The Nine-Box Dilemma: When the Manager’s Map Becomes the Territory

Years ago, I had what I now think of as my Nine-Box Dilemma with a manager. We were discussing a tool based on the classic nine-box talent grid, the familiar matrix that places employees somewhere between performance and potential. High performer, high potential, core player, inconsistent performer, underperformer. I remember looking at it and thinking that something was fundamentally missing.

My concern was not that performance should never be evaluated. Of course it should. My concern was that people were being transformed from complex human systems into coordinates.

I raised several questions. What happens to people whose strengths do not appear as consistent, linear output? How much does managerial bias influence where someone is placed? What does “potential” actually mean, and who gets to define it? How does a model like this account for neurodivergent employees, whose performance can be highly contextual? And most importantly: when a manager puts someone in a box, how much of that box describes the employee, and how much describes the manager’s own model of the employee?

The conversation ended much faster than the questions deserved. At the time, it was mostly an intellectual disagreement. Later, it became personal.

When You Become the Person in the Box

At one point in my career, I was evaluated as performing below expectations. Later, in a different team and under different leadership, I was evaluated as successfully meeting expectations.

I was still me. Same engineer, same ADHD, same brain, same career history. Many of the same strengths and many of the same imperfections. I did not receive a personality transplant between those two teams.

What changed was everything around me. The environment changed. The manager changed. The work changed. The relationship changed. The level of trust changed. The types of problems I was solving changed. And perhaps most importantly, the map through which my work was being interpreted changed.

That experience made my old nine-box discomfort much more interesting. If the same person can appear to be an underperformer in one system and a successful performer in another, then performance cannot simply be an intrinsic property of the person. Something more complicated is happening.

This brings me to one of the ideas from Neuro-Linguistic Programming that has become increasingly important in the way I think about engineering, careers, coaching, and leadership: the map is not the territory. As an engineer and a Society of NLP-certified NLP Master Practitioner, with certification personally signed by Richard Bandler, I find this principle particularly useful because it sounds simple until you start applying it to yourself. Managers have maps too.

The Seductive Simplicity of the Box

I understand why models such as the nine-box grid exist. Organizations are complicated. People are complicated. Executives need to make talent decisions across large groups, and a matrix gives everyone a common language. Performance on one axis, potential on another, nine boxes, done.

There is value in simplification. The danger starts when we forget that simplification happened. A metric can describe something useful. A category can help us notice a pattern. But there is an enormous difference between saying “this person is currently struggling to produce consistent outcomes in this environment” and saying “this person is an underperformer.”

The first statement describes a relationship between a person, a behavior, a period of time, and an environment. The second sounds like an identity. And identities are sticky.

Once someone becomes an “underperformer,” new observations tend to pass through that label. A delayed task confirms it. A communication problem confirms it. An unexpectedly difficult problem confirms it. A significant technical success may be recognized, but somehow it becomes an exception. The model starts selecting evidence that reinforces the model, and the map slowly becomes the territory.

The Filters We Don’t See

In NLP, we talk a lot about filters. We do not experience reality as a perfect recording of what happened. We construct internal representations of it. Experiences pass through beliefs, values, memories, language, expectations, criteria, and previous experiences. Managers do this too.

Imagine a manager whose internal map of a strong engineer contains these qualities:

• predictable delivery,
• proactive communication,
• visible participation,
• independence,
• early escalation of blockers,
• consistent output,
• strong ownership.

There is nothing inherently wrong with any of these. In fact, most are valuable. The problem begins when the manager forgets that this is a model of a good engineer, not the definition of one.

Another engineer might create enormous value differently. They might excel at ambiguous technical problems. They might discover relationships nobody else notices. They might perform exceptionally well when something is broken and nobody knows why. They might learn unfamiliar domains extraordinarily quickly. They might struggle with repetitive execution while being excellent at exploration. Their performance profile might be uneven.

As someone with ADHD, that last point matters to me personally. Neurodivergent performance can be highly contextual. Put the same brain in two different environments and you may observe dramatically different output. That does not mean accountability disappears. It means interpretation becomes more difficult.

