Things I've learned transitioning from engineer to engineering manager
blog.pragmaticengineer.com
blog.pragmaticengineer.com
I disagree with this a ton. The best managers I've had flipped #1 and #2. When your team is performing, the company profits. Far too often I've seen the manager put the company way ahead of the team, and that leads to attrition within your team or a major lack of motivation.
I’ve had managers make it clear that me and my teams were a distant #2 to the company needs. It was pretty demotivating (as you said).
On the other hand, part of a managers job is to align their teams with the organizational priorities.
Maybe it comes down to how you achieve that alignment?
I guess the way I view it is that company is the destination and team/people is the journey.
This way isn't the only way though, I've seen teams run 'company first' it generally involved a lot of unhappy people and I didn't like it, so I avoid that style personally. A friend of mine seems to pride himself on being a 'hard driver' and being 'brought in as the hammer', ; but to each their own.
He understood we were a team, and we weren't stupid. Your employees are working in a technical, educated field. They can smell bullshit a mile away. If you try to take advantage of them they will shit your bed and leave
Which to make the primary focus can and should switch up. If you are firefighting on a submarine all that matters is the task - the team are all dead if you fail. The individual does not matter there and then.
Back in the land of computers we have less extreme but analogous problems. There are times when you do need to focus on the task, however if you're a good leader you've used the times when that isn't the case to focus on building a Team and developing each Individual so that they can, and are willing, to dig out to achieve the Task.
Your team's health can absolutely be the number one vehicle you use to get the company the results they are after, but you can't forget the fact that the company employs you, and them for reasons other than their happiness.
That said, it's a mistake to treat developers like commodities or to dismiss their concerns and desires as second to the company, in my opinion. Churn hurts, and demoralized teams have negative impact on develpment.
So, take 'professional occupations' with a grain of salt. There are high-paying occupations that don't require a 4-year college degree.
Of course. The vast majority of whom are already employed at market rates.
I can easily be replaced but the company will be eating close to $100k in costs as well as taking on more short term risk.
Lots of companies die because they pursue quarterly growth. Team-focused managers focus on long-term health of the company.
"Put the company first" is another way of getting at "Engineering managers need higher trait Agreeableness in order to succeed". It's not maximizing Agreeableness, just putting it at a higher level than individual contributors. And for good reason.
The overall goal is to make smart tradeoffs and find win-wins. You've got yourself, your managers/executives, and your team. They each can get varying amounts of acclaim, compensation, and terms of employment (ie, hours worked). I kind of want to say that the company matters here, but it doesn't really. Well, at least not until you get to a large enough company scale that your reputation among people at the company starts to matter. Say ~100 people or so.
The point I think the author makes is that it is not the desires of the team that set up priorities at this point, and that it is contrary to what you are expected to do at the dev level (i.e. advancing your team projects first)
The thing is, if a manager does not set goals for it's team, it will find some by itself (refactor that old code, clean the doc, improve that funky demo, test out the latest shiny equipment) that are not necessarily the best choice for the company.
My experience has been that in that situation, very often, the team is relieved when it receives new instructions. It is not a "team or company" thing It is just an ordering of priority goals that makes sense to everybody.
Let's start a thought experiment. We'll start a company with one person. What will that person do? Manage? Of course not. There is nothing to manage, because there is only one person. Instead, although they have to plan and prioritise their work, if they aren't actually spending most of their time working, then nothing will get accomplished.
What if we have 2 people? Should we make one person a manger and have the other person do the work? Of course not. The manager will have virtually nothing to do while the working person will be overwhelmed. Although you may shift some responsibilities depending on who likes to do what, or who has various skills, unless both people are working then you will have lots of problems.
At what point do we need a manager? Well, we need a manager when the communication overhead and day-to-day chores impacts development. We only need that manager when there is enough work that they will be able to spend almost all of their day doing that work.
So what will they do? Basically anything that is stopping the workers from getting their job done. If there is a problem with communication, then the manager needs to organise things so that everybody has the information they need. If there is a lack of prioritisation, then the manager needs to prioritise/plan the work. If there is conflict on the team, the manager must find ways to resolve the conflict. I'll stop here as there isn't much point enumerating all the things a manager must manage.
