My next, next, next job

I originally wrote this as a private journal entry, then pushed it to a private (now public) GitHub repo to timestamp it for myself. I shared it with a few friends who suggested I go ahead and make it public.

Not too long ago I came across the concept of your Next Next Job. The question is:

“What do you want to be your next next job? And why can’t you get it right now?”

This has been a pretty useful framework for me over the past few months, which is about when I learned that I might not get a full remote work exception to move from Brooklyn to Cincinnati around the start of next year (i.e. when I realized I might have to find a new job pretty soon).

So anyhoo, it’s October 2023, and the reason I’m writing this in a private GitHub repo instead of my private notes is so that I can timestamp it and perhaps even make it public n years from now. One obligatory thing to mention is that plans obviously and inevitably change, so this is not a roadmap I expect to follow but rather a framework that hopefully helps me make career decisions.

As I was thinking about this, my "next, next, next job" came to mind as a natural extension of the prompt. Here’s a table of contents, of sorts:

• My next job
• My next, next job
• My next, next, next job
• Interlude
• Investing my time
• TODO
• What this isn’t

My next job

The focus here is much more on my next, next job (and the one after that), but my immediate job search does inform what follows. So here’s what I’ll be looking for:

• ~10-50 person startup
• Devtools company
• Talented eng team that I can learn from
• Frontend Web focus with fullstack work mixed in
• Building and/or writing in public (OSS, maybe)
• Remote

None of those parameters are set in stone, necessarily. Like, I could join a larger company if it checked other boxes, e.g. Vercel because it’s a devtools company with a ton of awesome open source under its umbrella, plus adding the ▲ to my resume could help me land my next, next job. Or I would definitely welcome less frontend/JavaScript work for a talented eng team and product I really like, e.g. joining Warp and learning Rust+WebGL.

Joining a devtools company is attractive to me for a few reasons. One, I use devtools every day and find great satisfaction in really good ones. Two, creating products for a technical audience (developers) means the quality bar is high, which means it’s hard technical work, which means I’d probably learn a lot and my coworkers would probably know a lot about web dev stuff. Three, devtools work seems to facilitate learning in public more than other kinds of work (like the B2C companies I’ve worked at as an eng thus far), which means my code (e.g. open-source devtool) and writing (e.g. company blog posts) and speaking (e.g. demos, talks, podcasts) would serve to build my public reputation as I build my skills/knowledge.

A more general way of stating the goal for my next job is to move my weekday work closer to my weekend work. I.e. my job could be more similar to the side projects I choose to work on over weekends and on weeknights. It seems logical that I should aim to spend more time doing what I naturally gravitate toward.

My next, next job

As of fall 2023 (I’m writing this on October 27th), my ideal next next job is founding engineer at Val Town. Here’s a screenshot, in case they take it down:

Founding Engineer at Val Town

Obviously that exact job will get filled, and the Val Town team will likely be a lot larger by the time I find myself two jobs down the road (side note: I’m super excited about VT and am rooting for them to realize the dream of vals as a Web primitive). So founding eng at VT is just a proxy for me. The job itself is exactly what I want, though: love the product/vision, lots of coding, building with Web platform primitives, working on a devtool, talking to developers, working in public, etc.

I want this line in the job description to describe me within the next year or two:

You...have had some sort of success shipping something on your own — a blog, open source project, etc.

As of recently, I am in the habit of writing stuff here on petemillspaugh.com, thanks in large part to discovering digital gardening. And even though I’m in the very early stages of my learn-in-public writing habit, it already feels very fun and rewarding. In the past week I wrestled with getting Pagefind search working in my garden, and the project maintainer bglw offered to add my write-up to the community resources docs page. That felt good!

Back to Val Town. Under the “Your skills and experiences” section:

• JavaScript, TypeScript, and bonus points for Rust
• Building devtools...
• Technical writing...
• Bonus points for experience with bundlers, compilers, transpilers, or sandboxing

All of those are either already in my wheelhouse (JS/TS, writing) or are things I’d like to add to my wheelhouse (Rust, bundlers, etc... What even is a wheelhouse, anyway? brb...ok, it’s the part of a boat or ship where the steering wheel is located.)

The job is also in-person in Brooklyn, which rules me out (ironically because I do live in Bkn presently but the whole point is that I’m moving soon). So (a) I’m not qualified yet, and (b) I need a remote job. But this exercise is very instructive in setting a concrete goal for role I want to be qualified and prepared for in what, 3-5 years? Or maybe I could literally come work for Val Town in a few years!

My next, next, next job

So if founding engineer at Val Town is a proxy for my next, next job, what comes after that?

This one is a bit fuzzier, obviously, but it’s fun to brainstorm. As of now, I am penciling in Independent Coding Course Creator as my next, next, next job.

The example that comes to mind is Josh Comeau, whose courses have been both useful and inspiring to me. It seems like Josh draws a lot of joy from teaching people, building his bespoke course platform, and maintaining a community.

I like teaching people—I think being able to explain things well is among the most important human skills—although I don’t have much experience teaching web dev, yet. I love building front-of-the-frontend stuff, so a course platform is a compelling opportunity. Creating courses would involve a lot of writing—major plus. It might also pair nicely with conference speaking, which I think would be a very effective learning tool much in the same way writing and teaching are (i.e. you better know the thing well if you’re going to talk to a crowd of people about it). I’d also get to work on what I want / think is cool (but driven by the incentive to build something useful, ofc). So what will make me qualified for that?

