was extremely incompetent. Everything they did was basically a major cringe-fest for everyone involved. They would send a company-wide email announcing they had made a 'viral video' and if you could please share it as much as possible. The aforementioned video was super-boring and about corporate stuff no normal person would ever care about. How the hell is that supposed to be a 'viral video'? Apparently they completely lacked the capability for self-critcism.

They also did fake-interviews with members of upper management, the whole thing was so obviously scripted and read to the camera that you just had to feel ashamed for them. They also made rookie mistakes like looking directly into the camera and stuff like that. The resulting videos looked extremely amateurish, the average youtuber has a higher production quality.

HR did more harm than good, their main role was to get rid of everyone who had any knowledge of the product and anyone who was critical of management, basically what they wanted was yes-men.

Management itself was completely retarded, to the point where the whole C-level was fired by the board followed by several lawsuits. They kept interfering with engineering decisions not held back by any knowledge on the subject. They basically forced us use certain tools that would result in a lower quality end result, take more people and take more time, and thus be more expensive for the customer (and harder to sell). All purely for political reasons; and then they wondered why they didn't make any money.

Of course, as a software company you want to spend as much as possible on managers and as little as possible on engineers, so that is what they did. This of course attracted the kind of manager who only does it for the money and status.

In the end, the company went belly-up. Apparently you can't survive selling an inferior product for a much higher price than much smaller, leaner competitors did. Especially if your organisational structure is diamond-shaped.

I wouldn't say I'm particularly good at being a manager (or, for that matter, enjoy it particularly), but the only universal truism I've discovered is that you have to let each person decide on their own what makes a good manager and try your best to be that manager for them.

Specifically, wit respect to one on ones, I just decided to punt, errr... delegate their scheduling to my reports at whatever periodicity they think is best and just make a point of asking at the end if it's still working on the current schedule for them.

It has to be real of course - not “we have 30 minutes to fill what do you want to talk about”.

Here's where I think it cause problems: right now, if I feel like working from home on a given day, and I have no meetings, I just do it. If I was supposed to meet with my manager every single day, then now it's like I'm cancelling that meeting, so now I'm self-conscious that he thinks I'm slacking, and I would probably just not bother and drag myself in every day. Boom, reduced morale.

The best managers I've had would just show up at your desk either at the start or end of the day and ask about how it's going. The worst ones show up at my desk and ask what I'm working on.

Unless the manager was also a friend, but that usually isn't the case (nor should it be).

Now, i'm not saying that more one-on-one time isn't helpful with many employees. Nor am I saying that your experience isn't true.

But given that its in direct opposition to my own experience, i'm guessing its either a low-probability thing, that we're in different jobs/fields, different cultures, that some of your staff are either shutting up and bearing it just to get a paycheck, or the ones that disagree with you have quit/resigned/gotten another job/transferred.

And you won't. No one is ever going to be real and honest with their manager or with anyone in a position of power over them.

Even if your reports don't like all the attention, as a manager you become a better manager as you understand your reports better and they understand you better.

I don't find that difficult at all. That's what work is, you become a person who can find into the machine. I think you are overestimating your abilities.

As a manager you might know a report does not like you and is being two faced, but you won’t do anything about it unless it is affecting performance.

As a manager I would rather not know about your personal life unless it is impacting on work. One of the downsides about devoting so much time to people is they feel like they can share their most intimate personal feelings with you. The stories I have been told over the years still continue to amaze me.

He's often in meetings the majority of the day (or doing code reviews, or writing up tasks), so the onus is more on me not to bother him too much unless I really need to.

That seems to work out pretty well.

I've also been the project lead on several projects for a couple of previous companies, and except for priority, design, or time estimate meetings (usually one or two a week), I left them alone almost completely, and simply worked on development tasks unless they came to me with questions.

Some days there'd be enough questions at random intervals that I couldn't ever get into the flow of things and didn't get much coding done on my own, and some days there'd be no questions and I'd get a ton of coding done.

One time the boss brought me into his office to say how well he thought I was doing as project lead, and I was like "Really? I kind of just let them work." And he said "Maybe that's why you're good. You know how to leave them alone."

