The 9-Box Grid Problem: Why Most Organisations Use It Wrong and What to Do Instead

Picture a room full of senior managers, sitting around a table with a 3×3 grid on the screen in front of them.

Someone reads out a name. The discussion begins. One manager says the person is “high potential, definitely.” Another quietly disagrees but does not push back because the first manager is more senior. A third offers a story from three months ago as evidence of something. Thirty seconds later, a dot appears somewhere on the grid.

That dot will follow this person for the next twelve months, possibly longer. It will influence whether they are considered for a stretch assignment, whether they make a shortlist for a leadership programme, and whether the organisation invests meaningfully in their development or quietly lets them plateau.

And nobody in the room has agreed on what “potential” actually means.

This is not a theoretical concern. It is the most common version of how the 9-box talent review happens in Indian organisations. And it is worth being honest about what it is actually producing.

What the 9-Box Was Designed to Do

The 9-box grid was developed by McKinsey for General Electric in the 1970s as a portfolio management tool for evaluating business units. It was later adapted as a talent management tool, placing individuals on two axes: current performance and future potential.

The idea, in principle, is sound. Organisations have limited resources for development, and they need a way to think about where those resources will generate the most return. The grid was meant to be a conversation starter, a way of making the distribution of talent visible so that decisions about investment, succession, and development could be made more deliberately.

The problem is not the grid itself. The problem is the assumptions that have quietly accumulated around the way it gets used, until the original purpose has been almost entirely lost.

The Undefined Variable That Runs Everything

When a manager places someone in the “high performance, high potential” box, the performance dimension is usually straightforward enough. There are numbers. There are outcomes. There are deliverables that either happened or did not.

The potential dimension is a different matter entirely.

Most organisations that use the 9-box have never formally defined what potential means in their context. They have left each manager to interpret it using their own internal model, which is shaped by their experience, their biases, their personal definitions of what leadership looks like, and, often, by how much they like the person being discussed.

In some organisations, potential means “this person reminds me of someone who succeeded.” In others, it means “this person speaks confidently in meetings.” In others still, it means “I think this person could do my job someday,” which is a projection that has more to do with the manager’s own self-image than with the person being assessed.

The result is a calibration session where the numbers on the grid look precise but the underlying judgements are not comparable. A “high potential” rating from one manager means something entirely different from a “high potential” rating from another. Plotting them on the same grid does not resolve that inconsistency. It just makes it invisible.

Research from McKinsey on talent management consistently shows that organisations struggle to define potential in operational terms, and that this gap is one of the most significant barriers to effective succession planning. The grid is not the issue. The absence of a shared definition underneath it is.

When the Talent Review Becomes a Political Document

Here is what happens in the calibration session once the discussion of each name has run its course.

The managers with the most organisational authority tend to win the disagreements. The managers who advocate loudest for their people tend to win regardless of the evidence they bring. And the managers who are newer to the organisation, or more junior, or simply less assertive, tend to see their people rated more conservatively than the evidence would support.

This is not a character flaw in any of these managers. It is what happens when a subjective judgement is made in a social context where power is not evenly distributed. The grid becomes a record of organisational politics dressed in the language of talent management.

There is a related problem that sits one step upstream. Managers in most organisations know, going into the calibration session, roughly where they want their people to land. They have already decided who they want to nominate for development programmes, who they want to protect from being poached by another business unit, and who they are managing out without yet saying so directly. The calibration session often confirms these pre-existing decisions rather than challenging them.

This is a significant problem if you believe, as the evidence suggests, that potential is not as evenly distributed as managerial intuition tends to assume, and that the people being overlooked in your calibration sessions are not all genuinely low-potential. Some of them are simply less visible to the people doing the assessing.

The Label That Does Not Come Off

There is a dimension to the 9-box problem that rarely gets named in the talent management conversation, but that practitioners working inside organisations know well.

Once a person has been placed in a box, it is very hard for them to move.

If someone receives a “medium potential” or “low potential” rating in one cycle, that rating shapes what they are offered in the following twelve months. They are less likely to be nominated for development programmes. They are less likely to be considered for stretch assignments. They may not even be aware of their rating, but the downstream effects are real.