• Knowing a lot of stuff, duh
• Long-running blog digital garden with readership
• Track record at known startup(s) because you need credibility to sell
• Teaching at a bootcamp and/or tutoring kids learning to code
• Work on open source and side projects

The economics of this path present much tougher prospects than most other next, next, next jobs I could have reasonably slotted here (e.g. Staff Engineer or EM). For some napkin math, if I can spend about a year making a course that 1,000 people would buy for $100, that’s $100k annual pre-tax income, which is more than enough. There are other ways to monetize, too, but selling a course for money is the simplest to reason about. Convincing 1,000 people to pay $100 for a thing (or 10,000 people at $10) is no small ask, though. Plus, it’s not sustainable—it’s unreasonable to expect to crank out a new course each year like that. The education stipend at big tech companies definitely changed the way I think about this ($5k annually at Credit Karma 🤯), but developer education is a fairly saturated space already, and the quality bar will only trend higher. But it’s fun to dream, and if I did go for it as an independent creator and hit the best-case scenario (JC/KCD level success), I could be financially independent, which would unlock a new incentive structure for what I work on.

Interlude

After drafting my next three jobs, I came back to add one career phase—an interlude—that I’m increasingly interested in and could be slotted between any job transition: Recurse Center. Recurse is an open-ended six or 12 week block of time and space with the simple goal of pushing yourself as a programmer. I first learned about it through Julia Evans (who did it twice!), then I started noticing other programmers I admire who have done a stint there. Like Jake Lazaroff (whose personal blog is 🤌) and Justin Bennett (co-host of devtools.fm, one of my two must-listen web dev podcasts). I’m sure I’ll keep discovering others who have done Recurse. I should ask them about it (update: I did talk to Justin who had positive things to say). I’m not sure when it would make the most sense for me to do it or what I’d even focus on (maybe offline-first web apps/PWAs), but I love the idea of a blank canvas and time to recalibrate / figure out what I want to work on.

Investing my time

Back to the present. With all this in mind, how should I allocate my time?

Heather Buchel’s article about what being a web designer used to mean got me thinking: if I was starting my coding career 10-20 years ago instead of 2-3 years ago, that’s totally what I would have been. A web designer. Or at least, that job path would make a lot of sense for me. Intimate with both design and frontend code. And I have absolutely heard the common opinion she referenced that front-of-the-frontend stuff isn’t "real engineering" and isn’t as worth my time. I’ve even convinced myself of it at times. But now that I’m more competent and confident in my coding foundations, I can lean into the front-of-the-frontend with more conviction. I think it’s completely reasonable to build my depth there. That’s what I’m best at, I think. I can spend a big chunk of time on Josh Comeau’s CSS course, because that knowledge will probably be fairly timeless and worthwhile—yes, even with AI. Just like how SQL will be timeless and worthwhile, which I bet most people would agree with—see: Lindy effect—even though language models are amazing at SQL, too. I do still strongly value fullstack competence, so I want to continue building expertise over time on server-side and database stuff, but I can also feel good about investing my time on f-o-t-f things like petemillspaugh.com.

TODO

If I were to follow Andrew Chen’s advice (the guy who wrote about the Next Next Job), I should be talking to people who already have my next, next job. Luckily for me, Rodrigo (my current CTO) and Aneury (my current manager) had jobs roughly similar to the next, next job I laid out above as a technical cofounder (Rod) and engineering hire #2 (Aneury) at SeedFi. So I can talk to them about it (update: I did!). I should probably talk to other people, too. Maybe I can try to find someone who matches one dimension of my next, next job really well. Like, one person who likes to write a lot, and another person who identifies as very front-of-the-frontendy, another person who has worked in devtools for a while, another person who joined as engineer #1/2/3, etc. Tying this back to my next, next job, I should totally go meet the Val Town team while I live in Brooklyn (another update: I did!). Maybe I could even meet the person who they end up hiring for that founding engineer role. Taking this networking TODO one step further, an even more impactful lever would actually be working for/with someone like that in my next job. E.g. a mentor or mentors who are early engineers at a devtools company and have f-o-t-f depth and/or like to write or learn in public or whatever.

FWIW, Andrew Chen also talks about building a "superpower" that will get you your Next Next Job. For me, I feel pretty strongly that my superpower can be writing in public. Not because I’m an exceptional writer or anything, either. I love to write because it helps me think and learn, but the reason it’s a superpower is because not that many developers actually do it frequently and consistently. So if I keep up consistent writing in public then I can be an open book and therefore a high-confidence hire for the right role as an early engineer somewhere.

What this isn’t

As I said up top, this is a thought exercise—not a plan that I should follow rigidly. Useful nonetheless, I hope. I’m adding this concluding section in December ’23 because I just read this relevant Richard Hamming quote from The Art of Doing Science and Engineering, which I’ll end on:

One of the main tasks of this course is to start you on the path of creating in some detail your vision of your future...

You will probably object that if you try to get a vision now it is likely to be wrong—and my reply is that from observation I have seen the accuracy of the vision matters less than you might suppose, getting anywhere is better than drifting, there are potentially many paths…

You must, as in the case of forging your personal style, find your vision of your future career, and then follow it as best you can.