A 9-box talent review plots every employee on two axes — performance and potential — then sorts them into nine boxes to decide who gets a stretch assignment, a development plan, or a performance improvement plan. The process runs in five steps: define performance and potential criteria, score each employee independently, plot the grid, calibrate scores across managers in a joint session, and turn each box into a specific action. Skip the calibration step and the grid turns into a popularity contest — the most common reason 9-box reviews lose credibility in 2026.
TL;DR
- A 9-box talent review plots performance against potential across nine boxes.
- Calibration sessions across managers decide whether the grid is trustworthy or biased.
- Potential scores fail most often when they run on gut feel instead of behavioral evidence.
- Each of the nine boxes needs its own action: promote, develop, coach, or manage out.
- Most companies run the review once or twice a year, tied to the performance cycle.
Why this matters
A 9-box talent review is only as good as the data behind the potential axis. Performance is usually backed by numbers — quota, output, project delivery. Potential is guesswork unless someone builds a rubric for it, which is why so many 9-box grids collapse into a ranking of who managers personally like.
Get the process right and the grid becomes the backbone of succession planning, promotion decisions, and development budgets for the year. For a deeper walkthrough of how the grid feeds succession pipelines, see this guide on the 9-box grid for succession planning.
How to run a 9-box talent review
Run the review in this order every cycle. Skipping steps — especially calibration — is what breaks the output.
- Define performance criteria first. Write down what "high performance" means for each role before anyone scores anyone. Vague criteria produce vague grids.
- Define potential criteria separately. Potential is not "has been here a long time" or "I like working with them." Tie it to readiness for more scope, complexity, or people leadership within the next 1-3 years.
- Score independently. Each manager scores their own direct reports on both axes before any group discussion happens. This surfaces individual bias instead of hiding it.
- Plot the grid. Place each employee in one of the nine boxes based on the two scores. Do this visually — a spreadsheet grid or a whiteboard works fine.
- Calibrate in a group session. Managers compare notes across teams. This is where a "high performer" in one department gets checked against the bar in another. Calibration is the single step that separates a useful 9-box review from an exercise in favoritism.
- Build an action plan per box. Every employee leaves the process with a next step — stretch project, coaching plan, retention conversation, or performance plan. A grid with no follow-up action is a wasted afternoon.
The 9 boxes, explained
Each box combines a performance level with a potential level. The labels below are the standard version used across most 9-box frameworks.
| Box | Performance | Potential | Typical action |
|---|---|---|---|
| Future leader | High | High | Fast-track, stretch assignment, succession pipeline |
| High potential | Medium | High | Mentor, expose to new scope |
| Rough diamond | Low | High | Diagnose the blocker — role fit, skills gap, or manager mismatch |
| High performer | High | Medium | Reward and retain, watch for flight risk |
| Core player | Medium | Medium | Steady development, most of the workforce lives here |
| Inconsistent | Low | Medium | Structured coaching plan with a review date |
| Trusted professional | High | Low | Recognize and keep in current role |
| Effective | Medium | Low | Maintain, low-cost development only |
| Underperformer | Low | Low | Formal performance improvement plan or exit |
The core player box holds the most people on any grid and gets the least attention, since it's not flagged high-potential or high-risk. That's usually a mistake — see the closing section below.
Why 9-box results vary across teams
Same framework, same company, wildly different grids from department to department. The usual causes:
- No shared performance rubric — one manager's "exceeds expectations" is another's "meets expectations."
- Potential scored on gut feel, tenure, or likability instead of documented behavior or values alignment.
- No calibration meeting, so each manager's biases go straight onto the grid unchecked.
Turn this into a candidate assessment
Build a culture-fit assessment that compares values, work style, personality, and culture profile signals before the interview.
Create a culture fit assessment- One rater per employee with no second data point to cross-check the score.
- Performance judged off a single incident — a bad quarter or one missed deadline — instead of a full review period.
- Potential criteria weighted toward "like me" traits rather than traits the next role actually requires.
Culture fit and personality assessment platforms give the potential axis something concrete to argue about instead of an opinion. MyCulture.ai's personality assessment for internal promotion decisions turns values alignment and work-style data into a documented input for the potential score, so calibration conversations start from evidence instead of impressions. A culture fit assessment platform doesn't replace manager judgment on a 9-box grid — it gives that judgment a second data point to check against.
Score potential with data, not gut feel
Add a documented values and work-style fit score to every calibration conversation.
Is a 9-box grid the same as a talent matrix?
A 9-box grid is one specific version of a talent matrix — the 3x3 layout that plots performance against potential. Other talent matrices use different axes, like readiness and flight risk, but the nine-box format remains the most widely used structure going into 2026 because it's simple enough for any manager to read in one glance.
How often should a company run a 9-box talent review?
Most organizations run a 9-box talent review once or twice a year, tied to the performance review cycle. Running it more often without new performance data just recreates the same grid with recency bias baked in, since nothing substantive has changed between cycles.
What's the difference between performance and potential in a 9-box grid?
5 minutes
to create your first hiring assessment
Use the assessment landing page to choose the right modules and see what the candidate report looks like.
See the assessment builderPerformance measures what someone has already delivered against role expectations; potential measures the capacity to take on bigger scope over the next 1-3 years. Treating tenure, confidence, or likability as a stand-in for potential is the fastest way to end up with a grid nobody on the leadership team trusts.
One last thing
The box everyone argues about is the bottom-left underperformer. The box that actually costs companies talent is the core player square in the middle — the majority of any workforce sits there, and it's the box most likely to get skipped during action planning because nobody's flagged as urgent. Underinvest in core players for two review cycles in a row and they're the ones a competitor picks off first, not the flight risks in the high-potential corner everyone's already watching closely.