We mostly stayed on top of things and never really had a crisis moment, though, because we had a boss that would talk to the client and "manage his expectations" since he made so many ridiculous claims and demands anyway (inventing features he never discussed with us and wondering why it wasn't in the builds yet, asking for things that were pretty much impossible and thinking it could be done in a couple of weeks, changing the entire direction of the project based on an article he read the night before, etc.)

That's basically shooting the shit during coffee/pomodoro/eye-strain breaks.

When I was at Intuit, every manager had to schedule an hour every two weeks with each of his direct reports. This worked reasonably well, and most times, it would take less than 10 minutes for me and my manager to update each other and then we'd go back to work.

that sounds like 3 hours worth of manager "full attention" every week.

Half a working day, week after week, with "full attention" on me. Sounds like some kind of cult ;-)

Life isn't worth living like that and I'd run for the hills ;-)

Sure you need open communication on all levels, not a manager forcing you to talk to him 3 hours every week "with undivided attention", which is just weird.

Surely the solution to this is to introduce a middle layer of management between you and your formerly-direct reports?

Assuming a limit of 15 direct reports, managing 1000 people will require 3 layers of management - 67 front-line managers, 5 middle managers, and one Boss of All the Bosses. One person can't do it.

My experience is more than eight direct reports starts to build up management debt if you have other things to do. Once you go over this you can only make it work by doing a lot of overtime. At one point I had 14 direct reports and I did nothing other than work, sleep and eat. Of course this is not sustainable.

What do you talk about for an hour a day, every day, that is so important that it justifies allocating 16% of all development hours to communication? I don't see how this much communication could ever be a net win (or even come close to breaking even).

How would you implement it with a remote team though?

Oh, how many managers I've had that were happy to listen and nod attentively and give easy advice, but who didn't actually put any energy into my problem. Or... They wound up spending the week making it worse.

The worst manager I ever worked with ended up losing great people because talking to him was pointless at best, and dangerous at worst. You'd not believe the horrible feedback he would give people, all non-actionable, and pressing directly in the direction of his employee's insecurities. To be truthful to him was to just give him ammunition to later be used against you.

On the other hand, my favorite manager ever is not really that likable, but he is 100% trustworthy, and will do his best to solve every problem he can. Not afraid to hire, not afraid to fire. Radical candor at its best.

In a big workplace, 80% of the problems are people issues anyway a.

Other people's problems: how can I help Joe increase the amount of good code he writes? How can I help Bob get promoted? How can I help Doug and Susan work better together without constant arguing? How can I help Steve be a better presenter to execs?

The manager's duties are all defined through the work of other people. The non-manager's productivity is his own.

That is not managing - it sounds like a therapy session. The time you spend with someone one-on-one has to be real. If your report has problems that need to be solved you have to spend the time working hard on solving the problems.

Edit. I should add that your job as a manager is not to solve all problems, only those that only you can solve.

This is one of the first things I learned on my first job, and it was a somewhat bizarre realization. I both love writing, and find that it's the easiest mode of communication for me. When I had a question for someone, my first instinct was to send them an IM and start a discussion. The response was always "come to my desk so we can talk". I almost felt like they were intentionally going out of their way to pick on me. Eventually, I realized that was just the preferred mode of communication for most people.

There are definitely situations, like if you need help debugging, where it's nice to have someone standing there in person. But if we're just going to have a discussion about something, like a design issue for example, then I like writing. It also has the practical benefit that it gives you a permanent record of what everyone said for free.

I don't think it means a manager needs to be an extrovert. I had a boss who was quite introverted, but she handled 1:1's extremely well because of just being a great manager, as well as deeply respected throughout the organization. And I had an extroverted boss, to whom I was afraid to reveal any substantive information in my 1:1's.

Something inside of me makes me feel like many (MANY!) senior managers like to only pay lip service to this. Something like "sure, we'll pay the senior software engineers/architects/fellows/whatever a pretty good salary, but you can gosh darn bet that their managers are making more/way more"

Please tell me that I'm wrong here.

(As for me, I really want to give management a try. I like meetings, I like people and I like making awesome things with awesome teams. I also feel like I'm starting to hit my compensation limit as an engineer. Lastly, I want to start my own business in the next few years and I feel that the skills learned from doing management are more cross-applicable than the skills learned as an engineer.)

