Getting a Dev Job in 2025

I try to not make a ton of videos about getting a job in tech, but with the state of the market and how tough it has gotten, I think it's worth sharing my thoughts. I just had a conversation with a bunch of junior devs yesterday, and before then I did a talk at a college that really showed me how rough things are for people trying to get their first job, or even those who were laid off and are trying to get a new one. It's rough out there, and I want to do my best to try and showcase what, from my experience, works. I do want to disclose something important, though. I haven't looked for a job in a while. I obviously don't need to get one right now, thanks to all of you guys supporting me, my team, and the channel. I have two companies, one for my content and one for my actual tech stuff that we sell with Ping, both UploadThing and Ping.gg, and those have put me in a position where I don't have to hunt, but what I do often have to do is hire, fire, and help other companies doing the same. So while I might not have as recent of experience actually looking for a job, I still have a ton on the other side, and having done hundreds of interviews as an interviewer throughout my career, I hope I can give a little bit of good advice considering the current state of the market. I don't want to mince words, though. It is rough out there right now, and in previous videos that I covered the state of the job market in, I showed that it wasn't that bad, but at this point, it kind of is. I know far too many people who are struggling to get jobs, and at the same time, I see the same mistakes over and over and over again. I want to break all of these down and try to give you a mini-bootcamp on what I think makes the most hireable engineers, but first, a quick word from today's sponsor. PostHog, the all-in-one suite of product tools that you almost certainly should be using. Every project I didn't use PostHog for, I ended up regretting it. They're an open-source platform that's a little bit goofy, as we can see with the cookie request there. That isn't even a real cookie request. They have everything you need, all open-source, by the way, from product analytics to web analytics, session replay, feature flags, experiments, surveys, which surveys is so clutch. All of these are great. I mostly use the product analytics, but everything here is useful. It's one of the best analytics platforms I've used, super easy to integrate in every web stack, Next.js, Astro, crazy custom things, old-school stuff, too, but it doesn't matter how solid they are, how beautiful the website is if they're super expensive. Well, almost all of their users, more than 90% of their customers, are actually using it for free, not just because they can host it themselves with the open-source project, but because they give you a million events for free. Even when you do have to start paying, it's super, super cheap. We're talking 0.0005 cents. If you actually managed to rack up the bill here, congratulations. You found Product Market Fit. You're doing really well. I can't imagine doing all of the stuff I do today without PostHog. They made my life as a product developer way easier, which is why I reached out to them to sponsor the channel. So thankful they did. I hope this was helpful for you. Check them out today at soydev.link slash PostHog, and make sure to tell them that Theo sent you. To understand where the job market is now, it's important to understand where it was not long ago, because most of the education system and the advice that you're going to get online is focused on how the job market worked in, like, 2019, 2021. It also worked like this way before then, too. The key thing about tech jobs is that there were way more jobs than there were tech people. So if I needed five people on my team to work on a project, and I only had three, finding those other two was really difficult because the number of available engineers was relatively low. As such, we were more likely to open up junior roles, not because we needed somebody at a lower level or we just wanted to pay somebody less, but because it was so hard to find great engineers that you would roll the dice and hire people who were less likely to be meaningful contributors in hopes that eventually they would be. In this example, you need two really good engineers. Maybe you'll hire five to six juniors, and if you have a one-in-three chance of them becoming really good someday, you might end up with four people who are kind of dead weight, but you would also end up with two that are really, really good. And eventually, those people will be able to run teams and do their own hiring and level up other engineers as well. Hiring a junior engineer isn't an exchange of money for code. It's usually an investment, an investment in a bet the business is making in hopes that long-term, you'll become good enough to be a really useful contributor to that business. It's harsh to say it that way, but it is reality. Most junior engineers slow down their teams. However, if they level up fast enough and they really start to get what's going on, they stop slowing the team down and eventually can accelerate it really meaningfully. So the best thing to showcase in an interview will always be your ability to grow and adapt on the fly because that's the best indicator that somebody might break out of that junior level and go far, far past it. Two really big things have happened that have made the job market work less like this. The first is the massive layoff waves that have been plaguing tech for a while now. If you haven't kept up, it's pretty brutal. In 2024, over 145,000 tech roles disappeared. 