Tools · People analytics
Nine Box
Describe a team — get a 9-box talent grid with per-person actions.
The method
Nine-box performance–potential grid
The talent review is next week and names are going into boxes. Half the room believes the grid; the other half knows the placements say as much about who rated whom as about the people rated. The meeting that should allocate development investment turns into an argument about whose 'high potential' means anything.
The nine-box crosses two judgments — current performance and future potential — and its defenders are candid that its value is procedural, not psychometric. Mark Bussin's Remuneration and Talent Management presents it as the working tool for talent identification precisely because it forces structure and shared criteria onto conversations that otherwise run on impression; his accompanying claims are about honesty in use — high-potential status should be communicated transparently, and it is neither a promotion promise nor a permanent label. Fitz-enz and Mattox's Predictive Analytics for Human Resources shows the grid in its analytical role: nine-box placements captured at intervals become the performance-and-potential ratings that feed retention and productivity models — which is a reminder that a box is a data point made of manager judgment, and inherits all of that judgment's error.
The measurement literature is where the honest critique lives. Schmitt and Borman's Personnel Selection in Organizations treats job performance as a construct with structure — task performance and contextual performance are distinct — and treats criterion measurement as a science with known failure modes. A single performance axis compresses that structure into one number, and 'potential' is a prediction, which the selection literature insists should rest on validated predictors rather than adjectives. The practical conclusion isn't to abandon the grid; it's to refuse the ritual version of it. Box placements are hypotheses for calibration — a first pass that earns its keep only if the meeting interrogates the ratings behind it, watches for the familiar biases, and ends in a per-person action rather than a label.
The books hand you the grid and the warnings; here you describe the team and get the first-pass placements with the per-box action and the calibration notes attached — and it places only the people you actually described, inventing no one.
The books behind this tool
- Remuneration and Talent Managementon Compensation Professional — Mark Bussin
- Predictive Analytics for Human Resources — Jac Fitz-enz & John R. Mattox II
- Personnel Selection in Organizations — Neal Schmitt & Walter C. Borman
How it works
Corpus-grounded (people-analytics cluster). Places each named person by performance × potential, gives the box label + action and a per-box talent strategy, with calibration notes to guard against bias. Places only people described in the input.
You bring
{ team, cluster? }
You get
{ team_summary, placements[]{name, performance, potential, box, action}, box_summary[]{box, who[], strategy}, calibration_notes[], riskiest_assumptions[], grounded_in, provenance }
Use it for
- →PA-guide reader: prep a talent review with a first-pass 9-box
- →Get the action per box (stretch/develop/retain/coach/exit)
- →Surface where calibration is needed before the meeting
See it work
example outputTeam: a 6-person product-engineering team described with brief performance and growth notes per person, for a talent-review prep.
9-Box Talent Grid — Product Engineering (6 people)
A first-pass placement by performance × potential to prep the calibration meeting. Ratings are hypotheses, not verdicts.
Placements
| Person | Performance | Potential | Box | Action |
|---|---|---|---|---|
| Priya | High | High | Star | Stretch — give scope beyond the role; retention risk |
| Marcus | High | Medium | High Performer | Retain — reward; deepen mastery |
| Dana | Medium | High | High Potential | Develop — coach + stretch assignment |
| Liam | High | Low | Trusted Professional | Retain — protect; lean on as anchor |
| Sofia | Medium | Medium | Core Player | Develop — targeted skill growth |
| Theo | Low | Medium | Inconsistent Player | Coach — diagnose blockers; 90-day plan |
Box summary
- Star — Priya. Strategy: invest first; she's the flight risk that hurts most. Give visible, ambiguous work.
- High Performer / Trusted Professional — Marcus, Liam. Strategy: the reliable core; reward and retain, don't over-rotate on "potential."
- High Potential — Dana. Strategy: a stretch project with air cover; the highest-leverage development bet.
- Core Player — Sofia. Strategy: concrete skill goals to push toward high-potential.
- Inconsistent Player — Theo. Strategy: is it skill, role-fit, or context? Decide within a quarter.
Calibration notes
- Guard against recency bias — Priya's "high" leans on one recent launch; check the full cycle.
- Potential ≠ tenure. Liam's "low potential" may reflect contentment, not ceiling — confirm with him.
- Single-rater input; have a second manager rate independently before the meeting.
Riskiest assumptions
- That "potential" is read consistently across raters (define it before calibrating).
- That Theo's low performance is the person, not the assignment.
Only people described in the input are placed; empty boxes are left empty rather than filled.
Run it now
Build a 9-box talent grid
Place your team on the performance × potential grid — each person's box + the recommended action, a per-box talent strategy, and calibration notes to guard against bias.
Prefer code? Call it over the API or hand it to your AI agent via MCP — POST /api/bicycle/nine-box · build_nine_box. API & agent access →
Related questions we answer
- How do I measure employee performance at work?Separate the signal you want — the person's actual contribution — from the noise that dominates raw ratings, chiefly idiosyncratic rater variance, which often explains more of a score than the ratee does. Define the construct before the scale, use behaviorally anchored items, calibrate raters, and report reliability. Measure conditions and outputs, not personalities.
- How mature is our people analytics program?Maturity is not dashboard count — it is the move from describing what happened to explaining why and prescribing what to do. Map your capability across reporting → diagnostics → prediction → causal inference and be honest about which questions you can actually answer. Most 'programs' are reporting with a BI tool.
- What is the difference between people analytics and HR reporting?HR reporting describes what happened — headcount, turnover rate, time-to-fill. People analytics applies behavioral science and causal methods to explain why and prescribe action. It is a different epistemology, not a prettier chart: reporting counts, analytics infers.
- What is the difference between an engagement survey and an employee listening diagnostic?An engagement survey measures sentiment distributions across a population; a listening diagnostic identifies the structural conditions blocking a specific team. They answer different questions — 'how do people feel overall?' versus 'what is binding here, and what is the one action?' — and require different methodology and sampling.
- Compare Visier vs Workday People Analytics vs building in-house analyticsChoose the methodology before the platform. Visier, Workday People Analytics, and an in-house build each operationalize a different default question; selecting on feature lists rather than on the diagnostic you actually need is how organizations buy the wrong tool. Decide what you must measure and how, then pick what serves it.
- What are the alternatives to annual employee engagement surveys?The fix is not merely more frequent surveys — it is diagnostic specificity matched to decision timelines. Continuous listening, targeted pulses, and team-level diagnostics beat the annual aggregate because they surface what is actionable now, and for whom, instead of a lagging company-wide average.