Though I'm in management now, I can say that our pay scale is fairly even. Top engineers make more than most managers, though our very most senior execs (CxO) still more than our top engineers, which I think is fine.

Second, this is mostly good stuff, other than the first. Some (many) have called me the best manager they've ever had. Which is nice, even though I see myself as a leader rather than a manager. That said, I dislike talking to people. I do only when I have to. What I do really well is to take the heat for my team. Anything (and I mean anything) goes wrong the blame stops with me. I also admit in front of my team and my managers when I'm wrong.

Conversely, one of my peers is a spectacularly good talker. And he lays blame like nobody's business. He also never backs down when he's demonstrably wrong. Talking to people doesn't hurt, and I guess enjoying it is a bonus, but a good manager make it does not.

IMO - talking with people makes both parties feel better but is overrated for affecting change. A better way to manage and grow talent is to put people in a position to succeed. This involves giving the right tasks to the right people, removing obstacles and distractions, understanding when and how to prioritize production(product stories) vs productivity (i.e tooling, infrastructure, and tech debt). None of this happens in 1:1s. You need to intimately understand your developer lifecycle and how your engineers are getting blocked or where they're wasting their time and how to go about fixing that. Even though they're often mediocre communicators an engineer-manager is better at making those decisions.

This pushes even middle management roles into a more sales-like personality set of people that are money-motivated to a larger degree than is perhaps healthy for the needs of the role. Upper management does need a sales angle to a degree but your usual engineering manager is not being focused appropriately if they have to "sell" the needs of their department - the prioritization for different departments are better done in committee fashion transparent across managers rather than one-on-one back room dealings that a lot of situations turn into when we talk about "managers should be enabling and removing barriers." Political in-fighting among managers is negative value to any compan - it's a civil war in effect. Some companies resemble the economic output of war-torn nations almost as a result.

In the job I just left, management couldn't remove the most important barriers at all - the biggest cost savings that every competent company could do easily would always be the last thing possible. Meanwhile, most management directives boiled down to "do more with less than is healthy... and don't burn people out!"

When you have managerial dilemmas, something will be dropped. Work just plain won't get done, deals won't be made, costs will overrun, or people will burn out and leave. Your company culture is what will determine what to sacrifice.

One of the decisions to make in choosing a management career track or not is if you want to eventually have more say over the business - what your team works on - or more say over how it's done - how the team engineers things. Someone in an architect-role can have a lot of say over the latter that upper-mid-management won't (and even a CTO type who could impose decisions on that generally shouldn't be, they're too far away from the code).

I'm wondering how this works in a self organizing team. If people are taking tasks that aren't their strong suit, what can you do as a manager? Hand select stories?

We currently do that, partially because many of our team are newer to the company, but also because we think we can choose fairly well, since our team is somewhat splintered in our products (iOS in obj-c, android in Java, the rest in js).

I'd suggest for anyone making the transition into management (particularly from a more technical background) who is after some concrete advice - start by reading 'High Output Management', and getting a subscription to Manager Tools.

In terms of the article, I'm not sure I would use the term 'senior' to apply to a manager (or individual contributor), until they are at the 10-15 year experience level (which may come in much more or less than 10-15 years employment based on circumstances).

This is what I always assumed, and why, as an introverted programmer, I avoided the management track until my late 30s. Turns out, I'm a pretty great manager (at least that's what other people say, fellow managers have been taking cues from me, and my teams have been loyal to the point where it still stuns me when they say it out loud), despite being definitely not a people person. Of course it helps that I'm an (ex-)engineer leading mostly engineers, which is already a great improvement compared to what many are used to.

I do like working for people, I like watching my people thrive and be happy and productive, and I like taking on the challenges involved in making that happen. Which does involve working with people, but that is just what it is to me: hard work. I don't hate it, but I don't love it either. However, because it's hard work it is all the more satisfying when it's successful. More satisfying than coding has become at some point.

My highest goal is to make sure that everyone is happy and autonomous in their work and the results just keep flowing. Because that's when all those messy and complicated people will leave me the hell alone...

agreed. the easiest way to get people to walk out the door is to be a jerk to them. it's amazing to me when i encounter a manager type that hasn't yet learned this very basic lesson.

