Operational Learning: The Missing Path

The conversation was supposed to be a routine one-on-one. A small table near the window. Two coffees. A calendar block that had been on the books for weeks, the kind of standing time that tends to drift toward catch-up when nothing’s on fire.

Halfway through, almost in passing, one of the strongest people on the team mentioned something they’d been working on. A vendor who had been quietly missing commitments. They’d been managing it. Building workarounds. Adjusting the timeline themselves. They’d been doing this for about a month.

They weren’t asking for help. They weren’t flagging a risk. They were describing what they’d been doing the way someone describes their morning commute. Routine. Already absorbed.

The recognition arrived slowly, in the moments after they finished describing the vendor situation. As they kept talking about the rest of the week, I started noticing the small adjustments they’d made along the way. The deadline they’d extended for a colleague rather than name the underlying cause. The calendar reminder they’d built for themselves to follow up with the vendor a third time. The decision they’d made, somewhere around week two, to stop including the vendor’s status in their weekly note because the status would have been the same three updates running. Each of these adjustments had been rational on its own. None of them, viewed individually, looked like absorption. Together, they were the shape of a month spent solving a problem the system was supposed to solve at a different layer.

They hadn’t decided against escalating the vendor relationship. They hadn’t weighed it and chosen to handle it themselves. The decision to escalate had simply never entered the conversation they were having with themselves about their week. It wasn’t a move they ruled out. It wasn’t a move they considered.

What I asked next isn’t the part I remember. What I remember is the silence that came after. Theirs, not mine. The moment when the absorption stopped being invisible to them. They had the same recognition I’d just had, in the same room, about the same set of small adjustments they’d been making for a month.

That was the moment the rest of this piece is built around. Not the absorption. The fact that the absorption had been invisible to them as a choice. The path up wasn’t unsafe in any dramatic sense. It was just absent. And the person across the table from me had been holding something for a month not because they were trying to be a hero, but because nothing in the operating environment had ever told them that sending it up was a thing the team did.

The path up has to exist before owners need it. Most of the time, it doesn’t.

That absence is more common than most leadership teams realize, and more expensive than most reporting layers can show. That conversation isn’t unusual. It’s the most common shape of broken escalation, and it’s also the hardest to see. Broken escalation doesn’t make noise. It doesn’t show up as missed deadlines. It doesn’t show up as failed projects. It shows up as decisions that get absorbed in places no one is looking, and the cost gets paid in directions the reporting layer was never designed to catch.

A strategic vendor relationship that should have been renegotiated three weeks ago becomes a margin issue two quarters later, and by then the leverage is gone. A scope cut that should have moved up the chain becomes a customer relationship that quietly degrades. A hiring decision that crosses functional boundaries gets made by whoever was willing to absorb the awkwardness, not by whoever was best positioned to make the call. None of these surface as escalation failures. They surface as something else, eventually, which is why most leadership teams never trace them back to a missing path.

This is the part the chassis arc keeps returning to. Operational systems that look healthy under low pressure can be broken under sustained pressure, because the failure mode isn’t visible at the surface. Tucker and Edmondson’s (2003) study of hospital workarounds names the mechanism clearly: frontline workers solve immediate problems individually, and the immediate solutions hide the systemic problem from anyone with authority to fix it. It’s structural. It’s invisible by design, not because anyone designed it to be invisible, but because no one designed it at all. The path up was never built. The path down was never built. So the path the owner uses is the one inside their own head, which is sometimes excellent, often adequate, and occasionally catastrophic.

The cost is real. It’s just paid in a currency leaders aren’t measuring yet.

I knew what I was looking at across that table because I had been the person absorbing things alone. Earlier in my career, more than once, I’d held decisions that should have moved up because the path up wasn’t findable to me either. I’d convinced myself it was discipline. It wasn’t. It was the absence of a path I could trust.

Escalation isn’t a culture problem. It’s a system problem that produces cultural symptoms. When the path up doesn’t exist, owners don’t develop bad habits. They develop the only habits available to them. They absorb. They work around. They wait. The pattern has a name in the literature on organizational voice. Implicit voice theories, in Detert and Edmondson’s (2011) framing: the unspoken rules about when speaking up is safe, learned not from explicit instruction but from watching what gets rewarded and punished.

