Fill the Bus, Then Pick the Road: Notes on Managing People
A practical guide to managing people and a complete breakdown of the counterintuitive rules that actually make teams work
There’s an entire industry built around telling you how to manage people, and it exists for a good reason: this is genuinely hard. Walk into almost any organization and you’ll find the same problems showing up – work slipping, good people running low on energy, the same issues circling back no matter how much effort goes in. It isn’t because the people in charge are careless; most of them are sharp and working hard. Managing other human beings is just one of those skills that doesn’t come naturally to anyone, and the gap between knowing what to do and actually doing it under pressure is wide.
A lot of the trouble comes from instincts that feel completely right in the moment. The project’s late, so we add people. Things feel chaotic, so we add process. A decision is risky, so we wait for everyone to agree. Each of these is perfectly reasonable, and each one tends to make things worse – which is exactly why they’re so easy to reach for. A surprising amount of good management turns out to be resisting the obvious move.
None of what follows is original to me, and I won’t pretend otherwise. The people who worked this out were running software projects, restaurants, factory floors, and (no kidding) ancient Roman households, often centuries apart, with nothing in common except the daily problem of getting real work done through other people. What’s striking is how often they reached the same conclusions. Different words, but the same answers. When a Roman estate manager and a modern software architect both land on “keep the team small and give each thing one clear owner,” it’s probably worth listening to.
I come at this from engineering, so that’s where my examples come from. But almost none of it is really about engineering. Swap the code for plates or pallets or patients and the advice barely changes, because the hard part was never the work itself. It’s the people – their worries, their pride, their need to feel that what they’re building actually matters.
So: a list. Think of it as the notes I’d hand a friend who just got their first team and is quietly terrified – not rules, just things that have helped. Some will sound obvious until you try them. Some will sound wrong until the alternative burns you once. Take what’s useful, leave the rest, and tell me where I’ve got it wrong.
Hire slowly, and when you’re genuinely torn, don’t.
This is the highest-leverage thing you will ever do, and almost everyone rushes it because the seat is empty and the empty seat hurts. But a bad hire doesn’t just fail to help – it actively costs you, for months, in cleanup and morale and the time you spend managing the mess instead of doing your job. An empty seat is cheaper than the wrong seat. Every time.
You can read more than you think at the gate, if you actually slow down and pay attention: not just “can this person code” but temperament, how they handle being wrong, whether they’re putting on a performance for the interview. The old hands will tell you the best people are neither timid nor impossible to steer – the terrified ones won’t push through hard problems, and the cowboys are a nightmare to point in any direction. And when your gut is genuinely split fifty-fifty? That’s not a coin flip. That’s a no.
Get the right people first, then figure out where you’re going.
This one feels backwards and it’s one of the most reliable findings in the whole business. The instinct is: set the strategy, then staff up to execute it. The better move is to fill the bus with the right people before you’ve settled the route – because the world is going to change the route on you anyway, and a bus full of the right people can handle a detour. A bus full of the wrong people can’t, no matter how good your map is.
It also fixes something else. When you get the who right, most of the “motivation problem” you hear so much about just... evaporates. The right people show up motivated. Your job stops being to light a fire under them and becomes the much nicer job of clearing the junk out of their way. Which brings us to the next one.
Once you’ve got good people, get out of their way.
If you hired carefully and then spend your days hovering, you’ve thrown away the thing you paid for. Good people don’t need a chaperone; they need room and air cover. Delegate real responsibility (not just tasks, but ownership) into a chain of people you actually trust, and then let them own it.
I know how hard this is, especially if you got promoted because you were the best at the work. The hardest day of becoming a manager is the day you accept that solving the problem yourself, again, is now the wrong move – even though you’re faster at it, even though it’s satisfying. Your value isn’t your output anymore. It’s everyone else’s.
Pick one goal everybody can say out loud without checking a document.
Walk up to anyone on your team and ask what the single most important thing is this quarter. If they have to open a deck, you don’t have a goal – you have a wish list with good production values. Real focus is brutal: it’s deciding what you are uniquely able to be the best at, what you actually care about, and what genuinely pays the bills – and then living in that overlap and saying no to everything outside it.
And “saying no” isn’t a vibe. It’s a list. Keep a stop-doing list that’s as long and as cared-for as your to-do list, because complexity accretes on its own (nobody ever schedules time to remove a process), and somebody has to be the one who prunes.
Adding people to a late project makes it later.
This is the one that gets the squint, every time, because it’s so contrary to instinct. The project’s behind, panic sets in, so we throw bodies at it. But think about what actually happens. The new people don’t know anything yet, so your best people (the only ones who could dig you out) stop working to onboard them. Meanwhile the number of conversations that have to stay in sync explodes: two people have one line of communication between them, but ten people have forty-five! Coordination eats the gains alive.
And some work simply can’t be parallelized no matter how many people you assign – the old line is that nine women still can’t produce a baby in one month. So when you’re late, you don’t get to add people. You get two honest levers: cut the scope, or move the date. Pick one and say it out loud, today, while it’s only a little embarrassing instead of a lot.