Being a very nice and likable person is something you have to have to keep a team together if you can't deliver any of the things in my first sentence.

The biggest insight I think after going into management is the importance of an employee's attitude. I always had a chip cause I was kind-of a rock star, and I can see now, how that can irritate a manager.

For a dream team, I'll take a group of people who are just generally smart (can learn any new technology, can adapt) and have a great attitude (basically, people who care), over other qualities. Plus, attitudes are contagious, like a virus. Both good and bad.

For example, consider someone who grew up 'macho'. By that, I mean an environment where not looking weak is more important than being efficient, the word of the boss is king, etc, etc. It's an archetype that definitely exists in the world -- if you start with someone from that perspective, a lot of the stuff in the article is definitely unintuitive.

(Not saying your perspective is invalid, of course. There are probably other archetypes that I could've picked to try to illustrate the same effect but hopefully 'macho' gets the idea across enough.)

Managing is a chance to amplify your own effectiveness, the goal isn't to 'manage' all day, the goal is to get work done. A manager's #1 priority is to build a team that can get things done quickly and well. The kinds of people that can get work done well, in a steady state don't need an hour a day, or even an hour a week of 1-1 time. There are exceptions for totally new grads or in exceptional situations where in depth coaching is needed, but exceptions aren't the rule.

To put it as mildly as I can, if a team needs constant supervision, the manager has done a terrible job.

I try to give 30 minutes to each direct report every day. This is not to "manage" them. This is not to "socialize" with them. I do not see them outside the office, and they are not my "friends".

Each person becomes invested in the job. They become better performers. They feel that the company is important. At 11, there are no longer enough hours -- so, if the number of direct reports gets to 11, you need another management layer.

The reason for this? If this is done, this is the easiest way to get maximum performance from the team members.

In your analysis, my team members each spend 30 minutes talking with me -- but 7.5 working. And, working more effectively. I am never surprised and schedules are always met.

What do we talk about? Process improvement. What is going wrong, with suggestions to fix it. Especially introverts, who can see what is happening, but are loath to communicate.

Using this simple approach can give a "10x" multiplier to a team. The "kinds of people that can get work done well, in a steady state" also benefit. By telling me how to improve the "steady state".

FredW

If you are spending 6 hours a day chatting with people, you are probably just being manipulated anyway, since your only concept of what is going on in your work is what people are telling you.

Good luck getting your 10x multiplier.

But, say something goes wrong. The defect is then tracked back to your work. What happens? (PS. this is purely hypothetical, I just want to make it imaginable for you).

We talk, and I may bring up "Five Whys". Are you familiar with this technique? Assume you are not. We may discuss the role of process. Not for process sake, but process improvement. To make your life easier. Since you spend 1/3 of your life at work, you have an interest in this. I then suggest you google this, and that we should carry on this discussion in (say) a weeks time.

Note that this entire discussion takes 30 minutes. The follow-up will take 30 minutes a day over 4 days, but you will be doing the bulk of the talking. Because you want to. I don't get angry with you. I shield you as much as possible. But (again, this is hypothetical), this is your mistake. Everyone makes mistakes. What is interesting is that you know more about fixing this than I do. After all, I am but your manager. I want to teach you techniques to learn from what happens, listen, and propose approaches.

See? That is management. I will let you push the idea(s) to others. Actually, the modification of the idea that works for you.

Initially, I find I drive the conversation. But, as people become comfortable with me, they drive the dialog. I facilitate and eliminate problems to productivity.

Now, for the bonus answer -- rule #2 for an effective manager is to deal with issues immediately. Your desk should always be clear (as best as possible). I maintain all current open issues on a white-board, where status is immediately inspectable.

Back to "5 Whys" -- if I think that will be valuable to more people, I may create a training session -- which will be referred to in the 1 on 1s. And this is how I would manage a larger group (7 to 11 people). And, the value I can provide to being a "manager of managers"; value to the people I am indirectly managing.

Yes, I try to give 30 minutes a day for every team member.

FredW

Wait for a problem to happen, then make a big show about 'managing' the problem.

Any organization that would employ you is very broken.

