9-Box Grid: Talent Review Template, Examples and Guide

The 9-box grid is a talent-review tool that places employees on a three-by-three matrix using two dimensions: current performance and future potential. Leaders use it to discuss development, succession, retention, role fit, and workforce risk.

The grid is useful only when the organization defines both axes, requires evidence, and treats placement as a discussion rather than a permanent label. Performance can usually be supported with goals, outcomes, behaviors, and review evidence. Potential is harder, because it is a judgment about future scope, learning agility, aspiration, and the opportunities the organization can actually offer.

This guide provides the grid template, explains all nine boxes with their common names, works through a calibration example, and shows how to connect the tool with performance management, succession planning, and employee development.

What Is a 9-Box Grid?

The grid plots performance against potential, each on a three-point scale, producing nine placements. Its value comes from separating two signals that a single performance rating collapses into one number. An employee can deliver excellent results in their current role without wanting or being ready for a larger one, and a recent hire may show clear capacity while still ramping up. A single rating hides both cases. The grid makes them visible.

Where the 9-Box Grid Came From

The framework did not begin in HR. In the early 1970s, McKinsey developed the nine-box matrix for General Electric as a way to prioritize investment across GE's large portfolio of business units, plotting industry attractiveness against competitive strength. It followed the Boston Consulting Group's growth-share matrix and became known as the GE-McKinsey nine-box matrix.

HR later borrowed the structure, substituting current performance for competitive strength and future potential for market attractiveness. The adaptation spread widely in corporate talent reviews from the 1990s onward. Knowing this history is practically useful, not just trivia: the tool was designed to allocate finite investment across assets, which is why it works reasonably well for prioritizing development spend and works badly when treated as an objective assessment of a person's worth.

The 9-Box Grid Template

The layout below is the standard arrangement, with potential on the vertical axis and performance on the horizontal. Box names vary between organizations, and the most common labels are shown here alongside the discussion each placement should prompt.

A caution on labels. Names such as Talent Risk, Iceberg, Bad Hire, or Deadwood circulate widely in published templates. They are memorable, which is exactly the problem. A label attached to a named employee in a document that may be retained, discovered, or forwarded creates real employee-relations and legal exposure, and it changes how reviewers think before evidence is even discussed. Many organizations use neutral coordinates, such as box 3B, precisely to avoid this. If your grid uses names, choose ones you would be comfortable reading aloud to the employee.

The Nine Boxes and What to Do About Each

The two corners get most of the attention, but the middle box deserves more than it usually receives. Core Players are typically the largest group and carry most of the operational load. Talent reviews that spend their entire agenda on the top-right corner tend to under-invest in the group that actually delivers the work, and those employees notice.

How Should Performance Be Defined?

Use evidence from a consistent time period and role context. Relevant inputs may include:

• Goals and business outcomes from goal management
• Role expectations and competencies
• Review evidence from performance cycles
• Feedback, quality, customer, or delivery evidence
• Context such as role changes, resources, leave, or newly assigned responsibilities

A single recent success or failure should not determine placement. Recency bias is the most common distortion in these discussions, and it is worth naming explicitly at the start of the session. Use performance calibration to compare evidence and standards across managers, and draw on a consistent set of performance metrics so that comparison is possible at all.

How Should Potential Be Defined?

Potential is not an objective fact. Define it for a specific future demand, time horizon, and type of growth. An employee may have high potential for technical leadership but not want a people-management role.

Possible evidence includes:

• Learning agility and response to unfamiliar work
• Ability to handle greater complexity or scope
• Transferable skills and judgment
• Aspiration and willingness to pursue the path
• Performance in stretch assignments
• Readiness gaps identified through competency management and a current skills matrix

Do not infer potential from confidence, visibility, similarity to current leaders, or willingness to relocate unless those factors are genuinely relevant and applied consistently. One practical discipline helps more than any other: require the reviewer to state which future role or scope the potential rating refers to. Potential in the abstract invites bias. Potential for a defined next role can be discussed with evidence.

A Worked Calibration Example

A manager brings six proposed placements to a calibration session. Two are placed as Star, one as Core Player, one as Trusted Professional, one as Enigma, and one as Talent Risk.

On its own the distribution looks reasonable. The session becomes useful when a peer manager presents a similarly sized team with nobody above Core Player. Three explanations are possible: the first team genuinely holds stronger talent, the first manager rates more generously, or the second manager is unfamiliar with the evidence standard. Only the discussion can distinguish them.

The facilitator asks the first manager to defend one Star placement with a specific example of work at a broader scope than the current role. The answer is that the employee ran a cross-functional project and handled an escalation well. The peer manager then notes that two of their own people did comparable work but were not credited because that work is routine in their function. This is the exact inconsistency the session exists to surface, and it changes the standard for both teams rather than punishing either manager.

The Enigma placement is examined next. The employee's results are below expectation, but they changed managers twice in eight months and took on a role that was scoped differently from the job they accepted. The group records the placement as provisional with insufficient evidence, sets a reassessment date, and assigns a role-clarity conversation instead of a development plan aimed at a problem nobody has confirmed exists.

How to Run a 9-Box Talent Review

1. Define the purpose. Clarify whether the review supports succession, development, retention, mobility, or workforce risk.
2. Define both axes. Publish the performance period, potential definition, evidence, and rating scale.
3. Prepare manager proposals. Managers should submit a placement with evidence, not only a label.
4. Review employees in comparable context. Avoid comparing unrelated jobs as if performance evidence were identical.
5. Challenge assumptions. Ask what evidence would change the placement and whether similar behavior is being interpreted consistently.
6. Document rationale and uncertainty. "Insufficient evidence" is more accurate than forcing a confident potential judgment.
7. Create actions. Connect the discussion with an individual development plan, coaching, learning, stretch work, succession actions, or role support.
8. Review later. Placements should change when performance, aspiration, role context, or evidence changes.

