Awesome piece below was written by my VP Rocco Seyboth and I just had to share. Let me know your thoughts in the comments!
Every morning, I see the unfiltered thoughts of 1400+ engineering leaders as one of the community moderators in the Dev Interrupted Discord server . We start every day with a Daily Interruption topic about how to make agile work in real life; scaling teams, building culture, hiring, continuous improvement, metrics — fun stuff like that.
Recently this Daily Interruption popped up and stopped me in my tracks:
How much control is the right amount of control?
Nick might as well have asked, “what is the meaning of life?” You can see my immediate reaction was bewildered introspection. 👇
A bit of context…
I was born with a default control setting of 10 (out of 10). I believe your strength is your weakness. At least in my case, this has proven to be true. Like many controlling people, I take ownership, obsess over little details and get the job done. Also, like many controlling people, sometimes I have a hard time working with others. I’m putting it mildly. 😄
So what do you think happened when I got my first job working at a software start-up?
As an individual contributor, I crushed. My super controlling nature propelled me to dominate every task that was assigned to me. I overachieved.
Then I got promoted. And just like my friend Dan Lines said about being promoted from dev to team lead , “a freight train hit me.”
Since then I’ve been on a journey to figure out how much control is the right amount of control when it comes to leading software teams. I’ve been managing people for sixteen years now and I can break that time into three distinct phases:
My three phases of controllingness
When I first got promoted to team lead I was highly controlling. I literally did most of my team’s work for them. I worked seventeen hours a day six days a week to ensure every single task was completed to my exact specification. The people that worked for me were unhappy (some actively disliked me personally) but we got results that the CEO cared about so it went unnoticed.
And I was good at managing up, so I actually got promoted for this behavior! I was in my early twenties and motivated by the wrong things (power, money, and, of course, control). I look back on the period with embarrassment and I’ve actually apologized to many of the people who worked for me back then.
I’m a person of extremes so when I realized micro-management was wrong, I naturally swung the pendulum in the exact opposite direction. I told myself I was hiring smart people and I should leave them alone. I’m good at hiring so it kind of worked. But, again, the people who worked for me suffered — this time in a way that they noticed much less. Good people actually want feedback! It’s not good for their work to go unchallenged because then it’s harder to improve. My teams in Phase 2 all had a lot of fun and liked each other but we were a bit chaotic in how we got work done and we were not living up to our potential.
I like to think I’m currently in Phase 3 which is more of a happy medium. I try to do four things to attempt to strike the right balance:
Write super clear job descriptions and goals so every person knows which areas they own and which outcomes they are responsible for driving for the business.
Provide extremely honest feedback… sparingly. You have to pick your battles. I find a lot of feedback is good upfront for new employees. And then, over time, you have to give people room to make their own choices or make mistakes. I find my people actually prove me wrong a lot of times when I think they are going to make mistakes anyway which is a good reminder to keep my mouth shut.
Occasionally I decide I’m going to personally own something that needs to be handled my exact way — like when an important new project needs to be bootstrapped with care and skill. In those cases, I let my control freak flag fly and I just do it my way. It’s ok to do this very rarely.
I warn everyone upfront I am a recovering controlling jerk and apologize constantly for when I step over the line which I still do all of the time. 😁
Phase 2 is just as bad as Phase 1
Managers in phase 1 get all of the credit for being the worst but we should not underestimate the damage that can be caused by phase 2 controlling managers — ones who do not “control” enough.
I found this response from drdwilcox (a VP of Engineering) fascinating.
When I reflect back on my time in Phase 2, I realized it was my own insecurity that stopped me from communicating and coaching more. Once I put my imposter syndrome aside and realized I just needed to do my best for everyone on my team, I was able to strike a balance between too much input and too little feedback.
Engineering metrics — a tool for good and evil
“Communicating expanding expectations that come from growth is such an important part of what I do with my leaders.”
Engineering metrics are probably the most common topic in the Dev Interrupted Discord as well as on our podcast (with the same name) . Everyone has ideas about which metrics are good and bad and everyone has a story about a time that a bad manager used metrics to control their team in a negative way.