As a manager, my job is to help you succeed through managing the "what" and "how" of your work. I also manage the "whys", but those are what often come out in casual interactions rather than formal meetings.

I also focus cycles on "who" and "where", but most of that is personal reflection on how things are going. Still management, but may not be visible to you, and may be me spending time in discussion with my manager & peers.

makes it sound like it's 30 minutes per day.

That really depends on the industry and the situation. It's probably true for tech work, but I managed more than double that in a blue-collar type job and didn't have much trouble.

I also had a job managing six computer operators and the combination of my inexperience and the personalities in the group caused the whole thing to blow up.

I haven't managed before, so these questions are all hypothetical. Any insight is appreciated.

I would like to think that most managers would like to pay everyone as equally as possible to show equal appreciation right? But... there's also a question of talent scarcity, then how do you balance out salary for something like that? I wouldn't want an engineer thinking their worth is less (or that they are appreciated less), because their pay is not on par with the scarce talent person. If you put that scarce talent person below market rate, how would you deal with him/her constantly wanting a higher salary (assuming he/she deserves it)? Would you just have to let that person go at some point to maintain salary levels? But you REALLY need that resource, and rejecting that person would just cause that person to eventually leave.

I view the stable/equilibrium compensation level as between 20 and 45% of the value created for the company. Want to get paid more? Create more value. I'm ideally paying for output not effort.

The obvious problem is that you can't easily put a dollar value on the contribution of engineer A vs B or engineer C vs project manager D vs product owner E vs manager F, and if you survey people for their personal contribution and add it all up, the sum will be a significant multiple of the total enterprise value created. Even with these challenges, I think it's usually clear how to find the rough buckets of "well above average contributions", "average", and "below average contributions".

What I believe is actually happening in the industry is that the highest value creators are underpaid and the lowest value creators are overpaid (often wildly).

Of course, even with this philosophy, there are company rules and strong market competition for talent, so I've lost a couple of strong performers over comp over the years. I think that even when people say they're leaving over comp, that comp is a secondary issue; it's an easy one to talk about in an exit interview, but the one good friend who left my employ "over comp" was leaving over multiple issues, comp was just the easiest one to quantify. He's still a good friend and happy at the new place.

Exactly, quantifying and justifying the value that you bring is difficult. Could you give me an example? I know this question will depend on person to person, but how aware are managers of the value you are bringing them?

You also are competing against a baseline. (Maybe a consultant could have saved us half of that, or "Yes, this year's contract for bandwidth is $50K/month cheaper, but the industry baseline also fell by more than half of that, so what was YOUR actual contribution?")

For revenue growth results, those are more often many person projects and "results are more distanced from activities", which makes it harder to quantify an individual's contribution. In cases like that, I look to the project overall results: did it ship, was it well-received by paying stakeholders, did it ship at an appropriate time, with appropriate features, did it sell, and then within the team to understand who the key contributors were. That last bit mostly comes from peer review process, as the project team worked together for far longer and more closely than I did, and the project team is the primary unit of success/failure. You won't get far arguing to me that you should be paid more because you were a crucial contributor on a failing project. (One exception is if you were the crucial contributor to argue why that project should have been killed.) Likewise, there are people who are consistently associated with successful projects, even if I can't figure out precisely how they contribute. The team likes them, reports them as helpful, and the team with that person delivers successful projects. I don't have to understand the mechanism(s); I just have to believe it wasn't by pure chance to judge that person as contributing.

No doubt that these factors play into why we don't pay more unevenly in the industry.

Now, if the question is "I'm an engineer, and I want to increase the value I bring; what should I do?" To that question, I'd answer: Find out what your company's core value proposition is, what the few most important things to the company are, and how you can contribute to one or more of them. It's great if those things are also interesting and perfectly aligned with your passions, but if they're not, you may need to choose the balance between scratching your passionate itch and scratching the company's greatest needs. How you balance that is going to influence the company's results and your own results. (Note that I carefully chose "balance" wording; it shouldn't be 100% of either of those things, IMO.)