The Meta-Level of Open-Mindedness

Here is where I think management gets really interesting. Many managers describe themselves as open-minded, people-first, curious, flexible, empathetic, supportive. I believe most of them genuinely mean it. The interesting question is not whether they are lying. The interesting question is at what level their flexibility operates.

A manager may be extremely open-minded inside their existing map. They may welcome five different implementations of a technical problem. They may encourage experimentation. They may enthusiastically accept suggestions from engineers they trust. They may give enormous autonomy to people who already fit their mental model of a high performer. That certainly demonstrates some flexibility.

But there is a deeper test. What happens when someone challenges the map itself? Not “could we solve this problem differently?” but “could the way you are defining the problem be incomplete?” Not “could I perform this task differently?” but “could your model of what good performance looks like be missing something?”

That is a very different kind of challenge. This is where managerial open-mindedness becomes a meta-skill. Being open-minded inside your map is not the same as being willing to change your map. You can be extraordinarily flexible at one level while remaining remarkably rigid at the level above it.

“I Am an Open-Minded Manager”

There is another trap here. The stronger our identity becomes, the harder that identity can be to examine.

Consider sentences like “I am a people-first manager,” “I trust my team,” “I create psychological safety,” “I am open-minded.” These sound positive, but from an NLP perspective they are also interesting because they operate close to the level of identity. And identity can become a powerful filter.

Suppose an employee says “I don’t feel trusted.” A manager who has strongly identified as someone who trusts their team may unconsciously hear something very different: “You are not the kind of manager you believe yourself to be.” Now the conversation becomes threatening. Instead of becoming curious about the employee’s experience, the manager may explain their intention. “I do trust you.” “This isn’t about trust.” “I’m only trying to create accountability.” “My goal is to help you succeed.”

All of those statements may even be true. But they miss an important distinction: intent and impact are different levels of information. You can intend to create accountability and create an experience of surveillance. You can value autonomy and behave more controlling when you become anxious about someone’s performance. You can believe deeply in psychological safety while creating an environment in which one employee no longer feels psychologically safe. You can care about people and still misunderstand one of them.

That does not make someone a hypocrite. It makes them human. The managerial problem begins when the manager’s self-image prevents them from becoming curious about the contradiction.

The Meta-Map

This is one reason my NLP training has affected the way I think about management far beyond communication techniques. For me, the interesting part is not learning a collection of techniques for “handling people.” It is developing greater awareness of the models through which we ourselves interpret people.

A good manager does not only need a leadership philosophy. They need the ability to observe their leadership philosophy itself. They need a map of the map. If you are a manager, ask yourself:

• What behaviors do I automatically associate with competence?
• What does “high potential” mean to me?
• Which personalities do I find easiest to manage?
• Do I confuse predictability with capability?
• Do I confuse communication style with commitment?
• Do I interpret someone differently after I have categorized them as a weak performer?
• Which kinds of contribution do my metrics fail to capture?
• Does this person struggle everywhere, or mainly in this particular system?

And perhaps the most important question: what evidence would make me change my current interpretation of this person?

If the answer is essentially “nothing,” then you are no longer evaluating someone. You are defending your map. And defending a map while believing that you are open-minded is, to me, one of the most interesting meta-level contradictions in leadership.

Metrics Should Be Sensors, Not Verdicts

None of this means we should abandon metrics. That conclusion would be just as simplistic as blindly following them. Organizations need ways to evaluate performance. Managers need evidence. Expectations need to be explicit. Delivery matters, reliability matters, communication matters. Metrics matter.

But metrics should behave like sensors. A sensor tells you that something deserves attention. It does not necessarily tell you why it happened.

Suppose an engineer repeatedly misses delivery expectations. That matters, but it opens a diagnostic process rather than completing one. Why is it happening? Does the work match their strengths? Are expectations ambiguous? Is this person actually struggling across contexts, or primarily under this particular configuration of work, manager, expectations, and environment?

Sometimes the answer will absolutely be that the employee needs to improve. Context should never become an excuse that removes personal responsibility. But sometimes the answer might also be that the system needs to improve, or that the manager needs to adapt. That possibility has to remain available. Otherwise we are not being open-minded. We are merely searching for evidence that confirms our first diagnosis.

