# Fifty Bits of Career Advice


As my team's summer interns finished up their rotations this week, I had my
usual end-of-internship "AMA" 1:1s. It's something I enjoy doing, but I realized
I was covering a lot of the same topics I've discussed with other previous
interns and early-career engineers over the years.

So I thought it'd be useful to collect these recurring bits of advice into a
single place, both for my own reference and for anyone else who might find them
helpful.

### Decisions and opportunities

1. There is a natural "explore vs. exploit" dynamic in career decisions. Early
   in your career, you're well served by leaning more into the "explore"
   direction: do as many internships as you have time for, try living in
   different cities, try working on different teams, try working on
   product-focused projects, try working on infrastructure projects.

2. Surprisingly few career decisions are true one-way-doors. This generalizes
   outside your career as well.

3. That said, all decisions are path dependent. The two-way door you step back
   through won't be the same one you left, in the same way that
   [you can never step in the same river twice](https://en.wikipedia.org/wiki/Heraclitus).
4. It's helpful to gravitate towards where value accrues in an organization.
   This is often not intuitive.
5. Similarly, it's helpful to gravitate towards where the smartest people are
   working. Cultivate a set of formal and informal mentors.
6. Similarly, it's helpful to gravitate towards where there are obvious problems
   that are low-hanging fruit to solve. For engineers: Environments where you
   are engineer-hour-constrained instead of product-alignment-constrained are
   much easier to progress quickly in. For managers and PMs, the reverse is
   likely true.
7. It's worth figuring out your tolerance for risk fairly early, since this has
   implications for many things: timing of job hops, personal finance posture,
   whether you prefer the (appearance of) safety of a large company or the risk
   associated with doing a startup.
8. I find I rarely regret my choice after choosing between two paths with
   reasonably similar "goodness", even though these decisions are often the most
   agonizing. After all, if the options weren't similar in "goodness", it'd be
   an easy decision.
9. Making hard decisions doesn't get easier, you just make more of them over
   time and become more comfortable with the discomfort. You also eventually
   build up your own
   [personal heuristics for tough decisions](/blog/2022/05/18/Tools-for-Making-Difficult-Decisions/).
10. Careers are marked in seasons. When you feel a season change coming, it can
    be tempting to hold out to avoid the discomfort of change. Sometimes you
    initiate the season change (e.g. you seek out an exciting opportunity) or
    sometimes the season change finds you (e.g. "oops, a few key people left and
    now your org is in chaos"). Either way, you'll spend less energy preparing
    for and adapting to the new season than you will trying to preserve the old
    one.
11. When receiving advice from others, note that sometimes you need to
    [reverse the direction of the advice](https://slatestarcodex.com/2014/03/24/should-you-reverse-any-advice-you-hear/).
12. When finding your first job or two, it's easy to look at interviews as a way
    of "convincing the company to hire you". However: the phrase "interviews are
    a two-way street" is a cliché for a reason. Interviews give you an important
    (though necessarily incomplete) window of what the company is working on and
    looking for, and you should try to maximize the amount of signal you get for
    yourself out of interviews. Prepare questions in advance, ask them, and
    write the answers down. This can help change your eventual decision to be
    more informed and less purely vibes-based.

