Hacker News new | past | comments | ask | show | jobs | submit login

This got me thinking about role of managers in a modern org (to simplify things let's say it's a Tech company).

It certainly needs a CEO/Visionary, it most likely needs HR and front/back office folks, it certainly needs PR and marketing people. But in a world where people communicate rarely in person, have their own management and economics 101 abilities, are smart enough to not work against their own interests (and look after the org's interests) - what's the role of the future manager?

It sounds inevitable that senior Engineers will double up as managers for their group as and when required (working with marketing etc.) instead of it being a dedicated managerial position.




One of my ex managers said that his role was to attend the meetings,shield the developers from the politics of getting projects approved and other institutional overhead so that we could focus on delivering product.


There's a bit of irony in there that these are mostly people created distractions and the org throws more people at it in the hope of solving it!

I think there will still be need for excellent managers to get a diverse group of people to do stuff together, to keep the vision coherent and to motivate people to deliver their best. It's just that it's a lot to ask of one single person and the position is vulnerable to Engineers grasping those skills and having a huge advantage in actually understanding the low level stuff much better.


Sounds familiar. As a USMC fire team leader, my job was mostly to protect my Marines from all the shit that rolls downhill. It's generally a very difficult job to do well, because of the way the explicit incentives are set up, but somewhat easier for me as I was basically apolitical and had no career plans in the Corps.


Blog.

Please?


I keep a blog about miscellaneous stuff at http://blog.byjoemoon.com (plug) but I don't think that's what you mean.

I got out in 2006, and I had already stopped writing by the time I made team leader, but I did keep a blog for some of the time I was in, and I collected it (the interesting stuff) at http://www.servicerecordbook.com.


"You can either be a shit funnel or a shit umbrella" - Tood Jackson, Product Manager, Gmail http://techcrunch.com/2010/03/14/key-to-gmail/


This is a great role for a manager to play for as long they have the power to actually shield the developers from politics.

When things go bad politically for the manager and they lose that power, the politics have a way of suddenly rushing in and catching the developers off-guard.

So developers, inspect your manager for leaks regularly.


> shield the developers from the politics of getting projects approved and other institutional overhead so that we could focus on delivering product.

This. This is very true from my perspective. I got a few glimpses of the amount of bull our boss is shielding us from, and boy I don't want her job!


Did you used to work for me?

I ask because when I moved into management and worked with a stellar group of extremely experienced engineers, far better AND more experienced than I ever was, that's exactly how I explained my job to them.


> "what's the role of the future manager?"

Personally I've always seen the organizational structure of film and TV units as being a natural model for software development [1]. They've been doing the somewhat-controlled production of fundamentally creative product by small, focused and often-times transient teams for quite some time.

And in that model, the current manager type is less like the director [2] and more like a producer.

That is: a person who primarily manages logistics and external requirements.

e.g. arranging the times to get the necessary people together to review builds and such. keeping statuses up-to-date. keeping an eye on tasks and deadlines. keeping their finger on consultants. etc.

[1] Far more natural than the manufacturing model that corporate America is trying to fit things into.

[2] I'm sure managers would like to imagine themselves as directors. But they're typically ill-suited due their not understanding the creative nor technical sides.


This sounds like the studio model that the AAA video game publishers have been using for some years now. While they have in-house production and coding for many packages, more and more the big-name titles are implemented by small remote teams that focus on the meat-n-potatoes while marketing and distribution and project management is handled by the parent company.


But in a world where people ..., have their own management and economics 101 abilities, are smart enough to not work against their own interests (and look after the org's interests) - what's the role of the future manager?

I don't know if there's as many people who have these abilities as you think there are.


Likely not but it still makes sense to give a chance to all the few talented people however many there are to actually create something instead of manage other less talented souls. Better for productivity and happiness IMHO!


Large corps are hardly run for employee productivity or happiness. They are run to maximize short term personal wealth of managers, and (less importantly) shareholders.


This sounds like a rationale for implementing the Peter Principle.


Microsoft has done this all along. Managers are programmers, they divide their time between programming and managing. The feeling within Microsoft, was how can someone who is not a programmer, manage programmers?


I'm pretty sure its illegal for non-lawyers to manage lawyers.


Lots of companies have in-house legal counsel - they can't all be managed by lawyers unless it is lawyers all the way up to and including the CEO.

What it might be is that law firms in most places can't have partners who are non-lawyers, but that's quite different.


This is incorrect. The company I work for has a legal department that reports to upper management.


My mistake, I just heard someone mention it before.


And look how well it's worked for them ...


The problems started when they forgot this rule for senior management and made the CEO a sales guy instead of an engineer.


It worked for them well for a long time. From what I hear managers there finally have managed to take over.


May we all be cursed with Microsoftian levels of failure.


A multi-hundred billion dollar valuation, a monopoly or two and the founder being the richest man in the world for years is good... right?


I don't know of many engineers who would want to double up. I do agree that a well managed company could get by with far less management than most currently have, but don't think that just having really smart people in place will solve the problem magically. I've worked at a company full of incredibly smart people who couldn't get anything useful done because they were all so busy proving how smart they were, often at odds with others in the group.


What makes you think mangers don't suffer from the same thing. Managers are also highly educated often more so(MBAs etc) etc except they will be trying to employ fancy management rather than fancy algorithms.