The result is a self-fulfilling system. The people rated as low potential receive less investment in their development, which means they are less likely to demonstrate the kind of growth that would prompt a re-rating. Twelve months later, the same managers look at the same person and confirm the original rating, this time with a quiet sense of “we were right about this one.”

The high-performer attrition research consistently shows that employees who are not visibly seen as “high potential” by their organisation are significantly more likely to disengage and eventually leave, regardless of their actual capability. The label becomes a signal the person eventually receives and responds to, even when it was never formally communicated.

What Good Actually Looks Like

Replacing the 9-box grid is not the answer. Replacing the assumptions that surround it is.

The organisations that use it well share a few specific practices. None of them are complicated. All of them require more rigour than most organisations currently apply.

Define potential before you begin rating it. Potential is not a feeling. It is a set of behaviours and capabilities that can be observed, assessed, and discussed with specificity. A useful definition of potential for most organisations includes something like: the capacity to absorb increasing complexity, to learn and adapt quickly in unfamiliar situations, to influence others effectively, and to sustain performance under pressure. These are things that can be assessed. They are not the same as “I have a good feeling about this person.”

A properly designed competency framework gives you the vocabulary to make the potential dimension discussable. Without that vocabulary, the calibration session will always default to instinct, and instinct varies too much across a room of managers to produce consistent results.

Separate the evidence from the judgement. The most useful change any organisation can make to its calibration process is to require that every potential rating be accompanied by observable behavioural evidence, not examples from the last three months and not character observations. What did this person actually do in a situation that required the behaviours you associate with potential? If the manager cannot answer that question with specificity, the rating should be held until they can.

Use structured assessment to validate, not just instinct to rate. The gap between what managers perceive about a person’s potential and what a structured behavioural assessment actually reveals is often significant. Not because managers are poor judges of people, but because they see a limited slice of a person’s behaviour in a limited range of situations. Assessment surfaces behaviours that the day-to-day work relationship does not.

The Assessment Development Centre as a Calibration Tool

One of the most practical shifts an organisation can make is to use Assessment Development Centre data as the evidence layer underneath the 9-box grid.

An Assessment Development Centre evaluates individuals against a defined set of competencies using multiple exercises and multiple assessors. The output is not one manager’s impression of one person in one context. It is a multi-dimensional picture of how that person actually performs across a range of situations that reflect the demands of the role they are being assessed against.

When an organisation runs an ADC before the calibration session, the conversation in the room changes. Instead of “I think Priya is high potential,” the manager says “Priya’s ADC data shows strong performance on learning agility and stakeholder influence, but a development gap in managing ambiguity at senior level.” The discussion is grounded. It is comparable across different managers’ teams. And the development plan that follows is connected to real evidence rather than to a box on a grid.

The data-driven approach to talent decisions does not make the calibration session redundant. It makes it useful. The conversation shifts from debating where to place someone to discussing what they need in order to grow.

When the Grid Connects to Development, It Earns Its Place

The most common failure mode after a talent calibration session is the file that nobody opens until the following year.

Dots are placed on the grid. A report is produced. Slides are shared with the CHRO and possibly the CEO. And then the process quietly stops until the next annual cycle, while the employees who were rated “high potential” continue to do the same work they were doing before, without a development plan that reflects their rating, a stretch assignment that builds the capability identified as critical, or even a conversation with their manager about what the organisation sees in them.

This is where the 9-box grid loses most of its practical value. The rating is only meaningful if it changes something. If the “high potential” individuals receive the same development as everyone else, if the “medium potential” individuals are not given a clear pathway to re-rate, and if the “low potential” individuals are neither supported with targeted development nor having an honest conversation about their fit, the grid is an annual exercise in documentation rather than a talent management tool.

The coaching and development conversation that should follow a talent review is not a bonus. It is the point. Without it, the review is a conversation about people that the people themselves are excluded from, and that produces no change in what those people experience in their day-to-day work.

SHRM’s guidance on succession planning tools notes that the 9-box grid is most effective when it is treated as a starting point for structured development conversations, not as a definitive categorisation of an employee’s ceiling. The distinction matters, and most organisations are using it as the latter.