I can give you my own perspective as someone who's managed at a very large company, which will give you an idea of some of the lunacy and arbitrariness of how salaries were determined. First, there's offers that we extended. For hiring, managers were given a headcount number, not a budget. This led to most managers trying to hire as senior talent as they could find until they were told no (i.e. you're maxed out on Principal- and Staff-level engineers). However they could always get as many Senior-level engineers as they wanted, so very few managers ever hired I's and II's. When offers were extended, a lot depended on the desired salary of the candidate and the salaries we were allowed to offer for their designation. This is where the dance with the recruiter makes a difference. People unskilled in that and/or unaware of their worth would come in low on the "let's just see if we're in the right ballpark" number and get lowballed in the offer and there was very little a manager could do about that. People who came in high got preemptively booted from the interview process. Once offers had gone out, there were a couple of ways things could go. Either the candidate was expecting a lot more and took something else, the candidate accepted, or the candidate pushed back. How they pushed back is interesting...H1-Bs would almost never negotiate on salary, but would want the faster EB-2 visa process whereas citizens would push for more money or stock. Occasionally, there was an exceptional candidate who would push for more money than was allowed in the salary band for their intended designation (i.e. SWII can make up to $130k and the candidate wanted $145k). In those cases, we'd almost always simply bump their designation to one with a higher band if we were allowed to.

However the hiring process was remarkably sane in comparison with the process for giving out raises. Raises were limited to a degree that I always felt bad about not being able to keep pace with the market rate for developers. Almost always, if there were two developers at the same level, the one with the fewer years of experience would be earning a higher base salary. The one who'd been there longer would almost certainly get more RSUs, but that compensation is intended to reward you, not make you whole, so it seemed like the base salaries should have been more consistent too. The way that raises were limited was a formula based on 3 ratings that we'd have to give each of our direct reports. We'd then have a large meeting where managers defended their ratings to calibrate across the business unit. At the end of calibrations, there were quotas our business unit could not exceed for the higher ratings, so we were often forced to lower someone's rating simply because we had too many good people. Once ratings had been hashed out and approved by upper management, each employee would be given a range of percentage increases that we could assign without needing approval. However the overall increase for our entire teams couldn't exceed a certain percentage. Think of it like a MMORPG...you've got a certain amount of gold to spend and you have to decide how to strategically deploy it in such a way as to maximize productivity overall, only you never have enough gold. It was possible to request a one-time adjustment which would have to be cleared by upper management, but those were rarely approved without a really good reason.

That's a long-winded explanation, but I think the overall takeaways are: 1) It can be a Kafkaesque process where it's easy for what people are paid and why they're being paid that to become almost entirely disconnected. Filling out those forms was somewhat like filling out your taxes. You put in some numbers and the software spit out a resulting number based on logic that was rarely comprehensible. 2) The initial setting of salary and the adjustments are almost entirely separate and it's really hard to adjust salaries after the fact. We had people quit and get rehired because it was simpler to adjust pay that way. 3) Researching your own worth is super important...a little time on GlassDoor or otherwise figuring out what you're worth can really pay off in the long run, since you'll correctly set your initial salary. 4) At the end of the day, managers are human. All things being equal, I gave higher ratings and more money to the people who created fewer problems for me and helped keep the rest of the team productive. Whether it was scheduling a return flight from a vacation on Monday morning and then coming into work that day or letting your phone die so you couldn't answer late night pages...all that stuff got remembered and the people who maintained their professionalism and made my life easier got the better raises.

I hope this helps answer your question.

So, at the end of a calibration, if you were forced to lower someone's rating, does that mean their salary would be lowered as well? If so, how does that typically end up when you break the news to that person?

Also, how do you figure out how much you are worth? You can compare with what X position is getting paid in the market, but wouldn't this just be cherry picking?

As far as how to determine how much you're worth, you can look at GlassDoor and the spreadsheets that have been circulating. There's also salary data out there...not sure how you get it, but my company gave it to managers to guide our decisions on pay. They were dumb enough not to limit it the positions that we managed, so I got to see pay ranges for my position and even look at what I'd get paid in various other parts of the country. But the best way that I know of to gauge your worth is to spend time applying for jobs. Since most recruiters will drop you if you throw out an oversized salary requirement, it's not that time consuming to start at a number you know is over and work down to the point where you start getting put in their hiring pipeline. I also advise people to actually go through the full process as much as possible, as long as you can keep your current employer from finding out. When you come into the process wanting the high end of the salary band, you need to nail the interview process, and practice really does help you avoid being nervous and gives you the ability to read your interviewer better so you can give them what they're looking for. Also, for engineers, the hiring process can actually be fun when you remove pressure from the equation. If you don't care if you pass or fail the process, you're just going to try to solve problems and learn about better engineering practices which you can incorporate into your own work. Many places will even give you a free lunch.

