The talent management toolkit: 9-box, talent reviews, succession depth charts, and build/buy/borrow

60-Second Summary

The 9-box grid plots performance × potential into 9 cells. It's the most-used talent tool in the world — and the most-misused. Use it as a calibration conversation starter, not a labelling exercise.
Talent reviews are the quarterly or semi-annual meeting where leaders calibrate talent across teams. The mechanics (data prep, facilitator, anti-bias scripts, action commitments) matter more than the framework.
Succession depth charts list 1–3 successors per critical role at three readiness levels (Ready now, Ready 1–2 yrs, Ready 3+ yrs). Cover the top 20–50 roles, not every role.
Build/Buy/Borrow is the talent supply decision: develop internally (build), hire externally (buy), or use contractors/contingent (borrow). Most companies default to Buy when Build would be cheaper and Borrow when Buy would be safer.
These four together = a complete operating system for managing talent at the company level — not just individuals.

Talent management is the set of practices that make sure the right people are in the right roles at the right time — and that there's a credible successor when each of them leaves. The four frameworks below are the canon. They are not optional for HR teams above ~200 employees.

What talent management actually is

Performance management = how an individual is doing this quarter. Talent management = how the whole portfolio of people is positioned for the next 1–3 years. The two are linked but distinct. The four frameworks here all operate at the portfolio level.

1. The 9-box grid (and its critics)

Developed at McKinsey for GE in the 1970s. A 3×3 matrix plotting performance (low / medium / high) on the X-axis against potential (low / medium / high) on the Y-axis. Each employee lands in one of 9 cells.

	Low performance	Medium performance	High performance
High potential	Enigma — coach or move	Future star — invest	Star — accelerate, retain at all costs
Medium potential	Inconsistent — diagnose	Core player — keep growing	High performer — promote when ready
Low potential	Underperformer — manage out	Solid — recognise & retain	Specialist — value the depth

What the 9-box does well

Forces leaders to separate performance (what they did) from potential (what they could do) — these are conflated 80% of the time.
Creates a common language across business units for who's a 'star' vs a 'specialist' vs a 'high-potential'.
Surfaces hidden talent — the medium-perf-high-potential 'future star' is often invisible without a 9-box conversation.

What the 9-box does badly (the honest critique)

Potential is mostly unmeasurable. Most 'high-potential' assessments are 60% manager bias, 30% recency, 10% real signal.
Forced distributions (5% top-right, 10% bottom-left) damage trust and ignore actual team composition.
The labels leak. The moment someone learns they're 'medium-medium', their engagement drops. Industry research (CEB / Gartner 2015–2022) shows 'hi-po' communication actually decreases retention if mishandled.
It's a snapshot. People move cells fast — a new manager, a new role, a personal crisis can move someone two cells in 90 days.
It treats people as portfolio assets, not humans. That framing creates the wrong leader behaviour.

How to actually use it

Use the 9-box as a structured conversation among leaders, not a database field. Don't tell people their box. Re-calibrate every 6–12 months. Track cell *transitions* over time, not snapshot distributions. And accept that 'potential' is your best guess, not a number.

2. Talent review mechanics

A talent review is the meeting where a leadership team calibrates its people across teams. The framework is simple; the mechanics are where the value is created or destroyed.

The 7-step talent review

1
Scope
Pick a population (e.g. all directors+, or one BU). Don't try to review everyone in one session.
2
Prep packet
Each people leader sends in: 9-box placement, role criticality, retention risk, last 12 months' impact, growth area, succession status.
3
Facilitator
Senior HR (or an external facilitator) runs the meeting. Their job is to challenge ratings, surface bias, and keep time. Never the CEO or the most senior leader in the room.
4
Calibration
Go person by person. The leader presents; peers challenge; the room agrees. 'Sue is a 9-box top-right' should be defended with at least two specific examples.
5
Anti-bias scripts
Before each cohort: 'Are we measuring the same things across people?' 'Have we considered who's been here longer vs newer?' 'Are we conflating likeability with performance?'
6
Actions
Every person discussed gets at least one action: stretch assignment, mentor, comp adjustment, succession step. Without actions the review is theatre.
7
Follow-through
30 / 90 / 180-day check on actions. If actions don't get done, leaders learn the review doesn't matter and stop preparing.

