Why I am considering getting into management

Why I am considering getting into management

I've reached a point in my career in which I'm questioning a lot of things about being a Software Engineer, don't misunderstand me, I love being a Software Engineer and getting my hands dirty writing code, implementing things, designing systems and discussing architectural trade-offs with other Engineers, the mix of art and science in what we do is amazing, but I'm afraid growth paths as an Individual Contributor (IC) are more complicated and less documented than the ones for People Managers (PM).

As an C, it seems like your growth will depend on the impression others have about your individual work, those others are usually PMs, as a PM, your growth depends on the performance of your team and the ability to achieve those quarterly goals.

Here are a few things I think are a bit odd when it comes to growing as an IC.

Product > Engineering

It might seem strange because in some companies the organization in charge of building and maintaining the user-facing systems it's called Product Engineering, but quite often Engineering is a sidecar of Product which is supposed to only support Product, with that in mind, all of the Product-related goals are to be supported by Engineering efforts: building new features, enabling experiments, investigating/fixing bugs and other types of issues, you name it.

Engineering is a feature factory for Product to a point that in some organizations, Product Managers even estimate Engineering effort and commits time and deadlines to Senior Management on behalf of Engineering teams.

This also translates on how everyone is rewarded, if you build new features, you get recognized and attention, if you reduce the latency on an endpoint by 20% by optimizing the database calls and indexing a few columns, you get a pat on the back.

This also has a huge influence in the accumulation of technical debt, which is one of the main elements of frustration for Software Engineers.

IC goals are subjective

Quite often, your growth is linked to your day-to-day job plus a set of goals set with your manager. These goals could be improve some system's reliability, become better at some skill or learning some technology and implement a poof of concept. But, how do you measure success on those goals? how do you measure a skill? are you going to apply a test?, what if throughout the performance cycle there's no project where a new technology can be used? what if throughout the performance cycle I don't get to work on the system I had to improve reliability on? What if I got to work on it but the reliability was improved as a byproduct of another project done by another engineer?.

Management goals are related to the team's performance, improving some business metric, improving talent retention. All your team knows they have to perform well because part of their growth depends on it, so, as long as you're gentle and have a little common sense treating people fairly, half of the work is done for you.

It's OK to be wasteful

When an Engineer is leading a construction, they calculate how much material is going to be needed and how much debris will be left to be managed and disposed and they purchase based on that. Any miscalculation is not good, they try to optimize resources to the minimum, the same happens with everything else outside software. Chefs optimize ingredients in their recipes and try to reuse as much as possible to reduce waste, tailors also optimize the use of the fabric.

Software Engineers on the other hand, well, we design and implement something and never look back. It's consuming too much memory, let's throw some more RAM into the servers and that's it. CPU is spiking, let's buy instances with more computing power. Servers can't handle the load, throw more servers in. We just throw resources at the problem without thinking whether our system design and and implementation can be better, without revisiting for improvements. This is also aligned with the previous item, because Engineering takes the back seat while product drives the car, Engineering pains are postponed until they're hurting really bad, and even then the solutions are rushed!. "Will it take more than 1 sprint?, we can't afford to delay features X, Y and Z"

The main reason I've heard is that "Engineering time is more expensive than machine time", but then, machines won't solve scalability issues caused by lack of design and rushed implementation, you're just paying interest on that loan given in "machine time".

Hiring ICs is broken

Yes, if you need/want to switch jobs, your options are limited to the companies working with technologies you have experience with. It doesn't matter how quick you can pick things up and learn new languages or technologies, if you have no experience some companies won't even consider you.

Just an example, in my current role at Careem I've worked with Java, maintaining services built with Springboot, Groovy, Bash, Terraform, CloudFormation, Python and Jenkins, building a CICD platform, Nodejs, Typescript and Expressjs building and maintaining services for another business vertical and PHP, Go, Elasticsearch and Kafka, building a search platform for the Food and Shops business verticals. Prior to Careem, I worked about 6 years in Python and Django, I had zero experience with any of the technologies I've worked with at Careem, where the main language is Java, if the interview process was focused only on Java and JVM tuning, they wouldn't have hired me. All of the technologies I had to learn them on the job. Asking other Engineers, reading documentation, reading a book, doing an online course and getting my hands dirty doing toy projects to learn.

Then, there's the coding interview. Complex algorithms and data structures questions I've only seen in programming competitions, infinite space search, optimal routes finding, graphs or 3D matrices traversing. Things I was once quite good at because I was really into it, when I was in Uni, but now, it would take me some time to dust off that knowledge. If I find one of those problems in my day-to-day work, I know I'd be able to figure it out with some thought and research, but in interviews you're often asked to solve 3 to 4 problems in a 90min time.

Systems design interviews are fine, you must be able to architect a system and think of the components you'll need to built in order to have an optimal behavior. But coding interviews feels like fighting Godzilla with a screwdriver while the day-to-day work is actually playing UNO with Barney and his friends.

Management skills are transferable

Engineering skills are not, well, they're partially transferable. Good engineering practices, Architecture and Systems Design can be applied at any company you land at. Unfortunately the skills with the specific technical stack can't be taken from one place to another if the technologies are different. A lot of companies will hire "Experienced Java Developers" and just cut the interview cold or just don't take your profile into consideration if you don't have 5+ years of Java programming experience.

Now, let's talk about the coding interview, the skills to pass a coding interview are not maintained by your day-to-day job, you won't work with graphs or puzzle-like problems every day, every time you're going for a technical interview you must spend a few weeks preparing, some people solve 400+ problems in leetcode in order to be prepared for the coding interview.

Management skills are practiced every day at work. Looking at metrics, dealing with people, negotiating deadlines, planning a roadmap strategically, evaluating trade-offs, mediating conflicts, even experimenting with team topology and different methodologies. All these skills are valuable at the time of the interview and also at the new company you land at.

Conclusion

Well... switching to management is not a decision I've made already, it's just something I'm considering, for all the reasons I listed above. I love getting my hands dirty with technology, designing and implementing a system and watching it work and serve millions of requests per minute. I enjoy working on refactors and tuning systems for better behavior, but the manager path seems to be much more straightforward than the IC path. The one thing I'm not so comfortable with is having so much impact on other people's career and life in general and, in crisis times, manager heads are the first to roll but, again, then landing in another job doesn't require solving 400 leetcode questions and pray you get some similar problems in the coding interview.

What do you think? have you ever thought the same?