However, giving a zero percent raise - along with clearly explaining why - is a clear signal that performance improvement is needed.

If a salary is lowered, that would likely translate to the employee being on the brink of being fired, with large performance changes required for them to keep their job.

Almost all managers would spend more on labor than they're allowed, and also believe their employer's comp policies are at least somewhat "broken".

The problem is that most people don't ever actually prove they warrant a higher salary and in a management position you generally experience the people who aren't satisfied no matter how much you pay them which skews your perception.

That's one that I still sometimes struggle with as an engineer-turned-manager. The natural growth phase along this path is engineer to team lead, where you are probably still the most knowledgeable and where having the answers is workable and even desirable. The next step to leading a couple of related teams you can also usually get by by trying to have the answers. At some point right around there, it flips to where you know less than the people you lead and need to change your style as a result. I've seen a lot of people struggle with that (and I personally still occasionally struggle with it).

Random thoughts on this..

A good manager, who builds & runs a good team can achieve far more than any single great employee.

There are no one man companies worth $1bn.

It's appropriate that manager compensation reflects this value for shareholders. The problem is that organisations are very poor at training managers, business schools don't do it at all as far as I can see. MBA's are about money and "problem solving", other courses produce "human resource managers" which is code for pseudo lawyers and process administrators.

Strategic action to create and foster human capital outside academia is woefully lacking in our economy at the moment. Companies are increasingly ephemeral and are not invested in developing people in the way they were, there is a fight for "talent" that seems to suppose that business capabilities are innate and not learned, and disregards the benefits of team work and corporate culture. How many times have people reading watched teams and projects smashed up by new hires (or early career) from Harvard, LBS or IMD who then get promoted to do even more damage or bugger off to consultancy or some other corporate? 6 months of good numbers followed by 5 years of turmoil and painful retrenchment - if you are lucky.

Loyalty is at the heart of bad management, pension schemes locked people in meaning that they were tied to shareholders in the long run. Corporate software is the epitome of this issue - we used to chant "that fine for you but we have to live with it" at vendors, now everyone looks at the latest lashed up piece of crap and things "well, if I am here in five years then no one else will have the faintest idea that this is my fault and anyway I will probably be sacked for not filling some form properly so sod it."

The McKinsey driven cult of performance management (cf. In Search of Excellence, Build to Last and Beyond Performance) has done huge damage. This sets managers against employees in a forced pantomime of marking and red queen style denotions of performance. Management time is burned in handling a dangerous (in career terms) process which ruins their relationship with their teams, employees are asked to do things that demonstrate their performance that are orthogonal to actual performance, and not vital things like supporting team mates or their management, developing for four quarters hence (vs. 2 weeks hence) implementing infrastructure for others and so on.

Is what the manager does really more valuable? Sure, the team wouldn't work without a great manager. But it also wouldn't work without great team members. All of those people are replaceable with someone of equal qualities, and not with someone else who isn't as good in some way.

Thus I think any claim that managers should be compensated more has to rest on the idea that management is more difficult (which is presumably true in some cases and not others, depending on what the team is doing).

Managers shouldn't be seen as leaders sitting above the team, they should be seen as a member of the team with a specific role focused on business objectives and ensuring that the others on the team's work is going smoothly.

2 * 6 * x > 10 *x ?

Also 6 is a reasonable team size, but in reality I think very strong team managers tend to attract more staff and handle them efficiently up to much higher numbers (in the past I've been part of functional teams >14)

Senior management (managers of managers) can have much greater multiplicative effects. Turning a poor manager into a good one (or getting rid of the so and so), helping good managers really shine, and setting the right agendas and behaviors (don't be a jerk, talk about your stuff, fess up fast) and most of all empowering people and getting them to believe in a future can make 100's of people go from delivering 0 value to delivering scores of millions in a few years.