### Organizations, leadership, and culture

13. Truly great managers are hard to find. Value the good ones.

14. Truly healthy organizations are hard to find. Value the healthy ones.
15. It's tempting to become "irreplaceable". It's far better to foster your
    replacements. (Both for your own sake and the people that grow to replace
    you)
16. The average tenure of an engineer at most big companies is no more than ~3
    years. An implication of this is that you _will_ see turnover on whatever
    team you join. Usually the sky is not falling even if a few folks leave
    around the same time. If you have the stomach for it, stepping into a
    departing senior engineer's shoes is a great growth opportunity.
17. At a steady state, 80% of your time when being a good tech lead is being
    [the rock that says "That seems reasonable"](https://www.astralcodexten.com/p/heuristics-that-almost-always-work).
    About half of your value comes from (skillfully) being this rock, and the
    other half comes from recognizing when it's necessary to push back, and in
    (skillfully) following through with doing so.
18. Many things are overrated. "Being in the room where the decisions are made"
    is, if anything, underrated.
19. Organizations are like people: They have personalities, they have
    insecurities, they have strengths and weaknesses and hopes and fears. Being
    in tune with organizational psychology is a great marker of a senior
    engineer. Influencing organizational psychology is a marker of staff+.
20. How teams and individuals say "goodbye" to people departing the company
    tells you a lot.
21. Promotions aren't as much an indication of "job well done" as they are
    recognizing that you're doing a qualitatively different job description as
    the last time you were re-leveled.
22. The "good feeling" of getting a promotion lasts much shorter than you'd
    think (for me, about a day or two). They're still worth pursuing and
    achieving as career milestones. But they're progress markers, not ends in
    themselves.
23. Kindness goes a long way. In 5 years, the people you remember will be the
    ones that helped you onboard, or bailed you out in a stressful oncall
    rotation. You'll spend comparatively less time thinking of the people who
    were "merely" brilliant. (Thankfully, brilliance and kindness co-occur
    suspiciously often)
24. You are _not_ the ultimate owner of your performance review. You can, and
    very much should, advocate for yourself and put a substantial amount of
    effort into writing the pieces that you control. However, it's ultimately
    your manager's job to advocate for you on your behalf. If you have a good
    manager, this is easy. If you don't, you'll need to do a lot of the
    narrative setting yourself. In either case, managers are invariably
    incredibly busy during perf season, so proactively making it easier for them
    to advocate for you helps you both out.
25. Most SWE career ladders value "cross team collaboration" for senior and
    above levels. As such, it's helpful to develop relationships with people
    outside of your immediate team. This is true for multiple reasons: first,
    it's a great way to get a larger sense of what the company is working on,
    and second, it gives you a natural set of people to act as champions for
    your work when you are in performance review season.

### Execution and craftsmanship

26. Unlike all the code you wrote before you started your career, you can't take
    most of the code you write professionally along with you (unless you're
    doing professional open source work, which, if you do, cool!). This is a
    good thing, since you get to
    [build things that outlast you](/blog/2025/04/08/Why-Developer-Tools/#:~:text=this%20developer%20had%20a%20quite%20positive%20view%20of%20it).

27. When you're new on a team, you automatically have a valuable perspective: a
    set of fresh eyes. Keep track of the papercuts and unintuitive things you
    run into -- when onboarding, when learning about how your team's systems
    work, and so on -- and share them with your team. Many of these things will
    already be known by them, but people develop an indifference towards
    papercuts over time, and having a fresh set of eyes is a great way to
    surface (and potentially fix) them.
28. Early in your career, there's a lot of value in being the person who lets
    nothing fall through the cracks. A few years into your career, it's about
    choosing which cracks can be safely ignored.
29. Relatedly, there will always be work that _never_ gets done. Even on the
    most relaxed teams, there will always be an ever-growing task backlog of P1
    and P2 tasks. This is fine. It's generally a waste of time to hunt down
    every little TODO on the long tail of your team's backlog.
30. Relatedly, it's sometimes worth declaring bankruptcy on your team's backlog.
    Unless you're a TL or manager you generally don't get to make this decision,
    but even as an IC it's worth keeping a pulse-check on how much backlog debt
    your team has at any given moment, and doing your part in managing this.
31. Having passionate users is a great way to get high quality feedback. Fixing
    a papercut or two of a power user can earn you a years-long source of
    feedback.
32. That being said, it's worth remembering that the most passionate users are
    not representative of the majority of users, and that you will still need to
    actively solicit feedback.
33. _Actually use_ the tools/products/systems/libraries/etc. you build. This is
    a necessary but insufficient condition for building something that is
    useful.
34. You should probably invest a bit in your physical workstation. A good
    keyboard, mouse, and monitor help both in feeling a sense of ownership over
    your workstation, and help with ergonomics.
35. You should probably invest a bit in your digital workstation. Yes, you
    should pick a nice color scheme for your IDE. But you should also invest in
    hyper-personalized configurations that make you faster at your job. Things
    like terminal aliases for jumping between PRs/branches, small scripts for
    common tasks, etc. These investments compound over time.
36. It's worth learning a pure command-line text editor like vim or emacs. You
    don't need to be an expert in it, but (for the foreseeable future) there
    will be times when you don't have access to an IDE, like when SSHing into
    remote machines, and you'll be glad to have a comfortable fallback.

