All posts

Blog

9-Box Grid Explained: Talent Review Made Simple (With Examples)

November 7, 2026 · By Zohaib Butt

9-Box Grid Explained: Talent Review Made Simple (With Examples)

Talent review season arrives, nine boxes appear on a slide, and the room spends two hours arguing about names before defaulting to last year's placements. The 9 box grid is not complicated — running it honestly is. Done right, it is the fastest way to decide who gets development money, who gets a stretch, and who gets a straight conversation.

The stakes are real. US organizations spent $102.8 billion on employee training in 2023 — $874 per learner (Training magazine, 2023 Training Industry Report) — and much of that money should follow the placement decisions a 9 box grid produces. Placements also drive retention: 94% of employees say they would stay longer at a company that invests in their learning (LinkedIn Learning, 2018), and a fair grid plus real follow-through is exactly that investment.

This guide explains the two axes, the nine tiles and what to do about each, then shows you how to run a calibration meeting that finishes inside an hour — with examples.

Key Highlights

  • The 9 box grid maps every person on two axes: current performance and future potential.
  • It is a development tool first — the placements decide where the $874-per-learner average (Training magazine, 2023) should concentrate.
  • A typical first pass puts most people in the middle tiles: core players around 55%, stars near 10% (illustrative).
  • Calibration moves ratings: in our experience only about 55% of first-draft placements survive the meeting unchanged (illustrative).
  • Stars with a written development plan retain near 80% over 24 months versus roughly 58% without one (illustrative model).
  • 94% of employees would stay longer where employers invest in their learning (LinkedIn Learning, 2018).
  • The grid takes an afternoon; the follow-through is what pays.

What the 9 Box Grid Actually Measures

The 9 box grid sorts your people on two axes. Performance asks: how well do they deliver in the current role, against the results the role actually requires? Potential asks: how far could they grow — learning agility, aspiration, and the range to handle bigger or different work? Every person lands in one of nine tiles, three rows by three columns, from star in the top-right to underperformer in the bottom-left.

Both axes need written definitions before anyone touches a name, or the exercise collapses into opinions. We define performance with verifiable results — quota, output, error rates, delivery against commitments. We define potential with observable behaviors: how fast someone learns new scope, whether they ask for bigger problems, and whether other teams would take them.

In a typical first pass, most people land in the middle: core players around 55% of the population, stars near 10%, high-potential-but-underperforming talent around 20%, and underperformers near 15% (illustrative distribution from our client work — your mix will differ).

Hbar chart of a typical first-pass 9 box grid placement mix, illustrative: core players 55%, untapped potential 20%, underperformers 15% and stars 10%. Illustrative distribution from US mid-market client work.

Axis What it measures What it does not measure
Performance Results in the current role, verified by numbers Effort, hours, enthusiasm
Potential Learning agility, aspiration, range for bigger scope Promises, tenure, loyalty
Placement Where the two axes intersect right now A permanent label or a person's worth

Key takeaway: write one-sentence definitions for both axes before rating anyone — definitions, not opinions, make the grid fair.

The Nine Tiles and What to Do About Each

The tiles only earn their keep when each one carries an action. A placement without a follow-up move is a label, and labels are what give the grid its bad reputation. The table below is the version we use in talent reviews — print it, argue with it, then use it.

Tile Who sits here The move
Star (top-right) Delivers now, ready for more Stretch assignment and succession exposure
High performer (top-mid) Delivers now, narrower range Broaden scope; test appetite for more
Dominant specialist (top-left) Outsized results, low people or adaptability range Pay for the specialty; do not force management
Future leader (mid-right) Good results, high growth signals Development plan before promotion
Core player (center) Solid results, steady scope Keep challenged; recognize consistently
High potential (mid-left) Not delivering yet, clear ability Diagnose the blocker; coach or re-role
Underperforming risk (bot-right) Strong potential signals, weak delivery Hard conversation and a dated plan
Drifting contributor (bot-mid) Adequate output, limited growth Clarify expectations for the role
Underperformer (bot-left) Not delivering, limited range Performance process, quickly and fairly

Money should follow the tiles deliberately, not evenly. In the budgets we build, stars draw roughly $4,000 per person per year in development, high-potential talent about $2,000, core players around $1,000, and underperformers near $600 in structured support (illustrative midpoints). The uneven split is the point — the grid exists to concentrate scarce development dollars where they compound.

