10 The Art of Judgment
This chapter is currently in draft form.
You won’t always have access to the top lever. Sometimes the CEO owns strategy and won’t budge. Sometimes structure is frozen by politics or history. Sometimes the people you need to prune are untouchable. That’s real life. The four chapters before this one gave you the levers and the order they sit in. This one is about the harder thing: choosing well when the lever you should pull and the lever you can reach are not the same.
Start with honesty about constraints. If the highest lever you can pull today is process, fine, but name the trade. You’re buying relief, not resolution. You’re managing symptoms while you accumulate the political capital for a higher cut.
The discipline is simple to state and hard to practice: diagnose up, intervene as high as you can. When a problem presents at level N, look one or two levels up before you act. Two teams that grind against each other look like a people problem; far more often they are a structure problem wearing people’s names. A process that keeps failing looks like it needs another ceremony; far more often the strategy above it never told anyone what to optimize for. Find the highest level where the real fault lives. Then intervene as high as you can actually reach, which is not always where the fault is.
10.1 There Is Never One Lever
But “intervene as high as you can” can mislead, because in practice you never pull just one lever. The chapters separate strategy, structure, people, and process so you can see each clearly. The job never separates them. At any given moment you’re sharpening strategy while you reshape a team while you redesign a process that’s quietly failing. The levers move together or they don’t move at all.
This is clearest when the one you most want is locked. The structure chapter described a frozen org I couldn’t touch; the move there was not to stop and wait but to work the levers above and below it at once until the system held. Perfect is the enemy of good here. A functioning org isn’t a consolation prize for losing the lever you wanted. It is the normal condition of the work.
You don’t have to be blocked for the same logic to bite. I’ve run a particular play many times: I’m brought in to raise the engineering maturity of an organization, and I start where the dysfunction looks most obvious, with the people. I coach. I help some people grow, and I help some people out. It produces results, just not great ones. Some movement, not transformation. That mediocre return is the tell that I’m leaning on one lever when the work needs several.
The results only changed when I stopped treating it as a people problem and started working the levers together. I went deeper on people, but differently, showing managers what serious mentorship actually looks like, with real artifacts, run through situational leadership rather than left to instinct. I reshaped the structure around how those people actually needed to interact. Then I reshaped process so the new structure had teeth, holding people accountable to it, and the accountability itself began producing the behavior I wanted. No one of those moves would have carried the load alone. Together they compounded.
From the inside it was anything but elegant. The mentorship piece in particular took forever, an ongoing grind with no clean finish line. I’d get traction on structure and process and then watch the people layer stall again, and I’d feel the frustration rise every time. The discipline was catching myself and routing that frustration back into a question instead of a complaint: what’s not working here, why isn’t it working, what do I address next? It was constant. Less like throwing a switch and more like playing a piano, both hands moving up and down the levers, adjusting the pressure on each as the music demanded.
So the art is not choosing the lever. It is choosing which levers to pull at once and how much force to put behind each. Most of your weight belongs on the highest lever you can actually move. The rest gets distributed across the others, enough to hold the line, compensate for what you can’t reach, and keep the system from drifting while your main cut lands. And the distribution is never fixed. As you earn capital, as constraints loosen, as the launch you were protecting finally ships, the weighting shifts. You are constantly rebalancing a finite supply of capital, attention, and force across several moves at the same time.
That is why this is judgment and not a formula. A formula returns one answer. Judgment decides which levers to pull together, how hard to pull each, and how to keep adjusting that allocation as the situation moves under you.
10.2 The Gap Between Seeing and Acting
That gap, between where the problem lives and where you can act, is where judgment actually happens. Seeing the right lever is the easy part; most experienced leaders can. The hard part is the honest assessment of whether you can pull it. Two different things sit in that gap, and judgment is mostly the discipline of telling them apart.
The first is capital. Sometimes you can see exactly what’s wrong and simply haven’t earned the standing to fix it. You’re three months in, the structure is incoherent, and a reorg now would spend credibility you don’t yet have and may not survive. That isn’t a failure of nerve. It’s sequencing. You hold the diagnosis, you build the capital, and you pull the lever when you can afford the cost. It’s the same restraint the strategy lever demanded: don’t spend capital trying to fix what you haven’t been given the standing to touch.
The second is courage. Sometimes you have the capital and you’re flinching anyway. You can see the conversation you keep not having, the underperformer you keep working around, the structural fight you keep deferring because it will be unpleasant. That isn’t sequencing. That’s avoidance wearing the costume of patience. The two feel identical from the inside, which is exactly why judgment demands brutal honesty about which one you’re in. The remedies are opposite: one says wait and build, the other says stop flinching and act.