Phase 1 managers often use bad metrics to stack rank engineers and pit them against one another or just force them to work harder. Phase 2 managers don’t share enough data and miss an opportunity to use good metrics to unite the team.
If you’re trying to become a Phase 3 manager, our community has tons of resources about how to use metrics to help your people improve and increase quality and efficiency among your teams.
What do non-managers think about all this?
Scott, a “never-ever-a-manager”, has insightful and hilarious things like this to say almost every day in the Dev Interrupted Discord 😆
All are welcome in the Discord so please join us and share your thoughts and controlling manager stories!
And please consider joining our newest event: Interact, the community driven conference for engineering team leads, managers, VPs and CTOs looking to improve themselves and their teams.
Originally published at https://devinterrupted.com on June 30, 2021.
Top comments (5)
I swear that my control issues are the reason I got into software development, haha! It's definitely caused a lot of issues for me in my career. I feel like good metrics need to directly relate to core business focuses. That varies wildly between companies, but at a basic level it comes down to "how much $ will this add?" Value takes lots of forms though. Maybe this PR fixes some infrastructure scripts which bring monthly server costs down. Maybe we can concurrently serve more users and increase sales and visibility metrics.
I hate to see metrics related to story points, lines of code, etc. If your team isn't capturing requirements well, it punishes developers who choose to work on more difficult issues that probably have a huge effect on the overall bottom line of your company, and can actually discourage people from innovating and taking the risks needed to remain relevant.
I feel like communication is a huge issue. There is an incredible degree of trust between my current development team, and we all feel safe calling each other out if we perceive that something threatens our ability to deliver or our culture. We speak to our sales team very frequently to make sure what we are doing is relevant. If we didn't have a high degree of trust and a safe environment to work out of, none of that would be remotely possible.
I was a developer for years before I became a dev team advocate. In that time, I saw lots of different leadership styles and I've gotta say, communication really is key.
No one wants an overbearing manager, but an absent one doesn't do much good either. Feedback is nice; it's nice to feel involved and part of a team. So whatever quirks you have as a manager, if you can nail communication, you can probably find the right balance between freedom and control for your team.
I don't quite like it when things related to management are formulated as recipes. It's important to use the correct approach with people depending on how they respond. Also, control is one thing but specifications don't quite fall under control unless you start specifying what algorithms to use, tools and so on.
To me, specifications should be the minimum needed so that the output can be objectively verifiable (eg: acceptance criteria). Eg: give me a REST endpoint to list users vs "give me a rest endpoint to list users that - receives a parameter for pagination - has a certain response structure - uses certain http codes for various conditions ... etc"
But also ... if there's a control secret then it's that a manager's need to "do the stuff that needs to get done" isn't about control. Is about definition what needs to be done (eg: YOU should know what's needed unambiguously first) and then it's about getting the people around to want to do that (and empowering them).
This means speaking to the motivation of people (whether they're motivated by a paycheck or the threat of losing it, by the tech involved, by the challenge ... etc).
I would say that if the ability to get people to do what's needed gets limited to controlling them (watching over their shoulder, doing it instead of them or taking every little implementation decision) then it's probably a big issue.
This article resonates with me pretty well. My current job I was hired as a senior software engineer to be a skilled worker for a SaaS startup. As we got more developers, mostly juniors, it naturally fell on me to become the lead and work with the team to coordinate dev efforts and get things done. Figuring out how much control to take has been a struggle for me personally and I have gone through each of the phases you describe.
While I am not sure I have found what works best I have found a mix of the strict controlling and loose "do what you want" has been working. As an example I will let the juniors go and try wild things with little oversight, then bring them back in and explain why what they did was a poor solution (if there is something wrong). I think it is important to give your a team a healthy chance to fail from time to time.
Now the next hurdle for me is optimally balancing the responsibilities of architecture, development, and team lead responsibilities. WIP.
Interesting. I suppose as with most things in life moderation is the best approach. Anyway, thanks for publishing.