Bar chart of illustrative development investment per person per year by tile, using midpoints: stars $4,000, untapped potential $2,000, core players $1,000 and underperformers $600. Illustrative midpoints from US mid-market budget work.

Key takeaway: every tile gets an action, and the money follows the tiles — a grid that spreads budget evenly has told you nothing.

How to Run the Calibration Meeting

Calibration is where the grid becomes real. Managers rate their own people first, then the leadership team meets to compare, challenge and adjust — because individual ratings drift wildly without a room to normalize them. How far they drift is worth seeing; the outcome chart below makes it visible.

Preparation decides whether the meeting takes an hour or an afternoon. With no prep, a 15-person calibration runs about 3 hours; with managers pre-rating their teams, about 2 hours; with pre-ratings plus one written evidence line per rating, it lands near 1 hour (illustrative, from meetings we have run). Evidence is the whole difference — "Jordan is a star" dies in the room, while "Jordan closed the two largest accounts of the quarter and onboarded two reps" ends the argument.

Hbar chart of typical calibration meeting length by preparation method, illustrative: no prep 3.0 hours, manager pre-ratings 2.0 hours, pre-ratings plus written evidence 1.0 hour. Illustrative from client calibrations run by our team.

Agenda item Minutes The output
Agree the two axis definitions 10 One sentence each, written on the board
Challenge the extremes first 25 Stars and underperformers confirmed with evidence
Sweep the middle tiles 15 Moves and promotions flagged
Assign one action per person 10 The follow-up moves, with owners and dates

Two rules keep the meeting honest. Managers who know the work firsthand attend, and nobody is allowed to rate someone they have not observed in 90 days. Disagreements get settled by evidence, not by seniority — the loudest voice in a calibration is usually the least accurate.

Key takeaway: require one written evidence line per rating before the meeting — an hour of calibration with evidence beats three hours of adjectives.

What to Do After the Placement

The grid is a starting line, not a verdict. Within two weeks of the review, every person gets the follow-up move their tile earned: stretch assignments for stars, development plans for future leaders, dated performance conversations for the bottom-left. The placements also feed succession planning directly — ready-now successors usually come from the star and future-leader tiles, which is why our succession planning guide treats the grid as its bench-rating engine.

Follow-through is where retention is won. In our illustrative model, stars with a written development plan stay at roughly 100% → 95% → 88% → 80% across zero, 6, 12 and 24 months; stars with no plan run 100% → 88% → 72% → 58% over the same span (illustrative — modeled on the retention patterns we see in client cohorts, not survey data). The difference is not pay. It is visible growth.

Grouped bar chart of star retention over 24 months with and without a written development plan, illustrative: with a plan 100%, 95%, 88% and 80% still in seat at months 0, 6, 12 and 24, versus 100%, 88%, 72% and 58% without one. Illustrative model from client cohort patterns.

How you talk about placements matters as much as the placements. Share the development plan, not the label — people should know what they are working on next and why, without ever hearing "you are a 6B" in a hallway. For the top tiles, that conversation doubles as retention; our high-potential employees guide covers the keep-them moves in depth.

Key takeaway: placements without follow-through inside two weeks teach your best people that the exercise was theater.

The Mistakes That Make the Grid Backfire

The grid fails in predictable ways, and all of them are avoidable. Using it as a layoff tool is the fastest way to poison the exercise — once people believe the bottom-left tile is a target list, every rating gets negotiated in self-defense. Using it as a label that follows someone for years is the second: placement describes this quarter, not a person.

Watch for the subtle failures too. Halo effect inflates ratings for likable people, so evidence lines exist to catch it. Recency bias lets one great or terrible month set the whole rating, so ask about the full period.

Central tendency packs everyone into the middle tile because extremes feel rude, which is exactly why the extremes get challenged first in the meeting. And comparing people to each other rather than to the axis definitions turns a development tool into a ranking fight.

Calibration is the safety net for all of these biases. In our experience only about 55% of first-draft placements survive the meeting unchanged, roughly 30% move one tile, 10% move two or more, and 5% get withdrawn (illustrative shares from client calibrations).