Pressure does not make anyone think faster.
You can lean on people and squeeze out a few more hours. Fine. But the work that matters (design, architecture, judgment) is thinking, and you cannot make thinking go faster by frowning at it. The data on this is pretty grim: past a small bump, all that overtime buys you is more bugs, more wishful estimates (”yeah, yeah, it’ll be fine”), and good people slowly burning out and leaving. The juice runs out fast.
So set a deadline you actually believe in, and then defend it like it’s yours – because it is. A date you privately know is fantasy doesn’t motivate anyone; it just teaches the team that your dates are fiction, and now you’ve got two problems.
Make people feel safe, or they’ll quietly stop taking risks.
Fear makes people conservative. When they don’t feel safe, they stop raising the weird idea, stop flagging the problem early, stop sticking their neck out. And the catch is that avoiding risk is itself the biggest risk, because you forfeit every upside that comes from someone being brave. A team playing not-to-lose never wins much.
What makes people feel unsafe isn’t only the boss who blows up. It’s the possibility of it – power that merely could land on them. Authority you never actually use still chills the room, because people manage to the worst case they can imagine. Your job is to make the safe move and the bold move the same move. (And a small, useful piece of folk wisdom from the people who’ve watched a lot of managers blow up: anger is almost always a fear projection. Once you notice you’re scared, you stop being quite so angry – try it.)
Go looking for conflict on purpose.
The most dangerous thing in a room is everyone being polite. If nobody’s pushing back, it’s not because you’ve achieved harmony – it’s because people have checked out, or they’re nervous, and the disagreements have just gone underground where you can’t fix them. Healthy teams argue. They argue hard, about ideas, and then go to lunch together.
So mine for it. When you can feel a buried objection in the room, drag it into the open and thank the person who voices it. The trick is keeping the fight pointed at the problem and never at the person – that’s the whole line between a great debate and a poisonous one. And if your meetings are boring, that’s not peace – that’s everyone quietly agreeing to avoid the conversation that actually matters.
Trust is built on vulnerability, and you have to go first.
People don’t come to trust each other through competence demos. They come to trust each other the first time someone says “I screwed this up” or “honestly, I don’t understand this part – can you help me,” and the sky doesn’t fall. That moment gives everyone else permission to drop the armor too. And the only person who can reliably create that moment is you, by going first.
It can be remarkably low-tech. One thing that genuinely works: get the team to share a little ordinary human context (where they grew up, how many kids were in the family, the worst job they ever had), because most of the friction between teammates comes from inventing unflattering stories about each other in the absence of any real information. Fill in the blanks and a lot of imaginary conflict just disappears.
Aim for clarity, not consensus.
Waiting for everyone to agree is how decisions die. You get mush, you get delay, and you usually get a worse answer than if one person had just decided. The better standard is “disagree and commit”: you let everyone be genuinely heard (and I mean genuinely, not the fake listening where you’ve already decided), and then you make the call, and people get behind it even if it wasn’t their pick. People don’t actually need to win. They need to know their argument was really considered.
A trick for the folks who freeze up because they’re scared of choosing wrong: walk the room through the worst case out loud. Nine times out of ten the worst case is “we lose a couple of weeks and learn something,” which is survivable, and naming that frees everyone to commit. Then close every decision the same way (who’s doing what, by when), and before anyone leaves, agree on what you’re going to go tell your own teams, so you’re not each broadcasting a different version by dinner.
Credit out the window, blame in the mirror.
When things go well, point out the window – at your people, at good timing, at luck. When things go badly, look in the mirror first (at yourself). This is the single cheapest trust-building move available to a leader and it’s astonishing how rare it is, because the ego pulls hard in the exact opposite direction.
There’s a flip side, too. The heroic, charismatic, look-at-me leader correlates badly with building something that lasts. Why? Because a company organized around one genius is really “one genius and a thousand helpers,” and the moment the genius leaves (or just gets it wrong), the whole thing folds. The quieter leaders who pour their ego into the team instead of themselves are the ones whose teams keep winning after they’re gone.
Tell the brutal truth, and keep the faith anyway. Both at once.
There’s a famous story about the highest-ranking American officer held in a brutal POW camp for years. Asked later who didn’t make it out, he said: the optimists. The ones who kept saying “we’ll be home by Christmas,” and then Christmas came and went, and they died of broken hearts. He survived by holding two things simultaneously – total, unflinching honesty about how bad it actually was, and an unshakeable faith that he’d prevail in the end.
That’s the posture. Don’t sugarcoat the situation and don’t let the situation crush the belief. Practically, that means building a place where the truth can actually be said: lead with questions instead of answers, run post-mortems that hunt for causes instead of culprits, and reward (genuinely reward) the person who raises the red flag. It even helps to appoint someone whose explicit job is to be the pessimist, because a relentlessly upbeat “can-do” culture is dangerous: it’s the one where nobody dares deliver the bad news until it’s too late to do anything about it.
Pay for results, not for being in the building.