2023, it was even more, 260,000. That's a lot of engineers who were employed that no longer are. And these people now are also looking for jobs. So if I'm again, imagine I'm an engineer running a team with five people, we need two more. Previously, I was very likely to consider juniors in hopes they could level up and get to that point. Now, there are 260,000 to 400,000 senior plus engineers that have actual industry experience that are competing for those slots. The likelihood that I can get an actual experienced engineer is significantly higher, especially if somebody on my team has worked with one of those people who was laid off. And they say, hey, my friend Joe here needs a job. I worked with him before. He's really good. They're going to skip the entire line and jump in front and be much more likely to get that role. There's a theme here, trust, and we're going to be going back to that in just a minute. Because first, I have to talk about problem two, AI. This isn't, oh, AI is going to ruin all our jobs, the big bad doom machine. Just trying to be realistic here. AI and the tools that are being built with it do represent a meaningful change in how jobs and tech roles work. On one hand, the need for less experienced engineers, as unproductive as they were, has gone down a bit. Because training Claude to do what you need is both slightly more reliable and slightly less higher cap than what junior engineers could do. And honestly, in a lot of ways, I feel like I'm mentoring a junior when I'm working with Claude. It's roughly the same thing. But more importantly, it has accelerated what good engineers can do very heavily. If I have the option of teaching an employee on my team how to work with a specific code base, or just spend 20 minutes trying to get Claude to make the change I'm working on, I'm going to do the thing with Claude most of the time. It's just easier and faster. And I, as an experienced engineer, can be much more productive when I use these tools and can cover a much larger surface area that previously would have required more employees. That's not the biggest thing that AI has changed, though. AI has ruined the process of interviewing for earlier roles. It is horrifying. If you put up an open role on the internet, like, hey, we're hiring for a junior, hit up this email address, you will immediately get thousands of emails from people AI generating the worst slop you've ever seen and calling it a resume, and then a five fucking paragraph or 10 page cover letter that they had chat GPT right for them. It is horrible. I cannot emphasize enough how miserable it is doing the actual recruiting and going through the never-ending large pile of resumes that are more and more full of AI crap. It's horrible. And I know it's hard to be sympathetic to the interviewer's side when you're trying to get that first role fresh out of school, fresh out of boot camp, fresh out of learning React yourself. But you need to know the misery that is opening up a junior role for applications and the gigantic, unbelievably large number of shit resumes that are going to come in as a result. It sucks. It sucks so hard that most companies just give up on managing that stack of resumes entirely because it's not worth going through because it's entirely slop. Just years ago, those roles would have gotten a few hundred applications. Now it's tens, if not hundreds of thousands. Sometimes I've seen people who had like 50 followers on Twitter post an open job role and in 24 hours get 10,000 applications. It's insane. So what do we do about this? If the number of senior employees looking for roles has gone up, the amount of shitty resumes has gone up, and the amount of roles as a whole has gone down, how can you possibly get a job? Is it time to just give up? I don't think so. Because there's one key thing that has been going down since all of this started. Trust. Trust is a hard-to-come-by commodity nowadays. Everybody is lying. Everybody's AI-generating and hallucinating shit. Every resume coming in is questionable at best. But you know what isn't? Your own experience and the people that you have worked with. The best thing you can possibly have when looking for a job is trust. And there's a lot of different ways to build trust. The one everyone says is go contribute to open source. I have a whole video called Don't Contribute to Open Source that describes why this is a dumb idea and also gives a little bit of advice on how to do this right. But I want to focus on some more practical things. One of the biggest ones is being part of the community. You have no idea how many people pop up in my Twitch chat, show up in my Twitter replies, hang out on Blue Sky, hang out in Discord, and just show without even trying that they know what they're doing. And it's very easy to start trusting these people. Almost too easy. If you know what you're doing enough to hang out and watch the videos that I make all the time, there's a good chance you could be successful in the space. A very good chance. And due to the fact that so many people are here, it might feel like you can't stand out. You are wrong. It is very easy to stand out by just being around and being useful. If you have a high signal-to-noise ratio where when you show a link in chat, it's likely to be relevant and interesting. Or when you reply to something on Twitter, it's likely to be useful or funny. I'll quickly build an association in my head with your username, your profile picture, brings value, is funny. These graphs in people's heads are relatively easy to manipulate, not in the derogatory trying to screw with people sense, but in the more realistic, people like the people that understand what they're doing sense. And it is so easy to build this type of trust if you just talk about the things you actually care about in spaces with other people who care about them. There is a catch though. If you don't actually care about tech, you're kind of fucked. A