The point is that everything a manager does is a result of coordinating large numbers of people or disparate information sources. Their job is to coordinate, prioritise, reduce conflict and communicate so that the workers can concentrate on getting their work done. The manager is there to "take one for the team" so that the team doesn't get embroiled in drama, trivia, or complications.
Getting back to the original problem: "you'll need to put the company first, team second and your team members third". Sorry to be rude, but that's just naive. Your function is not, through force of your will, to make all the workers do what the company wants. Your job is to coordinate information and reduce conflict so that the members can be successful. Your job would not exist if you did not have team members or teams.
And as impolite as this is, I can't finish without asking managers to contemplate the following: Is there more or less drama due to your actions? Are you demanding team members organise information for you, or are you organising information for the team members? Are you resolving conflict as it occurs, or are you creating conflicts in order to get your way? Do you ask your team to jump through hoops in order to solve a political problem, or do you jump through hoops to solve political problems for your team?
As an engineering manager, you can not succeed if the team does not succeed. It is true that your team can succeed for short periods even if some team members do not succeed, but it is an unsustainable condition. If your team members are not successful, then you have failed. <- Notice the full stop.
Everything starts from the owner ("king") who then delegates things to managers at various levels (starting from the board and executives) who then do their thing and either manage tasks, people and processes directly or delegate them further to lower level managers.
In a decentralized approach, if you need a small task done, you can hire me, give me some resources (money, workspace, equipment, whatever), and expect some results - but mostly leave me to myself. If you want more results, that's going to take more resources - and at some point these resources go from a salary and a desk to a budget and headcount, but in a larger manner, nothing changes; you have given me resources (more of them) and need results (more of them), and you can stop worrying about the details of how these results are achieved. In this approach, the engineering manager role becomes something like a CEO/COO of a small development shop doing custom software for their customers - and the main significant difference between this shop being internal or external is the (lack of) legal contracts required.
Right out of the manager's playbook: "Well, you see, I'd love to give you more than a 3% raise, but, you know, budget is tight. No one is getting more than that. Most people are getting less, you should feel lucky to be getting 3%." You know he's reading from a script that he was coached to read from.
Same strategy as calling into the support line at Time Warner cable or United Airlines with a real beef. You are immediately worn down and dejected because you realize you're talking to someone with zero authority to do anything. So what do you do? Suck it up and keep working, or quit.
We don't live in an ideal world though, and I do not know if the managers who prioritize the long term goals are more successful than the ones who focus on the short term goals (and who, frankly, usually are basically self-serving). Maybe it depends on the definition of success.
Hear! Hear! This completely, followed by managers being too lazy or being too ignorant and just never "getting it".
After many years, I've come to the conclusion that:
1. I sucked as an engineering manager.
2. Most people suck as engineering managers.
3. The only thing that works is a flatter structure with really exceptional managers as actual engineering managers that do what's right and stand up for the team, with a set of people under them that don't actually have management authority, but they can organize things, stay on top of efforts, and communicate status to those that can. This works because really exceptional managers are extremely hard to find and are paid well, so if you try to do anything but this, you will hire bad managers.
Not if it does so at the cost of another team.
I write this from personal experience, having seen teams been disbanded who did a great job shipping... stuff that brought in zero value to the company. This is something I personally would want to avoid if possible, hence me putting "company first" as #1.
Of course no two situations and companies are the same, so there might be cases when having a healthy team trumps doing what the company (thinks it) needs. For now I'm personally sticking to aligning all three where I can: have a great team building things that matter with satisfied members.
How do you get to this point? I have never been on a team where someone further up the business wouldn't come up to us at some point and ask, "Uh, what is the business value of what you are doing?"
Cherish this, and secretly examine every nuance of its ups-and-downs while you can. The basis of all of your future management efforts will be to get back to something similar to this point.
Just remember, a manager is a well-organized servant leader who is ultimately never responsible for success.
What to do when "conflicts of interest" arise is the measure of an Engineering Manager and where the line is drawn is defined by the individual at a personal level.
(now go figure why I'm burned out :-))
If your team is excited to build a service (say, for the technical challenge or the potential reputation gain) that objectively should be done by a different team (say, they have the right background, have already started, or whatever) -- do you put the teams interest first (and do it) or the company's (and don't).
