Why Leaders Solve the Wrong Problem

This is the first article in The Right Problem, a four-week series on root cause analysis and problem definition. The series is built on a single observation: most failed initiatives are not failures of execution. They are failures of problem definition that execution then faithfully carried out. Each week addresses one layer of why that happens, and what to do about it.

The most successful manufacturing operation of the twentieth century did not win by solving problems faster than its competitors. Toyota won by refusing to solve a problem until it understood the cause, while competitors raced to the fix. The discipline that built one of the most efficient and enduring production systems in history was not speed. It was the willingness to be deliberately slow in exactly the place where everyone else was fast: the moment between seeing a problem and acting on it.

Most organizations have that moment backwards.

Leaders are selected, promoted, and rewarded for looking decisive. The leader who arrives at a meeting with an answer looks capable. The leader who says "I don't think we understand this yet" looks like a bottleneck. The social architecture of most workplaces punishes the exact behavior that produces accurate diagnosis.

The result is not incompetence. It is a rational response to a system of incentives pointing in the wrong direction. The research has converged on this from three directions: cognitive psychology, organizational behavior, and the study of expertise. Each one tells the same story from a different angle.

The Reward Goes to the Person with the Answer

Picture a common scene. A problem surfaces in a meeting. Someone proposes a solution in the first few minutes. The conversation moves toward that solution. The meeting ends with a plan.

That sequence feels like leadership working. Often it is not.

Daniel Kahneman, in his account of how the mind handles hard questions, describes a mechanism he calls attribute substitution. When the mind faces a genuinely difficult question, what is actually causing this problem, it quietly replaces it with an easier one: what is the most obvious explanation? It then answers the easier question while believing it has answered the hard one. The substitution happens fast, below the level of awareness, and it feels exactly like clarity.

Kahneman describes two modes of thinking. System 1 is fast, automatic, and intuitive. System 2 is slow, deliberate, and effortful. Accurate diagnosis, the kind that requires holding uncertainty long enough to examine it, is System 2 work. Most workplace problem-solving unfolds in System 1 conditions: time pressure, social visibility, and an implicit reward for sounding confident. The first plausible answer gets proposed. The meeting fills in around it. The plan gets built.

The leader who names a solution early looks decisive. The leader who says the problem is not yet well understood looks uncertain. Those social signals shape what people actually do in rooms where problems get discussed, and they shape it in a direction that consistently moves faster than accuracy.

What it looks like when it's missing

A team keeps missing deadlines, and the response is a new project management tool. Turnover rises in a department, and the response is a recognition program. A hiring pipeline slows, and the response is a refreshed job posting. Each response is aimed at the most visible feature of the problem. The symptom, not the cause.

The project management tool gets adopted and the deadlines keep slipping. The recognition program launches and people keep leaving. The job posting gets refreshed and the pipeline stays slow. The solutions were executed well. The problems were defined wrong.

What to do about it

Before a problem-solving conversation moves to solutions, name what you actually understand about the problem. Not what you plan to do about it. What you understand about it.

Give the diagnosis the same protected time you would give a solution. Most teams spend ten minutes on what is happening and forty minutes on what to do about it. The rest of this series makes the case that inverting that ratio is one of the highest-leverage changes a leader can make.

The Bias Toward Action

There is a second force at work alongside attribute substitution, and it is harder to resist because it looks like leadership.

Action feels like progress. Sitting with an incompletely defined problem feels like paralysis. When something goes wrong, the instinct is to do something. The discomfort of not-knowing gets converted into the comfort of doing, and the doing gets aimed at whatever the problem appeared to be.

Management researchers have documented this tendency across industries and organizational levels. The finding, sometimes called the bias toward action, describes a consistent preference for doing over deliberating even when deliberation would produce better outcomes. In low-stakes situations, the cost is small. Problems get addressed approximately correctly and the organization moves on. In higher-stakes situations, the cost compounds. Significant resources get aimed at symptoms while causes remain untouched.

The organizational layer amplifies this. Companies that prize responsiveness create cultures where slowing down reads as a failure to lead. The leader who convenes a task force by Friday signals that they are on it. The leader who says "I want to spend two weeks understanding what is actually happening before we respond" signals uncertainty, even when that is the more accurate read of what the situation requires.

What it looks like when it's missing

A retention problem gets a new perk before anyone has asked why people are actually leaving. A performance issue gets a coaching plan before anyone has established whether the issue is skill, clarity, or a broken process. An operations review ends in a new policy when the real problem is one step in a workflow nobody has ever mapped.

In every case, action was taken. The action was aimed somewhere else.

What to do about it

Separate the diagnosis from the response, deliberately and visibly. Name the diagnostic phase as a phase, and protect it from the pull toward solutions.

Before any problem-solving conversation, ask what you would need to know to be confident you are solving the right problem. That question does not slow the work down. It redirects energy toward the part of the work where it is actually useful.