lot of people got into software dev purely for the money. They don't even like computers. I knew a bunch of engineers I used to work with that didn't use computers in their spare time at all. They would just use their phones a little bit and then read books and watch movies. And that's fine. Good for them. But in this new era of AI and the slow condensing of the field, the people who don't care are not going to make the cut. The ones who only came in for money that aren't spending their spare time watching videos on my channel, playing with new open source libraries and being part of these communities, they're absolutely fucked. And you have to go out of your way now to not look like one of them. And that sucks. It genuinely does because there's plenty of people who can contribute plenty of fine code. But if you don't care, it is unlikely you're going to be better than the AI tools that we'll be using next year. And that's a very hard pill to swallow. But once you've swallowed it, you'll realize, oh, I actually do kind of care about this. I know my peers don't, but I was the one they were asking for help with their homework on. Those people can still have a ton of success. And if you're one of those, if this isn't the first of my videos you've watched, if you've hung out in the community, you've played with the tools I recommend, you've disagreed with things I've said, you've been involved, you're fine. You just need to show the world that you do care and that you do know what you're doing. And the way to do that isn't spitting out hundreds of random GitHub repos. It's not throwing bad contributions to whatever projects you heard me talk about. The way to do that is being involved, hanging out, bringing value. Be realistic and honest. If there's something that I'm talking about that I missed a key point on, point it out. If I am live streaming myself talking about things that you're familiar with and I say, hey, I need a source for this thing. Does anyone have it? And you do, go find it and bring it in. Make your name associated with value and you will very quickly build trust with me. And I like to think I'm a good person to have trust you because I've helped dozens of engineers get jobs in the last year alone because I know a ton of companies that are hiring. But they're not hiring in the sense that they have a public listing that anybody can send a shitty AI generated resume to. They're hiring in the sense that they're strategically building a team that has specific technical needs and requirements. And they want to make sure every bet they make is a trustworthy bet because that's what hiring is. It is a type of gambling and it's the type of gambling that sucks because people's healthcare is involved. But you have to be ready to roll the dice a bit and understand that the other side is rolling the dice too. It's your goal to make that roll look as likely to be successful as possible. Increase the likelihood that you're going to be successful for those companies and the companies will see it the same way. You have to show them that you're very likely to be a good hire and very unlikely to be a bad one. And that's not as hard as it sounds. There's a lot of ways you can do this. I met this engineer, Taylor, at RenderATL and she was awesome. She asked a really good question after my talk about a niche use case for server actions in Next.js. And I decided to chat with her more after because I was curious. She wanted to make sure she was using the right solutions with server actions. So she went out of her way to try a whole of these different options and came to me asking what my thoughts were about them. My thoughts were, well, I haven't done the research you have. I would love for you to share that research so I can do more with it. And I told her it would be awesome if she had done a blog post about it. Her response? Other people care? Yes. If you care enough to do that type of research, to spend that time deep diving on different solutions to problems, document it and share it with the world. There's a decent chance nobody notices. There's also a decent chance someone does. And more importantly, if you're talking with people like me and I go to your profile on Blue Sky or Twitter or whatever and I go to your blog and I see you have this post that's really interesting to me, light bulbs start going off in my head like, oh, they get it. They're in. They're not just playing with these things to get a job. They care enough to deep dive on things I haven't even had the time to deep dive on yet. And that is the best signal I can imagine that someone is a trustworthy, potentially good hire. And there's one last piece that will always make hiring easier. Friendships. If you have connections with other people, be it peers that you went to school with, co-workers you had in previous jobs, or just a friend from your hometown that happens to have a role in an engineering company, try to maintain those relationships and try to nerd out and talk about these things with them. Obviously, if those people aren't actually nerds, they're just coding for the job and making the money, don't push it too hard. But if you can form a solid memory in their heads, a, oh shit, this person is associated with this thing going well, the likelihood that you can turn that into a job is very, very high. I learned this personally from my time at Twitch, because as rough around the edges as I could be to work with, I'm not one to compromise where it would hurt our users. I was very deeply associated with things actually shipping. I became the guy that was pulled in when projects were going to miss their deadlines to make them not only hit their deadlines, but often exceed them. So we could keep working on the things that actually mattered. And that association between me and shipping