Slashdot Mirror
Ask Slashdot: Transitioning From Developer To Executive?
First time accepted submitter fivevibe writes "I'm about to switch from a position where I did hands on development to one where I will be building and managing technical team. I will be responsible for designing and implementing the company's overall tech strategy. I am excited about this move but also nervous. It will require a different focus than I had up to this point, different skills, and different orientation. What should I be learning, reading, thinking about in order to make this transition successfully and avoid growing pointy hair?"
(Probably not the sort of ignorance you're thinking of, though.)
Start practicing saying "I don't know." You know a lot of technology right now, but in 5 years you'll know less, and in 10 the young kids will roll their eyes when you talk about how it "used to be." Set a big organizational goal ("double our storage space for next year") and then ask the technicians how to make it happen. Resist the urge to do anything more than "suggest" things or vaguely hint at solutions. Know how little you know.
What you shouldn't ever forget is how technology "really works." You know, "fast right cheap pick 2." If your company wants to go with a cheap solution to their problems, make sure you've prepared properly for it.
All the successful technician-to-manager folks I've worked under have suggested solutions, listened when technicians explained problems and tried to get managerial roadblocks out of their way. On the plus side, the best managers I've worked for were promoted techies. Good luck!
Ack!
1) find a mentor you can use to get advice and bounce ideas of of
2) contact some counterparts and see what professional pegs the belong to and join them and go to local meetings
3) and now the hardest part. While the developers are your friends you now have new responsibilities and may have to make some tough decisions. Be fair, but make the tough calls. If you don't, your team will suffer and do will you.
I'm a consultant - I convert gibberish into cash-flow.
First of all, read "The Mythical Man Month" by Fred Brooks, if you haven't already.
Be realistic and conservative on your delivery dates. Defend them to the death.
Avoid micromanaging people, if possible, and insist on clear communication and concise documentation.
My personal suggestion: don't give up completely on being a developer. Keep a small, but important task to yourself. You will gain an even better view on how your team is working.
Hack your mind out of its sandbox.
Hi,
A difficult thing will be: you have to trust people doing the job, even though you know that they are not good as you. You will get back solutions, that are not the same you would have delivered or may even not be up to the standards you expect. You must take a step back and ask "Does it suffice?" and not "Do i like it?".
There are two big dangers:
a) Trying to do your previous job in addition to be a manager. This will kill you. The result will be abysmal performance in both jobs.
b) Having no reserves in your schedules to talk to people. This will get you disconnected and you may not realize problems until they bite you in your posterior.
The most difficult thing for me was, that i learn things about people i never wanted to know. You have tragedies (child/husband/parent dying of some illness), relationship problems (both sides being in the company more often than you think), all kind of quarrels (If n is the number of persons you manage, the number of conflicts is O(n)) and so on.... You have to develop a thick skin concerning this. If you cannot, step back. Otherwise it will break you.
Another lesson learned: If you make a decision, never postpone it. Pull it through with max burn ;-).
After 8 years i had enough of that job and went into sales....
Good luck, Martin
After 13 years as a systems guy/programmer, I ended up as manager of 12 similar people. The University had gone through a restructuring and a few resignations and I thought it would be the right thing to do, since I was recognised as the most capable in the team (no false modesty here!).
Four years later, I left the University to go back to being a systems/guy programmer, working for a small Swiss proprietary fund (my current employer). Reasons:
1. Meetings. Endless. Bloody. Meetings. I'd been to fortnightly team meetings as a programmer. As a manager there was at least one sort of meeting with someone in the University every other day. Protestations that email or other collaborative software would save everyone time, mileage and money were met with indifference - other managers seemed to enjoy the stupid things.
2. Stress and Responsibility - two sides of the same coin. When you're in charge of a group, the buck stops with you. This can wear you down after a while. It certainly did with me. Whilst I was immensely proud of the team and what we accomplished, occasionally things do go wrong and for some reason the customers never remember the good times.
3. Health issues. My underlying, but previously unobtrusive OCD was exacerbated by 1 and 2 above. I grew afraid (shaking, uncontrollable fear) of meetings, eventually getting to the stage where I would leave them mid-way, or invent excuses not to go in the first place, or just not turn up. Whilst my managers were sympathetic, I became unhappy with the way I was doing my job, which of course reinforced the "bad thoughts"-side of my OCD. I was off sick from work repeatedly, sometimes for days at a time. I received professional help and medication for the OCD and got back on a somewhat even keel, but realised that I would never be happy in my job. When the opportunity to get back to programming and systems work arose, I took it enthusiastically.
Now obviously your mileage may vary and my comment may be utterly useless. I guess the point is that a good programmer may not be a good manager. A person who enjoys working directly on problems may not enjoy giving the problems to others to solve. And a person with any sort of mental issues may find them more exposed when working as a manager!
Coders often suck, especially at estimating effort of time
It's not necessarily that those coders suck, it's more that it's impossible to estimate the time to do some non-trivial new task, because there may or may not be hidden depths.
Even Donald Knuth can't estimate how long it will take him to do something, and he has a lot of experience with algorithms and coding. I think the numbers were that he expected TeX to take two years to write, but it actually took ten.
I think it's better for the manager to pad the numbers but not let the engineers know. Hold them to a tight-ish schedule, assuming that they will over-run sometimes. It's good to feel a little time pressure to keep you focused, but not so much that you get despondent. Allowing for explicit maintenance/refactoring time on the code would be important too if it's a project that has grown and morphed over the years and needs tidied up.
