There is a colleague I worked with for about three years, let’s call her Sara, who is by some distance the most quietly impressive engineer I’ve worked with. Sara is not the smartest person on the team. She would tell you that herself. She is not the loudest in design reviews, not the one who pulls all-nighters. What Sara does, and has done for as long as I’ve known her, is ship something every Friday. Not always something big. Some weeks it’s a feature. Some weeks it’s a postmortem written cleanly, a runbook updated, or three small bug fixes she’d been meaning to get to. Whatever it is, Friday afternoon, it lands.
Over those three years, Sara was promoted twice. She passed engineers who, on any given week, looked flashier than her, who could win design arguments she couldn’t, who knew more about distributed systems. She passed them quietly, and over time the picture had become obvious: Sara was the person you could count on, and the others were the people you crossed your fingers about.
Most career advice is dramatically overweighted on the heroic and underweighted on the boring. The boring stuff is what actually moves your career. Showing up. Doing the work. Finishing what you start. Compounding, week by week, over years.
”I’ll be impressive next month” versus “I’ll do my job well this Tuesday”
There is a story almost everyone tells themselves in the first five years of their career: “I’m not really showing what I can do right now, but next quarter, when the X project starts, that’s where I’m going to make my impression”. Then next quarter X gets descoped, or you get pulled onto Y, and the goalposts move.
I have been guilty of this. The story is seductive because it lets you tolerate this Tuesday’s mediocre output by promising yourself that next Tuesday you’ll bring it. The version of you who is bringing it next Tuesday turns out to be the same person, with the same habits, who didn’t bring it this Tuesday.
The shift that took me longer than I’d like to admit is from “I’ll be impressive next month” to “I’ll do my job well this Tuesday”. It is less satisfying as a self-image and far more useful as a habit. You don’t get to fantasise about the heroic launch where everyone applauds. You just get to be slightly better than yesterday at the work in front of you. Reviewed your colleague’s PR thoroughly instead of skimming. Wrote the test you said you’d write. Wrote that one comment in the runbook that future-you will need at 3am during an incident.
None of that wins you any awards. Over a year, it makes you a different engineer.
The trust-debt bank
The model that helped me make sense of this is what I think of as the trust-debt bank. Every time you reliably ship a thing you said you’d ship, you make a deposit. Every time you say you’ll do something and it doesn’t land, or it lands two weeks late without comment, you make a withdrawal. You don’t usually see the balance. Your manager, your peers, and the people who decide who to put on which project all see it.
After a year or two of consistent deposits, your balance is high. People assign important things to you, not because you’re the best engineer in the room, but because the variance on your output is low. In planning, your estimate gets believed. In incidents, you get paged because they trust you’ll handle it. In promotion discussions, your name comes up first, because nobody can think of a recent example of you dropping the ball.
After a year or two of withdrawals, you have the opposite problem. The brilliant launch you led six months ago doesn’t cancel out the three projects you were supposed to follow up on and didn’t. People don’t say so. They just route around you. You feel under-trusted and can’t tell why, because no single thing you did was bad. The pattern was bad.
You cannot make yourself look reliable in one heroic month. You can spend a year being reliable and then look up to find you’ve quietly become the person people send the important things to.
Reputations are a thousand quiet days, not three loud ones
Reputations are built in the unobserved hours. We imagine reputation as something forged in big moments: the brilliant talk, the launch, the complex bug heroically fixed at 2am. Those matter, but they’re a small fraction of the signal. The larger fraction is the steady backdrop: did you respond to the PR within a day, or did it sit for a week? Did you write the doc, or did you say you would and then quietly not? When on-call lands on you, do you answer pages, or negotiate them away?
These are small individually. They are not small in aggregate. The colleague who routinely takes on-call is seen, every week, by every engineer who didn’t get paged because she did. Nobody says “thanks for being on-call”. Everybody quietly files the data point. By the time anyone writes a performance review, your reputation is already hardened, formed by hundreds of small behaviours nobody consciously noted.
The visible launches are the icing. The reputation is the cake, and the cake is baked over thousands of Tuesdays.
Doing the unsexy work on purpose
A counterintuitive consequence: the unsexy work is some of the highest-leverage work you can do.
The unsexy work is the work nobody is going to applaud for. Filling out the postmortem properly instead of writing two lines and moving on. Updating the runbook after a deploy went sideways. Writing a clean rollback script instead of a quick one-shot. Reviewing the new hire’s first three PRs in detail. Documenting the dbt model with a paragraph instead of a one-liner.
These tasks have low individual visibility and high accumulated value. The thorough postmortem stops the same incident happening in six months. The runbook saves a colleague forty-five minutes at 3am. The new hire you walked through conventions is, six months later, a productive engineer instead of a stuck one. The dbt docstring is a rounding error today and load-bearing knowledge eighteen months later when you’ve forgotten the original context.
Engineers who do this work consistently are, over time, recognised for it. People notice that incidents involving your systems get cleaner postmortems, that your runbooks are usable, that you can point them at a doc instead of explaining things again. You become the person who keeps things in working order, which is a higher-status position than it sounds.
Why the people who reward heroics lose their best engineers
A short word on companies, because the consistency argument has a flip side. Some organisations reward heroics over consistency without realising it. They promote the engineer who pulled the all-nighter to fix the launch. They write Slack shoutouts for the person who saved the demo, but never for the person whose system never broke. They give bonuses for “exceptional effort” that was needed only because someone else’s quiet diligence had been allowed to slip.
This is, in my experience, an early sign that an organisation is going to lose its best engineers. The people who run consistent C-plus weeks for years notice, eventually, that their work is invisible relative to the firefighters. They see the firefighter celebrated for putting out a fire the consistent engineer prevented from ever starting, and the celebration is louder than the prevention ever was. They start looking elsewhere.
If you’re an engineer, pay attention to which kinds of work your organisation actually rewards, not which it claims to. If they keep promoting the heroes, you are in a heroics culture, and either you adapt to it or you find a place where consistency reads as the asset it is.
If you manage engineers, you’ll get what you pay for. Reward heroics and you’ll get more fires. Reward consistency and you’ll get more stability. The second is what the company actually needs, but it is harder to see, because by definition the contribution looks like nothing happening.
The compound interest connection
The shape of this idea is identical to compound interest, which I write about a lot in the finance posts on this site. Small, regular inputs over a long horizon beat sporadic large inputs, regardless of size. An A-plus week followed by three D weeks loses to four straight C-plus weeks, and it is not close once you compound them across years.
In money, the boring index-fund investor with twenty years of consistent contributions ends up ahead of the active trader who had one phenomenal year. In careers, the engineer who ships every Friday for three years ends up ahead of the one who shipped a brilliant project once, then disappeared. People overestimate the spike and underestimate the slope.
The good news is that consistency is more controllable than brilliance. You don’t get to choose how smart you are. You get to choose, every Tuesday, whether you do your job well today. You don’t need a great quarter to be on a good trajectory. You need a hundred unremarkable Tuesdays.
Takeaway
The seductive career story is the heroic one: the launch, the all-nighter, the brilliant intervention. The boring one is the true one: showing up on Tuesday, replying to the PR, writing the postmortem properly, taking the on-call. The seductive story gets you a Slack shout-out. The boring one gets you a career. Reputations compound the way money does, in small reliable increments, over years. The engineers who understand this stop trying to be impressive and start trying to be the person you can count on. That is unsexy advice, and after fifteen years of watching the same patterns play out, the only career advice I’m completely sure of.