has persisted throughout my entire career. And I had a ton of opportunities come up and I still do to this day, because somebody I worked with in the past has that association between me and actually getting shit done. Those associations last way longer than any role at any company. And the people who built those associations are going to have their own jobs in the future at other places you might be interested in. That is so essential to succeeding. Because when I'm hiring, I'm kind of doing like a tier list in my head of what I want. If I have an open role on my team, the first thing I do is think about the people I know that'd be a good fit for that role. Once I've exhausted that list, if I can't find anybody, the next thing I do is mention it to the team. I'll say, hey, I kind of want to hire for this role. What do you guys think? And if somebody else says, I think that's a good idea, I actually know a person who would be perfect for it. They just got laid off at this company, and I think they'd be a good fit. That's the easiest interview in the world. They're probably going to get the job. If after that point, there's nobody on the immediate team that has a good referral for the role, maybe I'll post it as an internal job posting to see if anybody from a different team wants to transfer to my team. Because if we've already hired them, and I can look at the work they've done and talk to their coworkers, the likelihood of a bad hire is significantly lower. And internal poaching is at an all-time high right now, because it's so scary to hire someone that isn't a good pick. After all of those steps are exhausted, after I realized I don't know anybody, my team has realized they don't know anybody, and I can't poach anyone from another team at the company, finally, I might consider a traditional public job role listing. Maybe. So if you do want to get a job fast, it's very important to do what you can to be in those first three tiers. And if you're stuck fighting at the fourth one, if you don't have friends in tech, if you're doing this alone, find a way to be less lonely. Because that's going to be the thing that kills your career. It's not the opportunities that businesses have, it's not the companies that are not giving you offers, it's not the degree that you did or didn't spend all your money on. The thing that will kill your career is letting the loneliness get to you. Fight that. We're getting some good messages, and I wanted to call out this one, because I think it's a phenomenal example of what I'm talking about here. I had a video call with a guy I didn't know from Turkey to help him with a question that he posted in Theo's Discord about how to set up CreateT3 app in a certain way. Now he's a good friend of mine. This is what I'm talking about. Ayub here had successfully made a connection, brought value, and built this mental association with, in that case, one person. But that could scale really far really fast. So now if that person from Turkey that he met is hiring, or is working at a company that is hiring, and he knows that you're looking for a role, you are so much more likely to snag that role. And if you do this for not just one person, maybe three or four, maybe a few of those are public, and if it's in the Discord and the public questions stuff, the answers are there too. And I and many others notice who is answering those questions and who is bringing the most value. Some of the best hiring help I have given companies was telling them who in my Discord answers questions the best, and they just straight up hired people from that. And now those people are running engineering at surprisingly large companies, like Clark. Yes, Clark has a high up engineer that was found because they answered questions while in my Discord. These are some of the best ways to build trust, to have these relationships with people who get it and are in the field too. Oh look, my team. I literally met Igor through Twitch chat. Gabriel is my lead researcher. I found him because he was always posting useful links and sources and all the other things I needed to be successful with my video creation and pulled him onto the team. We also have Addy here who has been beyond helpful. She's found a ton of connections and made a ton of friendships in the space, and now she's the one who reviews most of our videos before they go live. There's a ton of these, but like Ethan met Reese through the server. If you watched my video on the faster version of McMaster, those two became friends through my Discord server and they built that together. They might even do a startup in the future knowing how these things tend to go. It's not that hard to do. When you're a nerd and you hang out in spaces with other nerds, take advantage of that. Don't be alone. Don't let yourself be the loner that doesn't have friends to talk about this stuff with, and don't use the fact that you went to school during COVID as an excuse, which I've seen a lot. It does suck, but you can still make these connections other ways. I made a lot of friends through weird Facebook groups about hacking stuff when I was in college, and those friendships ended up being essential to me finding success in the field. Find people that you can relate with, that you can nerd out with, that you can have these deep technical and life conversations with, so that you don't have to go through this shit alone. Because if you're going through it alone, I honestly don't know if you're going to make it now, because this shit kind of sucks. So make friends, build trust, show the world that you kind of know what you're talking about, and you won't have as bad of a time. And for everyone else, if you're just here for the money, good luck. That's all I got to say on this one, I hope this was helpful. Until next time, peace nerds.