The Competence Trap

Experienced leaders carry a deep library. Years of problems and the solutions that worked for them. Pattern recognition that fires automatically when a familiar surface presents itself.

That library is an asset. Under certain conditions, it is also what produces the wrong solution.

Barbara Levitt and James March, in their work on organizational learning, documented what they called the competence trap: the tendency to keep applying solutions that produced past success, even when the current situation is different enough that those solutions no longer fit. A problem surfaces that resembles something a leader has solved before. They reach for what worked. They do not stop long enough to examine whether the current problem is actually the same one. The familiar solution gets applied before the current problem is fully understood.

This is not carelessness. It is expertise doing what expertise does, which is to compress the distance between recognition and response. That compression is efficient when the situation matches the pattern. It is expensive when it does not.

Chris Argyris added another dimension to this. Smart, successful leaders are often the worst at catching the competence trap in themselves, because they have rarely failed and have not had to examine their own patterns. When a diagnosis starts pointing toward a decision they made, or a system they built or endorsed, what Argyris called defensive reasoning quietly steers the analysis somewhere else. The leader believes they are being rigorous. The diagnosis is actually navigating around the uncomfortable answer.

What it looks like when it's missing

The manager who responds to every engagement problem with team-building, because team-building worked once and engagement problems are now team problems by definition. The senior leader who restructures whenever performance drops, because restructuring is the tool they know. The HR leader who adds process whenever a people problem surfaces, even when the existing process is part of what is causing the problem.

None of these people are incompetent. They are competent in a way that is working against them.

What to do about it

Before proposing a solution, write down the problem being solved. Not the symptom. The problem. If you cannot state clearly what is happening that should not be, who it is affecting, and what resolved would actually look like, the solution is not ready yet.

A second check: ask whether the solution being considered is fresh to this situation or familiar from a previous one. Familiarity is not disqualifying. But it is a signal to pause and examine whether the pattern match is real before applying the playbook.

What Toyota Figured Out

The LEAN tradition, and Toyota's version of it specifically, is the longest and deepest body of practice on exactly this problem.

Toyota's production system was built on a discipline that runs directly against attribute substitution, action bias, and the competence trap. Before any response was proposed to any problem, the problem had to be understood. Not described. Understood. The distinction mattered enough that it was built into the culture as a named practice.

Genchi genbutsu translates roughly as go and see the actual situation. The instruction is to leave the conference room, go to where the work actually happens, and observe reality before forming a hypothesis. Not what was reported. Not what someone remembers. What is actually there.

Taiichi Ohno, the Toyota engineer most associated with developing this approach, was known for drawing a chalk circle on the factory floor and instructing managers to stand inside it for hours before saying anything about what they thought was wrong. The deliberateness was the point. Most wrong diagnoses, in Toyota's view, came from solving a problem as it was reported rather than as it actually existed. The conference room version of a problem and the real version are often different things.

Toyota's core insight was that the moment most organizations move through fastest, the gap between recognizing a problem and acting on it, is exactly where care is most valuable. Speed is useful after a problem is understood. Before it is understood, speed is how you build the wrong solution well.

That reframe carries the whole series.

Why This Opens the Series

Attribute substitution, action bias, and the competence trap are different names for the same failure mode. They all describe the same shortcut: the movement from "there is a problem" to "here is what we should do" without enough time in the space between.

The discipline this series teaches is the willingness to stay in that space longer than feels comfortable. To ask what is actually happening before deciding what to do about it. To resist the social pressure that rewards confident answers over accurate ones.

This is not a soft preference. The research is consistent across disciplines: problem-solving quality is determined more by how a problem is defined than by the quality of the solution that follows. A well-defined problem generates useful solutions. A loosely defined problem generates effort aimed somewhere else.

The leaders who do this well are not slower than the leaders who do not. They are slower at the front end and faster overall, because they spend less time executing solutions to the wrong problems.

A short reflection before you go

Think about a problem your team is currently working on. A real one, not an abstraction.

Can you state in writing what is happening that should not be?

Can you name who is specifically affected, and what it is costing?

Can you describe what resolved would actually look like?

Next week

Week 2 goes deeper on the most widely misunderstood tool in LEAN thinking: the 5 Whys. The method is simple. The discipline is harder. And the place most people stop asking is exactly where the diagnosis is just beginning to get useful.

References

Argyris, C. (1991). Teaching smart people how to learn. Harvard Business Review, 69(3), 99-109.

Kahneman, D. (2011). Thinking, fast and slow. Farrar, Straus and Giroux.

Levitt, B., & March, J. G. (1988). Organizational learning. Annual Review of Sociology, 14, 319-340.

Liker, J. K. (2004). The Toyota way. McGraw-Hill.

Ohno, T. (1988). Toyota production system: Beyond large-scale production. Productivity Press.