I don't think micromanaging is the answer. If you ask me how long overall something should take I will be happy to give an answer - but I don't like giving a schedule of every thing that I will be doing, because I simply don't know in advance. Sometimes things move way faster than I expect, and sometimes I'm banging my head against a wall for a couple of hours because of an oversight in my design.
which is totally what she said
Keep on reading those journals to know what is possible and ignore those losers that call themselves "Architects", "Engineers" or even "Gurus" without some professional group of peers that think they deserve the title. You don't have to have earned one of those titles, go with what you have earned and keep in touch with it enough so that no amoral contractor can bullshit their way into robbing you blind. You don't have to be a cutting edge expert but you do need to keep up enough to tell one from a confidence trickster.
It doesn't all stop when you leave school or even the "shop floor".
2. They are being paid, make sure they do the work they need to by the time it needs to be done. Stick to schedules. I can not stress how important it is to stick to schedules. If a programmer can't meet targets you feel were set fairly then you may have to fire him/her. (...)
4. Designate a planner. This will probably be you. The planner takes the goal and the design and makes it into a step by step development cycle programmers can follow. (...)
You may not know it, but the people working for you probably think you're a lousy manager. You're the traditional project manager coming up with an estimate based on how hard it sounds like doing without taking any input from the ones actually working on the code, then drive people hard to meet your imagined schedule. From the "watch every commit" it sounds like you're trying to be the supercoder micromanaging everything everyone under you is doing. Chances are that if you're trying to do that much at once, your quality will turn to shit too even if you could outperform any one of them individually. Particularly if you're doing any part of the managing bit, making sure all your people are productive, clearing roadblocks, dealing with recruitment/staffing/budget issues, management reporting and so on. If you're serious you should get out of management, quick.
Live today, because you never know what tomorrow brings
Yes, I've seen several people "promoted" into a manager or VP of Technology type of position who were simply unprepared for the transition. It was felt by the (foolish, ignorant) top execs that the candidate would make a good manager since he was already such a good programmer or architect. The predictable result was catastrophe; the former star programmer became a stupid git who was hated and despised by everyone, and eventually was shown the door after having caused incalculable damage to the organization.
Unfortunately for all parties involved, software engineering and management are two completely separate skills. It's like saying a good surgeon would make a good hospital administrator--where do you pull that out of? They are unrelated and often oppositional jobs. Shoot-from-the-hip, cowboy programmers as well as "team player", 9-5 coders are equally unqualified to manage others, with the team players being slightly better simply because they're less likely to piss everyone off right away.
I've found in my professional experience, as have many others, that really great managers are born, not made. Some people seem to have instinctual abilities to see through the fog of war and focus on the goals, marshal their resources rationally, and avoid letting petty emotions, vindictiveness, oneupmanship, and all the politics that come with human interaction get in the way. You can call them out, tell them to their face that they're full of crap, challenge their assumptions, and they will calmly roll with the punches and adjust accordingly. Their bosses will lean on them, change their priorities, threaten them, etc., and they will push back, educate their superiors, win the time and money they need to achieve the objectives.
It's like sales. Great sales people seem to have an innate ability to close the deal, while crappy sales guys (the majority in the world, I think) piss off the customer, display their ignorance, and basically fumble the ball more often than not.
So, op, proceed cautiously. You are about to step off the cliff and hope you enjoy a soft landing. A few make it, but most don't. I would say, if you're pretty good technically, you should just stay in the tech field where you know you have a good future. But, if you want to learn from bitter experience, be our guest and delve in. Get back to us in a year and let us know how it's going.
it's = "it is"; its = possessive. E.g., it's flapping its wings.
Most "good" managers I've met are not good because of skills or training, but from simply being personable, intelligent, and able to solve problems (real world problems, generally very different from the types of problems programmers face). It takes a minimal amount of training to get a good manager, as long as you start with the right person, who possesses those innate abilities. There is plenty of management material to be found among software engineers, programmers, IT, etc. The problem is that, like you said, execs will often just assume someone good in one position will do fine managing that position, and promote them without any management training.
The truly unfortunate thing is that it generally takes much more skill and training to be a good software engineer than it does to be a good manager, and yet management pay generally BEGINS where every other job maxes out. My biggest problem with about 90% of all management I've worked with in my life is that I (or many other people I know) could do their jobs better than them with a week or two of training, and yet they're making twice what I make. At the same time, most of them also face less stress and work fewer hours.
Well, I think it is hard to generalize the way you have and be correct. I'm sure there are shops like the one you described - managers making much more than the devs, worrying more about their golf handicap than the project timeline. There are plenty of places though where the dev and manager payscales have quite a bit of overlap, where you'll find all of the senior devs making (much) more than the junior managers. I think this is right. At well-run companies there will also be quite a lot of pressure and stress put on the manager, simply because the manager is responsible for the success of every person on the team - so take all the things that can go wrong on the dev side (hit a snag and have to refactor, sick time, etc.) and multiply that by the size of the team. Good managers are also taking the heat for making the inevitable tradeoffs - "yes, we know big client X wants feature Y but we need to keep the release on track." Dealing with VPs several levels up trying to pull the project in different directions is also less than enjoyable.
Managers are also much more vulnerable to politics than individual contributors. You can be a great manager and still get canned if new upper management rolls in and doesn't like you or doesn't think you're the right person for their new policies - or if you get a tough project that doesn't go well and someone needs to be blamed. So, I think part of the compensation difference is because the job is simply riskier.