Cadence that works

Annual full review + semi-annual lighter check-in. Anything more is noise; anything less and the data goes stale. Pair it with the comp planning cycle so decisions are integrated.

3. Succession depth charts

A succession depth chart names 1–3 potential successors for each critical role, at three readiness levels. Most companies should cover their top 20–50 roles, not every role.

Critical role	Incumbent	Ready now (0–6 mo)	Ready 1–2 yrs	Ready 3+ yrs
VP Engineering	Amal (3 yrs)	—	Priya (Sr Director)	Jon, Sara (Directors)
Head of Sales	Tomás (5 yrs)	Riya (Sr Director)	Marc (Director)	Three internal HiPos
Head of People	—	External hire required	Anya (Sr HRBP) with development	—

The hard rules

Critical roles = roles where a 90-day vacancy would seriously damage the business. Usually 5–20% of roles, not 100%.
If a critical role has *no* internal successor at any level, that's a glaring gap. The action is either external pipeline-building or accelerated internal development — never 'we'll deal with it later'.
Successors should be told they're being developed, but generally not which role they're successors for. The latter creates entitlement and damages the incumbent.
Review the chart quarterly. The half-life of a 'Ready now' designation is about 9 months — people leave, get promoted, lose readiness.

The bus-factor question

Walk the top-20 roles. For each, ask: 'If the incumbent quit tomorrow, who runs this for the next 6 weeks?' If you can't name a person for more than 5 of the 20, you have a succession problem the board should know about.

4. Build / buy / borrow

The talent supply decision. For every capability gap, the org has three choices: build internally, buy externally, or borrow via contractors/contingent.

Strategy	When it's right	When it's wrong	Typical cost & risk
Build	Capability is core, durable, culture-sensitive. Time horizon ≥ 12 mo. You have an L&D capacity.	Capability needed in 90 days. Skill is rare and externally cheap.	Lower marginal cost, higher time-to-capability. Retention dividend if done well.
Buy	Capability is needed fast, available externally, and the role is well-defined.	Capability is so niche to your business that no external person can land it. (Common in deep-tech and emerging markets.)	Higher cost (recruiting fees, comp premium), shorter ramp, integration risk.
Borrow	Capability needed for a finite project or to bridge while you build/buy. Includes consultants, contractors, fractional execs, vendors.	Capability becomes structural — at some point a fractional CFO needs to become a real CFO.	Highest unit cost, lowest commitment, fastest exit. Watch misclassification risk.

The classic mistake: defaulting to Buy. It feels decisive, but a 12-month senior hire often costs 1.5–2x the equivalent internal development path and brings cultural drift. Defaulting to Borrow is the second mistake — fractional execs that should have become full-time, contractors that became permanent without ever being on the headcount.

The 3:1 rule

If you can identify 3 internal people who could do a role in 12 months, build. If you can identify 1, buy and develop a successor in parallel. If you can identify 0, buy (and ask why your pipeline is empty).

Running the system end-to-end

Quarter 1: run a talent review using the 9-box. Identify top 20% high-potentials and bottom 5% manage-outs.
Quarter 2: update succession depth charts for critical roles. Identify the gaps.
Quarter 3: for each gap, decide build / buy / borrow. Commit budget and owners.
Quarter 4: review progress. Re-rate the high-potentials. Update the chart. Feed into the next year's comp & headcount planning.

This is the cadence at most well-run companies of 500+ employees. Below 500, run a lighter version annually. Below 50, the founder and one HR lead can keep it in their heads — but write it down once a year anyway.

FAQ

Frequently asked questions

Should we share 9-box placements with employees?

Mostly no. The research is consistent: telling people they're 'high-potential' creates entitlement; telling people they're not damages engagement. Share career conversations and development plans instead — those carry the same information without the labels.

How do we measure 'potential'?

Imperfectly. Best practice: triangulate manager assessment + 360 + a structured rubric (learning agility, scope expansion, complexity tolerance, peer judgement). Even then, expect 30% error. Use it as a conversation, not a score.

Is the 9-box being replaced by anything?

Most large companies are evolving to a simpler 3-box (Differentiated / Core / Risk) plus separate conversations about successors. Same information, less false precision.