Few things demoralize a strong performer faster than watching someone who coasts collect the same reward. So differentiate – not out of harshness, but because fairness is what your best people actually care about. Let the recognition, the bonus, and especially the good opportunities flow to outcomes – not to hours logged, not to who talks most in meetings, not to tenure for its own sake. People will, sensibly, do more of whatever you reward.
And be honest about what “reward” mostly means here: money. There’s a polite fiction that the best people are above it, that they’re in it purely for the craft or the mission. They aren’t, and neither are you. Pay is how people keep score and how they know they’re genuinely valued, even the ones who’d never say it out loud and would be a little embarrassed to admit it. So when someone consistently delivers more, pay them more, and make sure they feel it. Meaning and praise are real and they matter, but they don’t cover anyone’s rent – and a strong performer who quietly decides they’re underpaid for what they bring is already halfway out the door, whether they’ve realized it yet or not.
Give people a real path and a real stake.
This overlaps with money, but it’s a different lever, and the difference is worth getting right. Good pay mostly keeps people from leaving; it takes away a reason to be unhappy, and that’s about all it does on its own. What actually makes someone push past the bare minimum is the feeling that they’re going somewhere, and that they own a piece of what they’re building. The people who figured out how to get genuine effort out of others (going back a very long way) kept landing on those same two levers: a credible path forward, and a real stake in the outcome.
A path means there’s somewhere to climb to and people genuinely believe they can get there – not a poster on the wall, a real ladder with real rungs. A stake means they share in what they build: a cut of the upside, equity, ownership of a whole area, sometimes just unmistakable credit for the thing being theirs. A stake can absolutely be financial, but the difference from salary is the point – it ties a person’s own future to the outcome, where a steady paycheck arrives whether the thing succeeds or not. Give someone something that genuinely resembles ownership and watch them start behaving like an owner. It’s not magic. It’s just self-interest, pointed in a useful direction.
One coherent vision, and exactly one owner per thing.
A product (or a strategy) that hangs together because one mind (or a few tightly-aligned minds) shaped it will almost always beat a pile of individually-clever ideas bolted on by committee. Coherence is a feature, and it’s fragile, and somebody has to guard it. So protect it: let someone own the shape of the whole.
And at the operational level, attach exactly one name to every process and every decision. The failure mode is sneaky, and it sinks good teams quietly: when something is “everyone’s” responsibility, it’s nobody’s, and it rots while each person reasonably assumes someone else has it handled. One throat to choke, one person to thank. Both.
Build the system so quality outlives the people.
If your standards live only inside the heads of two or three veterans, you are one resignation away from chaos, and you don’t get to choose the timing. In some businesses turnover runs north of 100% a year, and the ones that stay great anyway are the ones that wrote it down: the standards, the checklists, the “this is how we do it here,” automated wherever a machine can carry the load instead of a person’s memory.
Writing it down does something useful that people underrate: the gaps and contradictions only show up once you try to put the thing into actual words. Hand-waving in a meeting hides a hundred unresolved decisions. A blank page is merciless about it, in the best way.
Stop waiting for the silver bullet.
Last one, and it’s probably the most pragmatic one: there is no single tool, framework, hire, reorg, or magic technology that delivers the 10x leap. The hard part of hard work is irreducibly hard – that’s not a tooling problem you haven’t solved yet, it’s the nature of the thing.
What actually works is slower and a lot less exciting: many small, consistent pushes in the same direction. Picture a heavy flywheel. The first shoves barely move it, so you keep pushing, and at some point it’s spinning on its own – but there’s no single push you can point to and call the moment it took off. From the outside the win looks sudden. It never is. So plan to throw the first version away (you will anyway), don’t stuff everything you cut back into the second one, and just keep turning the wheel. The boring consistency is the strategy.
Let’s talk about where all this finally leads. Picture a “successful manager” for a second – for most of us, the image that shows up looks something like this:
The sharp suit or the elegant dress, a cocktail in hand, working a room full of important people with the easy confidence of someone who has clearly Made It. Now ask what any of that actually has to do with managing anyone. Someone can be handed the title, sit above a few names on the org chart, get introduced as the boss in every meeting. Does that make them a manager? Not on its own. Managing is a whole job in its own right, not a hat you put on while you keep doing the work you did before.
But nobody does both well at once: every hour you spend being the best worker in the room is an hour you’re not being a manager. How hard that switch feels depends on you. For some it comes naturally; for others (plenty of excellent managers included) it’s a deliberate, learned skill. Both end up effective – as long as you’re honest about which you are, and aim your effort accordingly.
If I had to squeeze this whole episode into one line: fill the bus, then pick the road. Get the right people on board first, then point them at the one destination that matters, clear the road, and stay out of their way – almost everything else here is a detail hanging off that. My honest opinion, for what it’s worth, is that the managers worth working for aren’t the ones with the cleverest framework; they’re the ones who quietly decided that people are the job, and do it on purpose.
Take what’s useful, leave the rest, and tell me where I’ve got it wrong. See you next time!