### Self-management & mindset

37. It's rarely worth making a decision you see as ethically or morally
    compromising. Being able to live (easily! joyfully!) with your decisions is
    a gift worth giving to yourself. To make this slightly less
    [Straussian](https://marginalrevolution.com/marginalrevolution/2023/01/ask-the-beast.html):
    There is often money out there that is not worth taking. But you have to
    figure out where to draw your own line.

38. If you're a natural "optimizer", you should probably find a way to mindfully
    and purposefully switch to a "satisficing" mindset when it's helpful.
39. It will sometimes feel like
    ["Everyone Is Getting Hilariously Rich and You're Not"](https://archive.is/3CZUA).
    You'll see former colleagues, classmates, and acquaintances win big in AI
    startups, crypto, or \$CURRENT_THING. Bubbles and the "leading edge of the
    hype cycle" always create real winners alongside the noise, but the winners
    are fewer and more random than they appear. Reminding yourself this is FOMO
    will save you a lot of grief.
40. That "I don't know what I'm doing" feeling early in your career is usually
    followed shortly after by "I'm starting to get my footing!" Believe it or
    not, you eventually start to crave this "beginner mindset". The senior
    engineers who you see switching teams or leaving for another company are
    often going to _feel that way again_.
41. It's worth cultivating an online presence, even as an extremely slow burn
    side project. Having a body of work online, especially writing you sink some
    good thought into, is portable across companies, contexts, and even career
    paths.
42. You are the ultimate owner of your career. The best managers and TLs enable
    you to move in the direction you want to go, but you are the one that has to
    set that direction.
43. You are the ultimate owner of your mental and physical health. You will need
    to learn your early warning signs for burnout and take steps to mitigate
    them. Take time off proactively rather than as a last resort.

### Finances

44. For short term financial planning, you should _mostly_ treat RSUs as
    worthless paper until they vest. For medium term financial planning, I find
    it helpful to do a
    [Monte Carlo simulation](https://en.wikipedia.org/wiki/Monte_Carlo_method)
    of various scenarios I think my current company is headed toward. There's
    some skill involved here, but it's worth the effort to get a sense for how
    to value your RSUs. This used to be an afternoon of work, but thankfully
    with LLMs you can script something like this in a few minutes.

45. By default you should _probably_ sell your RSUs as soon as you can and
    reinvest in a more diversified portfolio.

46. That being said, if you're conservative with finances, you're more likely to
    undervalue the future appreciation of RSUs (and thus, when comparing offers
    you may undervalue the net present value of RSUs as well).
47. You should be able to comfortably live off your salary (without any bonuses
    or RSUs).
48. If your company does 401k matching, you should absolutely max out the match.
    You should _probably_ max out your 401k contributions (when possible) as
    well, as a set-and-forget way to do retirement planning until you have time
    to develop a more nuanced plan for handling your finances.
49. It's worth doing a deep dive on _exactly_ how the tax system works.
    Internalizing how capital gains and each type of retirement account are
    taxed is a great way to avoid future unpleasant surprises.
50. If you work in tech, there's ~no excuse for not keeping and fiercely
    guarding an emergency fund of 6-12 months of expenses.

Your mileage will vary on all of this. What matters most is developing your own
heuristics over time. These are just mine, offered as a starting point. In a few
years, you'll have your own version of this list, shaped by different choices
and different seasons.

Cheers!