I've faced this situation a lot. I used to pick the former. If we have more impact, how can that be bad? Good for company & so on.
It frequently did end up good for my team, but not always good for the company.
Once I adopted the "good for company" mindset, I found I could often achieve what was good for both, but I did find value in putting the company first.
To put another way: the local optimum is not always the global optimum. As a manager, you need to think about the global & favor that over the local.
My two cents.
I wouldn't make this into a priority list -- whether it's right or wrong, I think it's the wrong mindset. Priority lists are hard to apply in practice because it's rarely clear who might be losing from a given decision.
For one thing, putting the "company" first is a little vague and hard to internalize. Is that the CEO? The long-term success? The stock price? With a "company" first mindset, you'll probably fall into some political games.
Instead, you should start with the actual users of the product and keep that in view. This will help solve the problem where your team executes on a project perfectly and then nobody wants it, or it gets canceled for some mysterious reason. It will also lead you to what's best for the company, make the team successful, and (hopefully) lead to rewards for the team members.
There's some useful advice here for newbie managers (manage your time!), but as a manager, most of really important things you work with -- like building teams and careers -- operate on much longer timescales than a few months. Looking forward to the next update 5 years down the line.
Remember that a "company" is an imagined reality that works because a lot of people believe it. "Team" is as well, but fewer people need to believe it. (Compare those to "nation").
Your team members are real though.
You are absolutely right. The company doesn't make you promises, offer you raises, or in fact owe you anything, people do.
Do you mean this? https://www.daedtech.com/developer-hegemony-the-crazy-idea-t...
I'm not sure I understand your position.
The central idea isn't wrong at all. It's quite clear that we've created a bunch of cultural shared "items" like the the limited liability corporation and such, which are useful but don't really have a physical existence. That doesn't mean they're easy to create or destroy either.
What if I went around the forest and picked one ant out of thousands of colonies and put them all together? Will you still call this a colony? Even if they begin fighting each other?
I wouldn't. It's the behavior of the ants that makes them a colony, which isn't a "real" physical thing.
Unless you're alleging that the ants conceive of a colony, the colony is a machine made out of pieces -- it's like saying a puzzle doesn't have a physical reality when clearly it does: a thousand random pieces don't make a puzzle, and that distinction lies in the physical reality of the system under examination.
In so much as it makes sense to talk about ants, and not clumps of mobile cells or automata of particles, it makes sense to talk about ant colonies. After all, it's the behavior of the particles that appears to make the ant an entity -- the ant isn't real, only the particle behavior.
Your point seems to depend on those entities which we can immediately recognize as entities being more real than other kinds of entities, but that has no logical basis and is an outright appeal to emotion.
If I take half the ants, do I still have a colony? If they now form two colonies, where did the second one come from? How many times can I divide the ants and still have a colony? Where did the colony go if I no longer have one?
Of course "all the ants that form the colony" is a real physical thing, just the same as all the cells that form the ant, but that's arguing in circles. You need to have the "idea" of the larger (or smaller) entity in the first place to be able to point at it.
And it's not the point of the OP anyway, because the ants certainly themselves don't have any concept of the whole colony and don't freely decide if they are one or not. That's just an external observation and classification we made after the fact.
My point is that what makes the colony distinct from the pile is the chemicals, genomic structure, etc that the ants share and are in their environment -- things which all themselves are quite physical. That collection of shared structure between the ants and environmental factors creates a system distinct but made from the ants. Corporations have similar physical components that their constituent humans interact with.
It seems to be begging the question to assume that humans require imagining the corporation but ants do not need to imagine the colony -- it's simply a dynamic system/entity in its own right that happens to have constituent "ants". (Similar in nature to the relationship between cells and animals.)
Every thought, idea, dream or lie could be described as neurotransmitters moving between neurons, but that's very rarely the level of detail we use.
Yes, corporations have physical manifestations. And if we use a formal or legal definition, you could maybe break that down to some database entry at some government branch or paperwork in a cabinet etc. Very real and physical.
But then I could hack the database and insert a new company and fake the paperwork, and arguably the result is not a real company. Or delete a real company and destroy the paperwork, but arguably the company will still exist.
