Hacker News Reader

Comment by robertclaus | Hacker News Reader

For most junior folks I meet, I recommend starting their career at a big company rather than a startup. In my experience, the dysfunctions at big companies are too much process (which you can still learn from) whereas the dysfunctions at startups are often a lack of process (functionally unknown unknowns). I've found it's much easier to solve process issues at a startup after working at a big company than vice versa.

View on HN

Hmm, I’d recommend the opposite. I don’t think you’ll ever full recover if you start your career at a large company and start to believe that that kind of dysfunction is normal.

Startups have issues too, but they’re more often related to the technical parts, which you can actually do something with and learn from early in your career.

If you go to the large company what you learn is navigating their particular brand of process.

This, starting out at a startup. When you later go onto a big company. They'd be surprised by your ability to build a product end to end, from design to implementation to infrastructure.

The only part that's missing is handling the enterprise processes, multiple stakeholders, chains of approvals etc. But those are easier to learn than building something from scratch.

It’s kind of depressing when people do not want to listen to you because you don’t have the right title though.

Yes, I know how to optimize a database. Yes, I know how to build infra. Yes, I also know how to build a system used by 30k people daily.

Yes, I’m fairly certain that our system that serves 3k people does not need to cost us $30k/month when I could do 30k for $1000/month.

But nobody wants to hear that. They’re all secure in their little fiefdoms, and your boss will certainly not be rewarded for replacing their 30 people team with three of you.

My gut reaction is, lots of people can build a product themselves from end to end. That someone has done this is no actually no surprise. Startups and big companies are not so isolated from one another that the denizens of big companies are ignorant about what's possible. Within just my part of the organization are people who have worked in quite diverse environments, including startups, academia, government, and even bigger companies.

I went from a startup to a big company, 25+ years ago. We buy startups. I'm kind of a jack-of-all trades. One of my roles is to serve on the technical side of taking a small company apart and figuring out how it works, so that it can at least be kept in production when the technical founders move on. Most of those products don't have a known theory of operation, until I write one for them. The term "house of cards" is overheard in the cubicles when the products have to be maintained.

I've seen the innards of those products. I've built those products, for my own side business. Being at a big company gives me a perspective from both sides, about what those products really look like when someone other than the founder has to build and maintain them.

I'm not touting the big company experience as idyllic. Most creative people bristle at bureaucracy, and can see through processes that are unnecessarily constipated. On the other hand, I think the biggest adjustment is just getting used to someone else being able to say "no" to your idea, without having a really great reason that you agree with. Over and over again. Keeping your creativity under those circumstances is hard.

I think my biggest issue with the corporate bureaucracy is that they have a lot of self inflicted issues that they keep saying they want to fix (e.g. we want to be fast and agile like a startup), only to keep repeating the same damn mistakes all over again (e.g. lets write a 300 page spec document before we begin, and start this project with a 30man team, even though there’s work for only 3 at the moment).

I agree if you can handle ambiguity. If you cannot and need/want hand holding/marching orders you will not make it at a true early-stage startup.

If you need/want that you have no business being a software engineer. In my opinion at least.

> For most junior folks I meet, I recommend starting their career at a big company rather than a startup.

Yes, absolutely this. I'm surprised to see recommendations to do a startup as a first job. That's not a good idea.

A new grad who has never worked anywhere might know how to code quite well academically but they have zero exposure to how to build a commercial product. Semester-long projects at school can't teach that. That's how you end up with codebases without source control, no code reviews, no architecture, tons of security problems and so on.

Go work a few stints at large respected companies (in whatever field you are in). Absorb everything you can about how and why they do things. Understand the processes and why they exist. There's a reason for everything.

Then go do startups. You'll know how mature products are engineered, and you'll understand which parts of the processes make sense for a startup (not much, but you'll know) and you'll know how and when to slowly start introducing more maturity as the startup grows.

By now in my career I've worked at every size company, from 100K+ people multinationals to 3 person startups and just about every size and maturity point in between. I would never hire a zero-experience person to a tiny startup, that makes no sense. It would be bad for them and bad for the startup.

100% agree, and this isn’t specific to IT. Whether you want to open a grocery store or an automobile garage or an investment fund, starting your career with 5 or more years working for the big boys is really important. There are a lot of problems these big orgs have “solved”, a lot of knowledge to be acquired

Kind of feels like a local optimum kind of situation. If you do things the way the big boys do, you’ll forever be running after them.

Learning from doesn’t mean replicating. It’s like coding, learning from something big and industrialised is very useful , even if you’re going to code smaller apps

Note that those aren't the only two options. There are established companies that are small and have no intention of growing. Those are some of the best places to work.

I'd recommend starting at a company of 30-50. It's large enough so some processes are already there, yet small enough that your actions have impact, and you have influence on the processes. Software companies at this size have a pressure to grow, so you'll be there when processes are defined. There are specialists for some things you don't need to care about, but you meet them, and can talk to them about their roles and learn about them as well.

IMO, startups rarely have a process issue. They have a management issue. Process is bureaucracy. It adds no value but it helps to minimize the harms of over hiring which is something all big companies do in their quest for job creation.

I dunno. In my experience, it depends.

Like, $PREVIOUS_COMPANY was a small shop. Not many customers, not many users. Handful of customer product deployments.

Even so, releases/deploys were broken most of the time.

First thing I did after I got responsibility was add dev/staging with an automated and clearly documented release process from PR -> Prod, including steps for QA etc. Nothing stupidly onerous.

Customers complaints stopped almost overnight.

People internally knew what to expect when it came to changes being released. People could see when things were released and where etc. People knew why their work might not be propagated out yet. Higher ups knew why things might take longer and why that was beneficial. We could also break down some very simple stats about how deployments/releases were going based on git tags.

Applying the limitations of a process can be useful when things have been going wrong, repeatedly, in exactly the same way.

However… there were absolutely cases where adding a process didn’t make sense and would have just been adding a process for adding a process sake.

For example, ee didn’t need tickets for every PR. If it was a ten minute fix just do it and explain it in the PR, that’s good enough.

i think this depends. for most people, i would agree. but for people who were destined to be great founders, it's a terrible idea because working for big companies will inevitably condition you to do things a certain way. if you're someone who is smart enough and truly destined to be a great founder, starting from a blank slate and finding your own unique way of doing things is how we create new processes and systems and ideas that has never been done before. i don't think mark zuckerberg would have benefitted from working for big tech before starting fb.

I think it all depends on your manager, obviously. I’ve had good managers and mentors at startups. I wouldn’t have done it any other way.

I think you should be given the opportunity to try both, even better before you finish your studies (if that's the case).

Reply on news.ycombinator.com