How Often Should the Review Run?

Most organizations run a full talent review annually, often following the main performance cycle so that evidence is current. A lighter mid-year check is useful for succession-critical roles and for anyone whose placement was recorded as provisional. Running the full exercise more than twice a year rarely produces new information and tends to reduce the seriousness managers bring to it.

Should You Force a Distribution?

Some organizations require a target spread, such as a fixed share in the top boxes. The argument for it is that unconstrained ratings drift upward and calibration becomes meaningless. The argument against is that a forced distribution guarantees somebody is placed low regardless of actual performance, which damages trust and produces placements that cannot be defended with evidence.

A middle position works for most organizations: publish the expected shape as a guideline, require an explanation when a team's distribution departs sharply from it, and do not enforce quotas at the individual level. The purpose is to prompt a question, not to hit a number.

The Small-Population Problem

Nine boxes need enough people to be meaningful. A team of six spread across nine cells produces mostly empty boxes and differences that reflect noise rather than signal. Below roughly twenty employees in a comparable population, calibrate at a higher organizational level, or use a simpler four-box version, rather than pretending the grid is telling you something it cannot.

Bias and Fairness Safeguards

• Use written definitions and behavior-based evidence.
• Train reviewers on common rating and potential biases.
• Review demographic and organizational patterns where lawful and appropriate.
• Give reviewers access to comparable evidence, not reputation alone.
• Separate career aspiration from assumed ambition.
• Record disagreements and evidence gaps.
• Use more than one perspective for high-stakes decisions.
• Do not describe the grid as bias-free or legally definitive.

Feedback from 360-degree reviews can provide additional perspective, but it also requires careful design and interpretation. Reviewer training matters as much as the framework itself, which is why manager enablement tends to determine whether a talent review produces fair outcomes or simply formalizes existing impressions. The broader calibration process guide covers the facilitation practices in more depth.

Should Employees Be Told Their Placement?

Practice genuinely differs here, and anyone who tells you there is one correct answer is overselling. Most organizations do not disclose the box itself, for three defensible reasons: the placement is a provisional group judgment rather than a finalized assessment, potential ratings carry more uncertainty than employees typically hear in them, and a label communicated without context is more likely to demotivate than to develop.

The stronger argument on the other side is that undisclosed talent decisions quietly shape careers, and employees identified as high-potential often behave differently once they know. Withholding it can also make the process feel like a secret list, which corrodes trust when it eventually leaks, as these things tend to.

The workable compromise most organizations land on is to communicate the substance without the coordinates. Tell the employee what evidence was discussed, what the organization sees as their strengths and readiness gaps, what development is being offered and why, and what would change the assessment. That conversation delivers everything useful about the placement while avoiding a label that was never designed to be a verdict. Whatever you decide, decide it before the session rather than improvising afterward, and apply it consistently.

What to Do After Placement

The grid creates value only when it changes follow-through. Possible actions include:

• Development goals and experiences
• Coaching or mentoring
• Learning pathways
• Stretch assignments or rotations
• Succession readiness actions
• Retention and career conversations, particularly where retention risk was raised in the session
• Role clarity or performance support
• A separate performance improvement process when formally required

A 9-box placement should not replace a promotion process, pay decision, performance review, or documented corrective process. It also should not be the only input to a succession slate. The grid identifies candidates worth developing; readiness for a specific role is a separate assessment against that role's requirements.

When Not to Use a 9-Box Grid

• When potential cannot be defined. If nobody can state what future role the rating refers to, the axis will collect impressions instead of evidence.
• For very small or non-comparable populations. See the small-population problem above.
• As an input to redundancy or termination selection. Placements are provisional judgments recorded for development purposes and are a poor evidentiary basis for those decisions.
• When the organization has no development capacity. Identifying high-potential employees and then offering nothing creates expectations you cannot meet, which is worse than not asking.
• As a substitute for career conversations. The grid tells you what leaders think. It does not tell you what the employee wants, and talent strategy built without that input tends to misallocate development spend.

9-Box Grid Software Checklist

• Configurable performance and potential definitions
• Evidence, comments, rationale, and uncertainty fields
• Role-based access for sensitive talent data
• Calibration and approval workflow
• History of placements and decisions
• Connections with employee, job, performance, and development data from Core HR
• Succession and readiness reporting
• Development actions and review dates
• Appropriate security controls
• Required integrations and implementation support

Use reporting and analytics to review patterns, but do not treat the grid as a predictive model of an employee's future success.

Common Mistakes

• Leaving "potential" undefined, or defining it without naming a future role.
• Using reputation or confidence instead of evidence.
• Forcing every employee into a box despite insufficient evidence.
• Assuming high performance automatically means high potential.
• Spending the entire session on the top-right corner.
• Applying the grid to a population too small to support nine categories.
• Using stigmatizing box labels in retained documents.
• Ignoring employee career interests.
• Using the grid as a permanent label.
• Discussing placement without creating development actions.
• Using one grid placement as the sole basis for promotion, pay, or termination.

Final Takeaway

The 9-box grid can structure a useful talent conversation, but it does not make talent decisions objective. It began as a tool for allocating investment across business units, and it remains best suited to that purpose: deciding where to direct finite development effort. Define performance and potential, require evidence, calibrate judgments, document uncertainty, protect access, and create specific follow-up actions. The quality of the review depends on the discipline around the grid, not the nine boxes themselves.