From the outside, this looks like ownership. The org celebrates it. Performance reviews reward it. And the system slowly accumulates risk it cannot see. Spend was about where to use friction on purpose. Send is the harder pair: where the decision goes when the owner runs out of room to spend. Both have to be designed. Neither survives if the design is left to chance.

The first move is making the path legible. The discipline isn’t the leader’s visibility. It’s the path’s visibility. Owners have to know where decisions go before they’ve been holding them too long. That means naming the path explicitly, in writing, in onboarding, in one-on-ones. Decisions of this shape come to me. Decisions of that shape, we go together. Decisions of that other shape, you make and I back you. Said once, this is a memo. Said repeatedly, in real moments, against real decisions, it becomes a map. People use the map because they trust it. They trust it because they’ve watched it be used.

The first time someone tests the map, you find out whether the path holds. They bring you a decision they think belongs in your hands. Sometimes it does. Sometimes it doesn’t. Either response teaches them what the path actually allows, and which routes are theirs to walk alone. The map becomes a map after it’s been walked, not before. Until people have seen what happens at the joints, the path is just words on a page.

The second move is harder, and it’s the one most leaders skip. Once the path is named, owners will start using it. Some of what they bring will be theirs to make, not yours. The temptation in those moments is to catch it anyway, because catching feels like leadership. It isn’t. It’s design erosion. The right move is to hand it back, with framing. This is yours. Here’s the shape of it. Here are the two or three things I’d weigh. I’ll back you when you decide. Refusing the catch sharpens the path. It tells the system which decisions belong where.

The first refusal is harder than it sounds. The owner sometimes comes back the next day with the same question framed slightly differently, hoping the path has shifted. It hasn’t. The framing has to hold the second time, and the third, until the owner trusts that the route up isn’t going to move under them. Refusing once is a mood. Refusing consistently is design.

Without this move, the leader becomes the path. A path that runs through one person isn’t a path. It’s a dependency.

The harder move comes after that, and it’s the one you can’t engineer directly. You wait for replication. Not the catching. Anyone can catch. The framing. The handing-back. The naming of which decisions belong where.

The first time it happened, I almost missed it. A senior IC, in a meeting I was only half tracking, framed a decision back to a peer the same way I’d been framing them. This is yours. Here are the two or three things I’d weigh. I’ll back you when you decide. They didn’t know I was watching. The discipline had landed in someone else’s voice. That moment is the only signal worth waiting for. Until it shows up, the path runs through one person, which means the system is one absence away from breaking.

For senior leaders, the question isn’t whether you personally model the catch. It’s whether the organization has enough leaders modeling it that the discipline replicates without you. If the answer is no, the work isn’t escalation training. It’s escalation architecture. And it has to be designed at the layer that has authority to design it, which is usually one or two altitudes above where the absorption is actually happening. The owner absorbing a decision alone for a month is rarely the place to start. The place to start is the layer above them, asking why no one above the owner ever made the path findable.

This work earns its reward in places the reporting layer doesn’t reach. No one will write a quarterly review about the team escalates well now. The proof is what doesn’t happen. Fewer decisions absorbed in private. Fewer absences that last a month. Fewer surprises two quarters later that turn out to have been visible the whole time, just not to anyone with the authority to act. The signals are quiet. The discipline runs the system anyway.

Most of the cost of broken escalation is already on your books. It just isn’t labeled that way yet. Somewhere in your operation, someone is holding a decision that should have moved up three weeks ago, and the reason they’re holding it is not that the path is unsafe. It’s that the path was never built. They aren’t waiting for permission. They aren’t avoiding accountability. They’re operating the way the system has trained them to operate.

What is the most expensive decision in your organization right now that no one has officially decided to make?

Sources

Detert, J. R., & Edmondson, A. C. (2011). Implicit voice theories: Taken-for-granted rules of self-censorship at work. Academy of Management Journal, 54(3), 461–488.

Tucker, A. L., & Edmondson, A. C. (2003). Why hospitals don’t learn from failures: Organizational and psychological dynamics that inhibit system change. California Management Review, 45(2), 55–72.