Why Your Organization Is Running Faster and Getting Less Done
There's a quote that opens Nelson Repenning's book, There's Got to Be a Better Way, that stops you cold. A manager in one of his classes described it this way: "I know my organization's in trouble. When I sit in my office Monday morning, I make a list of all the things I need to get done that week. I work my ass off all week, and then I check the list Friday afternoon. I didn't get anything done."
If that sounds familiar, Nelson has a name for what you're experiencing. And more importantly, he has a way out.
The Capability Trap
Nelson is faculty director of the MIT Leadership Center and has spent his career studying why organizations that are working incredibly hard still fail to improve. The central concept in his book is the capability trap—a cycle that begins with good intentions and ends with an organization that can only survive by fighting fires.
Here's how it works. Every leader knows they're supposed to balance short-term results with long-term investments in capability. But under pressure, those long-term investments get quietly cut. And when they do, delivering the same results next quarter gets a little harder. So you cut a little more. Before long, you're in an environment where every resource is consumed just keeping the lights on—and nothing is left over to solve the underlying problems.
The trap closes.
What makes it particularly insidious is that over time, organizations get good at it. They develop impressive muscle for last-minute heroics, for pulling things out of the fire at the eleventh hour. Nelson described watching this at Harley-Davidson, a company that had elevated firefighting to something almost admirable. The human cost, of course, was enormous. And the race they were winning was one no one wanted to be in.
AI as a Workaround Engine
One of the sharpest exchanges in this conversation came when the question turned to how organizations are currently deploying AI. The pattern is familiar: hand out the tools, encourage people to experiment, and see what sticks. The hope is that something useful emerges that can later be standardized and scaled.
Nelson's concern—and it's a pointed one—is that this approach may be doing the opposite of what's intended.
He drew on a concept that runs throughout his book: the gap between formal process and actual practice. In most organizations, there's the official way things are supposed to work and the real way things get done. Healthy organizations are constantly closing that gap through problem-solving. Struggling organizations let it widen until the workarounds become the work.
Introduce AI into that environment without first understanding it, and you don't eliminate the workarounds. You multiply them. What was five ways of doing something becomes fifty.
Nelson put it plainly: "Trying to automate a process you don't understand is probably one of the fastest ways to lose money that organizations have invented."
He cited a PhD student of his studying AI in clinical settings. Doctors were using voice-to-text tools to generate patient notes and loved the efficiency they provided. But specialists on the receiving end had stopped reading the notes entirely. They were too long, and the AI made errors they couldn't trust. A win on one end equated to a loss on the other, and the overall outcome was almost certainly negative.
The lesson isn't that AI is dangerous. It's that a solution looking for a problem is always dangerous.
The Utilization Trap Inside the Capability Trap
Another thread Nelson pulls at is the confusion between utilization and productivity, and why maximizing the former can actively destroy the latter.
No factory manager would try to run their lines at 120% of capacity. The logic is obvious in physical systems: you need slack to absorb variability, or the whole thing seizes up. But in knowledge work, that lesson evaporates. The assumption becomes that any resource not fully occupied is wasted.
The result is systems running permanently overloaded. Overflowing inboxes. Blurred priorities. And because everyone is constantly firefighting, no one can actually identify the bottleneck—it's hidden behind a wall of work-in-process. The organization stays busy. It does not get better.
Nelson's framing here is useful: overload is an express lane into the capability trap. When people have too much to do, they almost always default to short-term delivery. Long-term investment disappears. The cycle accelerates.
The AI parallel is uncomfortable but accurate. If the pitch for a new tool is that it will push everyone to 120% utilization—that you'll essentially get a full-time contributor who never sleeps—you've built the trap into the business case before you've written a single line of code.
When Something Goes Wrong, Look at the System First
One of Nelson's strongest positions—shaped by years of work in high-hazard industries—is where to place blame when errors occur.
His answer: almost never on the person.
In his experience investigating incidents and near-misses, the underlying story is almost always the same. The system was not set up for success. It relied on heroic behavior from the people doing the work. And somewhere, something that should have been designed out was left to human judgment under pressure.
This isn't just a philosophical point. It has a practical consequence: if people believe they'll be blamed when something goes wrong, they stop surfacing problems. The near-misses go unreported. The early warning signals disappear. And the organization loses exactly the information it would need to actually get better.
Start with the system, Nelson argues. Build a job that any normal person could do correctly on a regular basis. Get the process right. And then—only then—if someone still won't follow it, that's a different conversation.
Two Things a CEO Can Do Tomorrow
Nelson closed with two concrete moves for leaders who feel trapped in the day-to-day.
The first: go see the work. Carve out an hour a week to actually observe what people are doing and what they're dealing with. Nelson's collected years of examples of seemingly minor friction—a broken printer, a manual workaround in a critical process—that turned out to be bottlenecks no one in leadership knew existed. It can feel like micromanaging. It isn't. It's the only way to stay connected to the organization's actual workflows.
The second: stop having five number-one priorities. Rank the work. Not "here are our top initiatives"— actually rank them in order, one through five. And then build the culture where if you're on the critical path for project one, you're not touching project three. In Nelson's experience, this simple act alone—real prioritization, not prioritized-sounding language—frees up more flow than almost anything else an organization can do.
The capability trap is common because the path into it is paved with reasonable decisions. Deliver this quarter. Skip the long-run investment just this once. Put out this fire first.
Nelson Repenning has spent his career studying what happens next, and more importantly, what it takes to find a way out. The answer, as it turns out, isn't working harder. It's working on the right things, at the right pace, with enough slack in the system to actually see what's broken.
There's got to be a better way. Apparently, there is.
To learn more about ELB Learning’s business transformation services, click here. To watch the full Human Capital Gains podcast episode on the same subject featuring Nelson Repenning, click here.