I think many (middle) managers are afraid of Scrum and other agile development because their role is sidelined. As "individual contributors", engineers are generating actual business value. Engineers can self-organize in "quasi-communist" Scrum teams (as they own the means of production ;) and collaborate with product management who work with customers to derive and prioritize product requirements.

In this scenario, what is a "people manager" supposed to do?


The job of a people manager at Google is generally to keep the engineers happy and make them want to continue working for Google. Somebody has to be the point person for general career-development issues, so that if you get bored and frustrated with your current position, they can say "Here, check out this other group over in this other area of the company, you might enjoy working with them" instead of quitting abruptly. Somebody needs to make the high performers feel valued, and let the low performers know that they need to improve. Somebody needs to be the point person to handle things if a family emergency comes up and you need to take several weeks off, making sure that your responsibilities are covered by other people and all the administrative stuff is done.

The good managers are really good at this. The bad managers...well, I don't work in that part of the company.


But aren't those manager doing engineering work as well?


People managers rarely do engineering work as well - it's very hard to both manage people well and contribute significantly to the code for a project. It also sets up a sort of perverse incentive where you want to look good as a coder by taking on the glamorous, high-profile programming tasks, but this can actually demotivate your teammates because they're left with the drudge work. You don't want to be in competition with your reportees.

They're often drawn from the pool of former high-ranking engineers, or at least the good ones are. It's a very different skillset though; part of being a good manager is knowing when to step back and let your reports handle things, and the worst managers are those that haven't managed to separate their engineering ego from their team building ego.


Very true. I worked for a startup some years back where the technical cofounder had a nasty habit of reserving all the interesting problems for himself and fobbing the scutwork off on employees. It was intensely frustrating. Yet somehow they never understood why they couldn't retain programmers...


I've helped organizations transition from waterfall to agile, and there was absolutely no decrease in the workload for management.

You still need some form of functional management to get the right staff, make sure they have what they need to be able to do their jobs, and deal with day to day issues. This might sound trivial, but it really isn't, and somebody who does this job well will enable a lot more success than somebody who treats it as a minor task.

And you also still need some form of product and project management to ensure that the work being done is aligned with strategy, marketing, sales, finance, etc, and to serve as an interface with those groups. Failure to do this effectively creates massive inefficiencies.

As such, I really don't think any moderately intelligent competent manager is at all concerned by Scrum, Kanban, XP, TDD, Lean, or any other Agile practice.


Smooth things out in the event of engineers being on the verge of shooting at each other. That is, manage human beings to let them work at their best, not timelines and general process flow to achieve a financial goal.


The scrum teams don't own the means of production. Individual contributors are human capital. The means of production are owned by the people in control of the finances and business model. If the company were to shift direction and your product discontinued, it's possible the scrum team would be disbanded and the resources that were being used to support them would be reallocated by the "owners" to a new team that aligned with their decision to shift strategy.


The team could also quit en masse and go found a startup to do what they'd been doing before, if they really believed in it. It's happened to more than a few big companies.


Yes they do - just like lawyer owns his skills and can take them elsewhere and continue his work.

A factory worker on the other hand, can't (legally) take the machines with him.


Ideally, but power still seems to be centralized to dealmakers (probably a function of human nature), whether they're specially skilled or not. Managers are essentially low-level dealmakers.


Ppl who can make win/win situations out of lose/lose situations are important. Isolated technical decisions are often the easy ones. The tough ones are getting two people or multiple groups to come to an agreement on technical decisions w/o any feeling like they were just burned or stabbed in the back.

This is basically Steve Jobs. Steve is the consumate manager. Not an extremely nice guy, not a designer, not all that technical. But people value and respect his opinion. And even more importantly people want to deliver what he wants.

You know a great manager/dealmaker because he's the person who when they change groups, you want to go with him because you feel like he'll take care of you and your projects.


Agreed. Good post. I probably shouldn't have said "ideally", because that makes it look like I don't see value in dealmaking.

I should have said, "ideally (for you)"


From the "Dear John" letters of people departing Google (Dennis Crowley for example) it seems like in the absence of a large political institution, the major political currency at Google is your ability to attract developers.

Given that, it seems like what a manager can provide a group of developers and designers is "workforce maintenance". That is, they will attend to the needs of the team, make sure people don't leave, and make sure they attract more people when they need them.

In other words, HR. The central HR department at Google has no reason to really care about the success of your unit over another, so your manager becomes the person who takes responsibility for the human resources of the unit.

This frees the engineers/designers to be able to focus on creating a great product. And as a side benefit, they happier and their jobs are easier because the manager is doing what it takes to keep them in the unit.


How is that in any way a "Dear John" letter?


The goal of management is essentially the same as the goal for any kind of political structure: remove issues which couldn't have been dealt with in a cheaper way.


And self-preservation of the management political class. How often does a corporate reorg reduce the number of managers?


Layoffs are routinely used as an opportunity to flatten organizations by removing layers of management.

I don't think I've ever seen a corporate reorg where middle management wasn't amongst the hardest hit groups. They're expensive, and you can usually fire a huge swath them without thinking, tell the remaining ones to start working late and limp along until times get better.


Are you kidding? Middle Management are often the first to go. They produce no real work, they just provide less work for upper management.


No thats what it becomes, not what it should be.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: