Be Like Water
The best operators don't force outcomes. They find the path of least resistance, then quietly make it the only path anyone takes.
Essay · 15 June 2026 · 5 min read
My poh-poh ran a congee stall in Sham Shui Po for thirty-one years. If you'd asked her for a growth strategy, she'd have handed you a bowl and told you to sit down.
But she was running one of the tightest operations I've ever studied. The suppliers came at five-fifteen, so she was there at five. The broth needed six hours at a low simmer, so the fire was lit by five-thirty. Lunch peaked at noon, so the pot was ready at eleven-thirty — never early, never late. She'd built the whole business around a single fixed constraint, the broth's timeline, and let everything else fall in behind it.
She never called it a system. It was the best one I've ever seen.
Water isn't the soft option
Everyone knows the line from Bruce Lee: be like water. It gets quoted to mean go easy, stay loose, don't force things. That's the wrong half of it.
Water isn't gentle. It carved the Grand Canyon, layer by layer, over five million years. What it does is refuse to spend effort it doesn't have to — it finds the lowest point, goes there, and doesn't stop going there. Patient and relentless turn out to be the same trait, measured over different timeframes. 唔怕慢, 只怕站 (m̀h pa maahn, jí pa jaahm), my poh-poh used to say — don't fear going slow, only fear standing still.
I spent my first decade as an operator doing the opposite. I treated effort as the lever: work harder, push more, get further. That works right up until you're standing at a wall, shoving, with a door six feet to your left.
Find the constraint
Here's what I had backwards for years: I thought output was a function of effort. It's a function of the bottleneck.
Later, as a COO, I watched the same thing in every operation I ran. The work that flowed had a clean handoff and one owner per decision. The work that stalled, stalled in the same place every time — a missing input, or a person who was the only one who knew a thing. And piling more effort upstream of that point didn't move more work through it. It just stacked up in front of the constraint, waiting.
That's the operator's version of being like water. Stop pushing on the whole system. Find the one point where the resistance actually is, and put your attention there. Everywhere else, take the path of least resistance on purpose.
Make the right path the only easy one
The best systems don't rely on anyone trying harder. They make the correct move the path of least resistance — and the wrong move faintly annoying.
Manufacturing has a name for it: poka-yoke, error-proofing. A USB plug that only fits one way. A car that won't leave park until your foot's on the brake. You're not depending on memory or discipline. You've shaped the system so the right action is simply the easiest one available.
My poh-poh did this without the vocabulary. One pot. No substitutions, no off-menu requests. You got what was in the pot, or you ate somewhere else. She'd narrowed the path until it only ran one direction — and the people who walked it came back the next day, and the next.
The friction you've stopped seeing
The thing that quietly kills most operations isn't laziness. It's friction you've stopped noticing because you've spent years routing around it.
Do the arithmetic. A step that takes thirty seconds, run a hundred times a day, is fifty minutes — every day. Across a team of ten, that's most of a full-time role disappearing into a step nobody chose. And the longer it's there, the harder it is to see: first it's a workaround, then it's a habit, then it's "how we do things."
The fix isn't to move through it faster. Stop. Look straight at it. Ask whether it needs to exist.
It usually doesn't.
My poh-poh retired at sixty-six — not because trade dropped off, but because she was ready. She'd built something that ran on its own rhythm. It didn't need her to force it. She'd found the natural shape of the thing and arranged everything else around it, and after thirty-one years the path was so worn it ran itself.
Find the constraint. Work with the grain. Be patient.
Water always gets through.
