Talent to Value
talent-strategycritical-rolesmckinseyorg-designagentic-organizationperformance-management
Talent to Value
McKinsey & Company's talent-strategy framework (in use ~a decade): identify the 30–50 critical roles that drive ~80% of value, define a success profile for each, and match top performers to them. Most of those roles sit well below the C-suite (50–65% are three levels below the CEO), which is why they get overlooked.
The AI-era update: "Talent and Agents to Value"
Per Rewiring Talent to Value in the Age of AI (McKinsey) (2026), the four original steps each get an agentic twist and a fifth step is added:
| Step | Original | AI-era version |
|---|---|---|
| 1 | Define the value agenda | Continuously re-allocate — value pools move faster than planning cycles |
| 2 | Identify critical roles | Identify critical roles and critical agents; the unit of value is a human–agent system |
| 3 | Match talent to roles | Assess how much a person can amplify value with AI ("super users") |
| 4 | Operationalize via leaders | Re-look at the top team's AI fluency |
| 5 | — | Manage system performance: agents on decision quality/reliability/speed/cost; humans on impact, workflow improvement, ethical use, collaboration |
New critical roles it names: agentic-workflow mission owners, agent-ops platform leads, agent-governance liaisons — plus elevated AI product owners, human–agent workflow architects, and data/knowledge specialists.
Why it's useful
- Gives a board-legible vocabulary for a question the vault keeps circling: which work goes to agents, and how you measure a mixed human–agent team.
- Step 5's dual performance model ("who did the work?" → "how well did the system perform?") pairs naturally with the eval layer (LLM as Judge, Judgment Models (System 1 AI)) needed to actually score agents.
- The "super user" lens is the talent-side view of the Advantage Gap.