If we imagine I could create or delete a company perfectly in all its physical aspects, documents, electronic records, inventory, products etc - but not modify anybodies thoughts or memories -, you would be very clearly be able to tell if the company is "real" or not.
If nobody knows about it ("believes" it exists) and nobody works for it, it's not real. The office, machines, computers and a bunch of (fake) employment contracts do not result in anything if nobody shows up.
If everybody still believes it exists and just continues to work, they might need a few new staplers and other things, and at some point get the paperwork corrected.
Definitely recommended if you are considering making the transition - or if you are like me and just want a better understanding of how your manager works.
https://www.amazon.com/Up-Organization-Corporation-Stifling-...
Here you go, the simple formula for your management success: it's more software, of better quality, made faster, for less money.
Better to spend more time and money building the right thing than wasting time building the wrong thing.
It's not an official advice to anyone, but if you see what can be improved in your workplace, then go improve, but blindly accepting a position may throw you in your personal hell. All many years before I went that route, invested my time in tools and approaches and it paid back making me and my coworkers happy to do our job. High-stake goals do not work that way, and I can't clearly see what you can achieve by breaking yourselves core. For me it seems like a path to personal failure while making some "dirty" money. Reality strikes back.
This is my personal anecdote on this subtopic, ymmw.
Technical line management is a separate skill set to product & project management, dumping all that on a new manager is an express ticket to burnout city.
edit let's also remember that being promoted from a senior engineer/tech lead to dev manager means you need to find a replacement to hand off technical responsibilities to, managers can't time manage & code and expect to excell at both.
How do people become managers, should I just tell my manager that I want to be a manager now. Would he be thratntened that I want his job ?
Join a rapidly growing organisation and help other people to do their jobs better. You will be a manager before you know it.
Alternatively move company and apply to be a manager and probably say you had more management experience than you really do. I'm sure at 35 you've had to coach some people to be better and do interviews, these can be highlighted.
Do not tell your boss you want his job/to be a manager because anything you ever do that makes her feel threatened from then on will be noted and used against you. People move companies to gain good promotions for a reason.
That's not great advice. Don't lie to get a job.
> Do not tell your boss you want his job/to be a manager
Every company I've worked at has needed more great engineering managers. There's more than enough work, and anyone who says "I want to help" is my hero.
Where I work, when engineers say they want to become managers, we give them gently escalating management responsibilities, enroll them in management training, and give them a mentor for guidance. And later we check to make sure it's still something they want -- it's truly a different job, so they should be able to say "not for me" and go back to being an IC.
It wasn't advice. I was saying it is one of the ways I've seen people transition into management. Interestingly, the way you said management training happens at your company, I've never seen that happen anywhere, so it sounds like a fantastic place to work. How many people who have been through the management program now have a team?
My current employer is more like what you describe, however, and I'm thankful for it.
Has every company you've worked at been successful or at least rapidly growing? There's plenty of cake to go around in a company doubling in size every year, but the daggers come out when there are layoffs twice a year, particularly when the people being laid off aren't SV software engineers with tons of other options.
I don't entirely think that piece of advice applies if your manager/boss is ambitious. If they're ambitious, then they're most certainly looking for their own replacement. Heck, someone that can in the interim do their tasks while they focus on "moving up".
Don't tell your boss you want their job. Tell them you'd like to have an opportunity to contribute in a greater capacity and how you can do that. If there's something specific that thrills you, ask to do that. Put 100% into making sure that thing gets done well.
I think it's normal and professional to have a conversation along the lines of "I'd like to start getting experience managing people, do you think there'd be a chance for me to do that?". If they're threatened by that conversation, I wouldn't want to work for them frankly.
The reality is there's a lot of older coders out there, I think it matters more what you _want_ to do.
I've always told myself that I'll code until I can't keep up. So far that hasn't happened. Once it does, I will happily transition to more of a management role to be of the greatest value that I can. Along the way (and even now), though, I've had a good number of management-like responsibilities which will help a lot. So, I encourage you to take advantage of those opportunities as they come along. There isn't a bright, hard line between management and individual contribution. There is a space between.
A good manager will look out for your career path and help you get to management, if that's what you want.
I think a good manager sees the second two as an important factor in putting the company first. Demoralized teams are bad for the company.