Operational Learning: What's Running Underneath

The new director of operations has been in the role for six weeks. Long enough to know the rhythm. Not long enough to have earned the right to change it.

She sits in the Thursday ops review for the fifth time. Same room. Same cadence. Same twenty-two-slide deck that takes forty minutes to walk and produces exactly one decision, usually a request for more data.

The supply chain lead reads through demand variances. No threshold attached. No line that tells the room whether these numbers require action or are simply movement within normal range. The room debates the number itself, not what it means. Ten minutes pass. Nothing is decided.

The quality lead surfaces an exception that happened nine days ago. A customer commitment created an unplanned production sequence, and the team handled it. But the exception never got classified. Nobody asked whether it was a one-off, a recurring pressure, or a signal that the standard needs revision. It just got handled and moved on. The learning died in the doing.

Nobody asks whether last month’s capacity decision should be revisited. The conditions that drove it have shifted. Two data points suggest the original assumptions no longer hold. But there’s no defined trigger that would earn the right to reopen it, so the decision sits unchallenged. Not because it’s right. Because nobody has a mechanism to say it might not be.

The plant manager mentions that the eastern facility handles the same deviation differently than the western one. He says it casually, almost as an aside. Nobody flags it. Nobody asks when the divergence started or why. The room moves on to the next slide.

A workaround from Q3 is still running. It was supposed to be temporary. Three people in the room know this. None of them raise it, because the workaround works well enough and there’s no moment in this meeting where that kind of observation has a home.

This room knows how to report. It doesn’t know how to decide.

She writes nothing down. She watches.

There’s something here worth protecting. She just has to find it underneath everything else.

Most organizations have built some version of the learning engine. The concepts aren’t the problem. Signal triage, reopen triggers, exception classification, drift indicators, portable memory. These are sound. Most senior leaders, when they encounter these ideas, recognize them immediately. They’ve felt the absence of each one in rooms exactly like the one that director is sitting in.

The problem is that these concepts have nowhere to live.

Learning only matters when it becomes runnable under pressure. And the gap between “we understand this” and “we practice this every week” is where most operational learning dies. Not because organizations are careless. Because the weekly rhythm is full, and status reporting feels safer than structured learning. A signal check competes with a slide deck. An exception review competes with an update that everyone already read. A drift flag competes with the meeting ending on time.

Status wins because it’s comfortable. Everyone knows their part. The format is predictable. The risk of surprise is low. But the cost of that comfort compounds. Every week the rhythm fills with status instead of learning, the re-decision tax grows. Triggers that should fire don’t, because they have no moment in the cadence to surface. Exceptions get resolved but never classified, so the same pressure produces the same workaround next quarter. Drift goes unflagged, and quiet drift becomes permanent divergence.

The engine is there. The surface isn’t. An engine without a surface is just generating heat. The energy is real, but it has nothing to grip. No traction, no contact with the road, no translation of effort into movement. The concepts spin. The weekly rhythm absorbs them as background noise. And the learning that should be building week over week simply doesn’t, because there’s no point of contact where it touches the work.

That point of contact has a shape, even if most organizations haven’t named it. Call it the operating surface: the minimum set of touchpoints where learning actually makes contact with decisions in the weekly rhythm.

The director in that Thursday meeting is already looking for it, even if she wouldn’t use that language. Every gap she noticed maps to a touchpoint that should exist but doesn’t. No signal threshold means no signal check. A nine-day-old exception with no classification means no exception triage. An unchallenged capacity decision means no trigger review.

A casual mention of plant-level divergence that nobody flags means no drift check. A temporary workaround with no home in the agenda means no memory capture.

Five touchpoints. Five places where the learning engine makes contact with the operating rhythm. Not a framework to adopt. Not a process to install. The actual moments in a week where the work of learning touches the work of operating. Argote and Miron-Spektor’s research on organizational learning routines points to the same conclusion: learning doesn’t transfer through documentation or training. It transfers through repeated interaction with the work itself, inside the cadence where decisions are made.

The beauty of the operating surface is that it doesn’t require a new meeting. It requires fifteen minutes of an existing one. The trade is direct: compress a status segment that produces updates into a structured check that produces decisions. The meeting may get shorter. But that’s a side effect, not the point. The point is that the meeting starts generating learning alongside action, and that learning accumulates instead of evaporating.

The temptation when you see a broken cadence is to rebuild it. New agenda. New structure. New expectations. Walk in, redesign, ship it Monday.

That director watching her fifth Thursday review is doing exactly what I do every time I step into a new role. She’s resisting that temptation. Not because she can’t see what’s broken. She sees it clearly. But she’s doing the work that comes before fixing: understanding why the cadence runs the way it does.

I see it in the first meeting too. The round-robin status updates that take forty minutes. The recurring deck that took someone hours to build and that nobody in the room will act on. The agenda item that exists because of an incident three years ago that nobody remembers but everyone still reports against.