Maybe Performance Is a Relationship

This leads me to a different way of thinking about performance. Instead of asking “what kind of performer is this person?” perhaps we should more often ask “under what conditions does this person perform at their best?” and “what conditions reliably reduce their performance?”

That does not remove accountability. It creates better accountability, because now we can distinguish between the things the employee needs to change and the things the system might need to change.

Perhaps performance is better understood as something closer to:

Performance = Person × Task × Environment × Manager × Time

rather than:

Person = Box 7

That model is less convenient. Unfortunately, human beings are inconvenient.

Behavioral Flexibility Goes Both Ways

One of the NLP principles I appreciate most is behavioral flexibility. The more possible responses we have available, the more effectively we can interact with a changing system.

Managers often apply this idea downward. How can I communicate differently with this employee? How can I motivate them differently? How can I coach them differently? Those are useful questions. But real managerial flexibility also requires applying it inward. Can I change how I interpret this person? Can I revise my criteria? Can I abandon a model that previously worked for me? Can I recognize that the employee I thought was the problem might perform perfectly well in a different environment? Can I acknowledge that my management style may itself be one variable in the outcome?

That is much harder than changing the wording of a feedback conversation. It requires changing the map.

Looking Back at My Nine-Box Dilemma

Today, I think about that old nine-box conversation differently. My problem was never really the nine boxes. A grid is just a grid. The deeper problem is the temptation to mistake a useful abstraction for an accurate representation of a human being.

My own career has been giving me that accidental experiment ever since. I have been documenting part of it in my Back to Engineering series: from consulting to training to engineering, then across teams working on different technology stacks and different levels of the infrastructure.

But one of the earliest signals required no career transition at all. The team did not change. The role did not change. Much of the work did not change. The manager changed, and with that came a different map through which my work was interpreted.

Later, moving to another team under different leadership gave me another data point. Again, the person had not suddenly been replaced. Yet the interpretation of that person’s performance changed significantly.

One environment produced one interpretation of my performance. Another produced a different one. Neither evaluation needs to have been completely dishonest. Neither manager needs to have been malicious. Both could have been looking at real evidence, and still the conclusions could differ. That is precisely the point.

The observer is part of the system being observed. A manager is not standing outside the employee-manager relationship measuring an independent object called “performance.” The manager is one of the variables producing the performance they are measuring. Once you recognize that, a nine-box grid starts looking very different. It can still be a tool, but it should make us ask questions rather than give us confidence that we have discovered what kind of person someone is.

The Map Is Useful Until It Isn’t

We need maps. Without them, management would be impossible. We need frameworks, expectations, performance conversations, and difficult decisions. But good managers should maintain a particular relationship with their maps. Use them. Learn from them. Challenge them. Notice what they exclude. And redraw them when reality stops fitting.

Perhaps that is the deeper form of managerial open-mindedness. Not “I am an open-minded manager,” but “I have a model of what good management and good performance look like. I know that model is incomplete. Show me where.”

There is humility in that sentence. There is curiosity in it. And perhaps most importantly, there is room for another human being. Because the map is not the territory. The employee is not the box. And sometimes the most important performance improvement a manager can make is not moving someone to another square. It is realizing that the grid itself may need to change.

---

What Does Your Map Look Like?

I have spent roughly two decades in software engineering, from hands-on development and open source to technical leadership, while increasingly exploring the other side of engineering careers: how we think, make decisions, respond to pressure, define success, and sometimes get trapped inside our own models.

Today, I bring those two worlds together through my work as a Principal Software Engineer, Career Coach, and NLP Master Practitioner. The interesting problems are often not purely technical. Sometimes a developer is trying to understand why they thrive in one environment and struggle in another. Sometimes an experienced engineer is questioning what the next stage of their career should look like. Sometimes a manager wants to understand why an approach that works extremely well with some people creates friction with others. And sometimes the most useful thing we can do is step back and examine the map itself.

If you are a software developer, engineering leader, or manager who wants to explore your career, working patterns, leadership style, or the assumptions shaping your decisions, you can book an exploratory coaching conversation with me at aykutbulgu.com. No boxes. We start with the territory.