In practice the two get tangled, and the most expensive mistake is letting the lever you can’t reach become the alibi for the one you can. We once spun up a new product. Small team, but it mattered; the company was counting on it for growth. The PM and the engineer were both struggling badly, and our weekly check-ins had curdled into a ritual: I’d give the same coaching, they’d give the same response, and nothing moved. Meanwhile, by happenstance, I was in front of the C-suite regularly, and for three months the thing we talked about most was how badly this one project was going. It was bleeding my credibility at the top.
Two levers were in play, and they were not equal. Product and engineering didn’t report to each other; they only unified at the CEO. The product side was weak and I had little real influence over it, and the personalities on the two sides grated against each other in a way that made everything worse. So I told myself the situation was structural and mostly out of my hands, and I kept working the lever I felt good about: mentoring, coaching, trying to grow the people into the roles.
But I did have a lever fully in my control. The engineering seat was mine. I should have moved that person and put someone else in within weeks, not months. I didn’t, partly because a dozen other fires were burning and partly because the unfixable product side gave me a story for why nothing I did would matter anyway. That was the mistake. The product reporting line being beyond my reach was real, but it had nothing to do with whether I made the staffing change I actually controlled. I let the locked lever become the excuse for the open one.
It cost me. Three months of visible dysfunction spent capital I didn’t need to spend. I carried enough in that role to absorb it and eventually earn it back, but I should have acted far sooner. The lesson stuck: when you catch yourself explaining your inaction by pointing at the lever you can’t move, go check the levers you can. The constraint is usually real. It is also usually not the reason you’re standing still.
10.3 The Two Ways Judgment Fails
Judgment fails in two characteristic directions, and both are worth naming because you’ll feel the pull of each. The first is reaching too low. Process is visible, safe, and infinitely adjustable, so the tired or the cautious leader retreats into it, another standup format, another planning ritual, while the structural rot one level up keeps producing the very problems the ceremonies are meant to fix. The second is reaching too high too soon: the reorg reflex. Restructuring feels decisive, and for a leader who wants to look like they’re leading, it’s seductive. But structure has a human tempo. Pull that lever faster than people can absorb the change, or before you’ve earned the capital to spend it, and you exhaust the organization and yourself for motion that reads as progress and isn’t. One failure hides from the hard cut; the other swings the axe before the blade is earned.
Timing matters as much as level. Sometimes you spend capital early and hard, because the drift you’re watching will compound into a year of waste if you wait. Sometimes you hold, because a forced restructure right before a critical launch will cost more than it saves. Don’t confuse urgency with importance. The urgent work clusters at the bottom of the pyramid, where feedback is fast and visible. The important work sits at the top, where feedback is slow and the payoff delayed. The leader who only chases the urgent never climbs.
Sequence matters too. Get strategy sharp and every downstream cut comes cleaner. Get structure right before you hire at scale, or you pour good people into bad plumbing. When you can’t control the sequence, when you’re hiring into a structure you haven’t fixed yet, say so plainly: “We’re staffing an imperfect org; here’s how we’ll hold organizational distance down until we can refactor it.” Naming the compromise keeps your team aligned on the real problem instead of mistaking the workaround for the design.
And not every misalignment demands a reorg. Sometimes a shared metric, an embedded specialist, or a standing weekly alignment collapses the practical distance between two teams enough to ship, without spending the capital a structural change would cost. These are not the real fix, and you shouldn’t pretend they are. They buy time, and they make the eventual structural change cleaner and cheaper when you can finally afford it. Buying time is a legitimate move as long as you know that’s what you’re doing.
What ties all of this together is the refusal to hide. You will always be tempted toward the safe lever, the visible ceremony, the change that costs you nothing. Resist it. Spend your capital where it compounds, even when that means fewer rituals, harder conversations, and progress no one can see on a dashboard for another quarter.
10.4 Closing Reflection
Strategy, structure, people, process. In that order of leverage.
You’ll get more from pulling the right lever than from pulling the wrong one harder. Diagnose up the hierarchy first. Put your weight on the highest lever you can actually move, and keep working the ones below it at the same time. Spend political capital where it compounds. And when you can’t reach the top lever, stabilize what you can with the levers you’ve got, name the trade, and keep chipping at the barriers until the higher cut becomes possible.
This is the Dao of Delivery. Not laws, not checklists. Judgment under constraint. Your teams need outcomes, not rituals; your organization needs flow, not theater. Checklists don’t ship products. People working within clear structures, toward sharp strategic goals, ship products.
Pull the highest lever you can reach. Keep your hands on the rest. That’s the job.