Bar chart of what calibration does to first-draft 9 box grid ratings, illustrative: 55% confirmed, 30% moved one tile, 10% moved two or more tiles and 5% withdrawn. Illustrative shares from client calibrations, not survey data.

Key takeaway: the grid is a development instrument — the moment it doubles as a layoff list or a permanent label, it stops producing truth.

Using the 9 Box Grid in a Small Team

Companies with 10–50 people often skip the grid because it feels corporate. The opposite is true: the smaller the team, the more every misallocated development dollar shows. Run the same two axes across the whole company in one sitting — in a 30-person firm, that is one hour with the owner and one or two leads, not a bureaucratic season.

Scale the follow-through, not the formality. With one tiles-per-person table and one evidence line each, a small team can finish ratings, calibration and actions in a single afternoon. Then do the only hard part: actually fund the moves, from a stretch assignment for the star to a straight conversation for the bottom-left. Structural details like the grid's logic are settled practice; what varies is the discipline of the room using it.

Key takeaway: a 30-person company needs the same two axes and the same evidence rule — just fewer chairs around the table.

What Your Business Should Do

Days 1–30: Define the axes before the names.

  • Write a one-sentence performance definition tied to results people can verify.
  • Write a one-sentence potential definition covering learning agility, aspiration and range.
  • Pre-rate every person yourself before anyone else sees a grid.

Days 31–60: Calibrate with evidence, not adjectives.

  • Hold a 60-minute calibration with managers who know the work firsthand.
  • Require one observable example for every rating that managers dispute.
  • Commit to one development action per tile, not just the top-right box.

Days 61–90: Follow through where it counts.

  • Turn each star and future-leader placement into a written 90-day growth plan.
  • Pair every bottom-left placement with a specific, dated performance conversation.
  • Re-run the grid next quarter and compare movement, tile by tile.

Key takeaway: the first grid establishes a baseline; the second grid, run honestly, is where the development budget starts working.

The Bottom Line

The 9 box grid earns its keep as a money-allocation tool: two axes, nine tiles, one evidence line per rating, and development dollars that concentrate where they compound. US firms spent $102.8 billion on training in 2023 ($874 per learner, Training magazine), and 94% of employees stay longer where that investment is visible (LinkedIn Learning, 2018) — the grid decides where it lands. Define the axes, calibrate with evidence in an hour, follow through within two weeks. If you want this mapped to your own team, book a free strategy call or start with the free resource library.

Work With Zohaib Butt

1. Free Resource Library — Zohaibbutt.com/store

World-class training and coaching resources, all in one place. Download the assessments, templates and playbooks you need today.

2. Online Courses — learning.zohaibbutt.com

Self-paced leadership and coaching programs for managers and owners who want structure, not scattered tips.

3. Free Strategy Session — funnelsthatsell.com

Book your complimentary strategy call and we will map your biggest people-development bottleneck to a 90-day fix.

Free assessment pack (PDF)

Place every person with evidence, then give each tile an action.

Download the PDF pack

FAQ

What is the 9 box grid?

The 9 box grid is a talent review tool that plots every employee on two axes — current performance and future potential — producing nine tiles from star to underperformer. Each tile carries a development action: stretch assignments for stars, growth plans for high-potential talent, performance conversations for underperformers. It is a way to concentrate development money and attention where they compound, rather than a label that follows people around.

What are the axes of the 9 box grid?

Performance measures how well someone delivers in their current role against verifiable results. Potential measures how far they could grow — learning agility, aspiration and the range to handle bigger or different work. Every placement is the intersection of the two, which is why a strong performer with no appetite for more scope and a struggling learner with high range land in different tiles. Both axes need written definitions before anyone is rated.

Is the 9 box grid fair to employees?

It is fair when ratings rest on written evidence and calibrated across managers, and unfair when they rest on one manager's impressions. In our experience only about 55% of first-draft placements survive calibration unchanged (illustrative), which shows how much individual ratings drift. The fairness rules are simple: verifiable evidence per rating, managers who observed the person, and challenges settled by evidence rather than seniority.

Should you show employees their 9-box placement?

Share the development plan, not the label. Employees should leave the conversation knowing what they will work on next, what evidence shaped it, and what growth is possible — without being told they are "a 6B." Label-sharing turns the tool into a ranking people defend against, while plan-sharing turns it into a growth contract. For your top tiles, the conversation doubles as your best retention move.