I see all of it. And I don’t touch it. Not yet.

The work that matters first is listening. Learning the who, what, when, where, and why behind the current rhythm. Not the rhythm I would design. The rhythm they run, and the reasons it exists. Because every cadence protects something, even the broken ones. The round-robin protects visibility. The unused deck protects someone’s sense of contribution. The legacy agenda item protects against a fear that may or may not still be valid.

Until I understand what the cadence is protecting, I can’t know what’s safe to change. And if I change something without understanding the why, I’m not improving the system. I’m just imposing a preference.

In my current role, I’m still in this phase. Still listening. Still learning. We’re making positive changes, taking steps in the right direction. But the work isn’t complete. It never is. That’s the quiet part about operational learning that doesn’t show up in frameworks: the operating surface isn’t a project with a finish line. It’s a practice. You keep finding friction, understanding why, and making space for the teams to build better decisions. The moment you treat it as done is the moment drift starts.

My job is never to upset the apple cart and change for the sake of change. It’s to find the friction points, understand the why behind them, and create conditions where teams become curious about their own decisions. The operating surface is something you uncover by paying attention, and then you clear the path to it.

The operating surface, once uncovered, raises a harder question. Now project that director forward three years. She’s done the work. She sat in the cadence, understood what it protected, made careful changes over months. The Thursday ops review produces decisions. The signal check catches what matters. Exceptions get classified in real time. Drift gets flagged before it becomes permanent. Learning captures happen, and the next quarter’s team inherits context instead of starting from scratch.

Then she gets promoted. Or she takes a role at another company. Or the organization restructures and the rhythm she built gets absorbed into a different reporting line.

What happens to the operating surface?

I’ve lived this. I’ve built learning cadences that worked. Systems that surfaced exceptions in real time, that caught drift before it became failure, that produced real decisions instead of action items. And I’ve watched those systems break when the leader who reinforced them left. The structure held for a few weeks. Then the uncomfortable parts got quietly dropped. The exception triage shrank. The drift check disappeared. Status reporting expanded to fill the gap, because status is easier and nobody was asking the harder questions anymore.

Weick and Sutcliffe’s research on high-reliability organizations surfaces the same pattern: sustained performance depends on routines that survive personnel changes, not on individual vigilance. The organizations that maintain learning under pressure aren’t the ones with the best leaders. They’re the ones where the learning is embedded in the rhythm itself.

The operating surface has to survive the person who built it. If it depends on one leader’s presence and persistence, it’s a personality, not a system. The touchpoints need to be tied to roles, not names. The rhythm needs to be embedded in how the team operates, not in how one leader runs a meeting. That’s what makes the difference between a learning cadence that produces value for a quarter and one that produces value for a decade.

The trade-off is worth stating plainly. When you build an operating surface into your weekly rhythm, you protect decision quality, institutional memory, and standard integrity. Those three things compound. Better signal triage this week means fewer decisions get relitigated next week, and the re-decision tax drops. Captured memory means the next leader inherits context instead of guessing. Monitored standards mean drift gets caught as a flag, not discovered as a failure.

What you give up is the comfort of long status reviews and the feeling of control that comes from hearing every function report out. That feeling is real, and it’s hard to release. Status reviews make leaders feel informed. The problem is that informed isn’t the same as effective. A room full of updates produces awareness. A room with an operating surface produces learning.

However, this is not an argument for eliminating status entirely. Some updates matter. Some visibility is earned and necessary. The discipline is knowing which updates produce decisions and which produce only comfort. Keep the former. Compress or eliminate the latter. Use the space you create for the touchpoints that build the learning engine into the weekly rhythm.

The organizations that do this well aren’t the ones that adopted a framework. They’re the ones where a leader sat in the cadence, watched it run, understood what it was protecting, and then made one careful change. Then another. Then another. Over four weeks, the ops review shifted. Not because someone mandated a new structure. Because someone made room for the right questions, and the room discovered it could answer them.

Take your most important weekly operating review. Not the one you run for your boss. The one where your teams make decisions that affect next week’s work.

Remove one status-update segment. The one where someone presents information that the room already has, or that generates discussion but never action. Replace it with one touchpoint from the operating surface. A two-minute signal check. A quick trigger review. An exception triage. Just one.

Run it for four weeks. Watch what happens to the meeting’s output. Watch whether the team starts arriving with different expectations for what the room is supposed to produce.

Then ask yourself the question that matters: 

What would you remove from your operating cadence to make room for learning that sticks?

Sources

Argote, L., & Miron-Spektor, E. (2011). Organizational learning: From experience to knowledge. Organization Science, 22(5), 1123–1137.

Edmondson, A. C. (2019). The fearless organization: Creating psychological safety in the workplace for learning, innovation, and growth. Wiley.

Weick, K. E., & Sutcliffe, K. M. (2015). Managing the unexpected: Sustained performance in a complex world (3rd ed.). Wiley.