We are excellent at measuring performance...but terrible at measuring potential

Why future capability, learning velocity, judgement and adaptability matter morethan retrospective performance metrics

Large organisations have extraordinary apparatus for measuring performance: objectives, scorecards, dashboards, ratings, revenue contribution and increasingly granular productivity data. Yet most of this machinery answers a backward-looking question: how well did this person perform in a role whose requirements we already understand?

That question is necessary, but no longer sufficient. When strategy, technology and operating models are changing faster than job descriptions, the more consequential question is whether an individual can grow into work that does not yet fully exist. The C-suite therefore needs a second talent lens; one that assesses future capability through learning velocity, judgement, adaptability and the capacity to perform at greater scale and complexity.

The answer is not to abandon accountability. It is to stop treating past performance as a proxy for future readiness. This approach carries the risk of promoting excellent specialists into ill-fitting leadership roles, overlooking unconventional talent, reinforcing historic advantage and discovering succession gaps only when a critical role becomes vacant.

The rear view mirror problem

Performance measures are attractive because they appear concrete, converting contribution into comparable numbers and support pay, promotion and workforce planning. But their precision can create false confidence. Results reflect not only individual capability but territory quality, team strength, manager support, inherited accounts, market conditions and the degree to which a role matches someone’s existing expertise.

More fundamentally, performance is role-specific. Delivering a known process efficiently does not demonstrate the ability to redesign it. Hitting a target in stable conditions does not reveal how someone will respond when the assumptions behind the target collapse. Technical mastery does not automatically translate into enterprise judgement, influence across boundaries or the ability to lead through uncertainty.

Harvard Business Review’s Claudio Fernández-Aráoz defines potential as the ability to adapt and grow into increasingly complex roles and environments. His five indicators: motivation, curiosity, insight, engagement and determination, shift attention from accumulated credentials to the capacity to acquire what the next context requires. This distinction is becoming more important as AI shortens the useful life of some technical knowledge and changes the composition of managerial work.

What should replace the single-score view?

Future capability – asks whether a person can create value against the organisation’s emerging strategic agenda, not simply its current operating plan. Assessment begins with the few capabilities the strategy will make scarce – perhaps AI-enabled decision-making, ecosystem leadership, regulatory judgement or business-model reinvention – and works backwards to identify evidence of readiness. McKinsey argues that talent should rather be matched dynamically to the roles where the most value is at stake, supported by skills baselines, internal mobility and goals.

Learning velocity is the speed and quality with which someone turns unfamiliar experience and feedback into improved action. It is visible in how quickly assumptions are updated, whether lessons transfer between contexts, and whether the individual actively seeks disconfirming evidence. It is not the volume of courses completed. The best evidence comes from a sequence: exposure to novelty, reflection, behavioural change and stronger subsequent outcomes.

As information and analysis become cheaper, judgement becomes more valuable. Leaders must frame the right problem, distinguish signal from noise, understand second-order consequences and make responsible choices where data are incomplete or objectives conflict. Deloitte’s 2026 Human Capital Trends warns that AI adoption is outpacing organisational oversight and can weaken human ownership and critical judgement. For succession decisions, executives should therefore examine decision quality: what trade-offs were recognised, whose perspectives were sought, how risks were calibrated and whether the leader changed course when evidence changed.

Adaptability is not passive resilience or cheerful acceptance of change. It is the ability to preserve purpose while changing methods: to reconfigure resources, work across unfamiliar boundaries and remain effective when authority, technology or market conditions shift. Deloitte reports that seven in ten business leaders identify being fast and nimble as their primary competitive strategy for the next three years. Yet many talent systems continue to reward consistency inside established roles more reliably than constructive

A better executive operating system

A potential-led system should not collapse these dimensions into another opaque score. It should create a disciplined portfolio of evidence.

• Define potential against strategy. Replace generic “high-potential” labels with explicit hypotheses about the scale, complexity and capabilities future roles will demand.
• Use multiple observations. Combine structured career interviews, simulations, stretch assignments, stakeholder evidence and longitudinal learning data. A manager’s confidence is not an assessment method.
• Separate performance, potential and readiness. Performance is contribution in the current role; potential is capacity to grow; readiness is proximity to a particular next role. Treating them separately makes succession conversations more honest.
• Create controlled tests. Give talent consequential but bounded assignments: a turnaround, cross-functional launch, unfamiliar market or enterprise-wide decision. Observe how the person learns and exercises judgement, not only the final result.
• Calibrate for opportunity and bias. Ask whether each candidate has had comparable sponsorship, visibility and access to high-value assignments. Potential cannot be inferred fairly from exposure the organisation distributed unequally.
• Track development as a business asset. Measure internal mobility, capability acquisition, succession coverage and the conversion of stretch experience into stronger performance, not merely training participation.

A strategic return

This is not a softer alternative to performance management. It is a more demanding approach to enterprise risk. McKinsey’s analysis of 1,800 large companies found that organisations combining people development with strong performance were about 1.5 times more likely to remain top-tier performers, experienced roughly half the earnings volatility, and recorded attrition around five percentage points below performance-driven peers. During 2019 – 21, their revenue grew twice as fast as that of performance-driven companies.

The implication for CEOs and boards is direct: the leadership bench should be managed as a portfolio of future options. Retrospective metrics tell us who has created value under known conditions. Potential evidence tells us who may create value when the conditions, tools and rules change. In an era defined by discontinuity, the second question is increasingly the one on which strategy depends.

Conclusion

Organisations do not suffer from too little performance data. They suffer from allowing what is easiest to count to dominate what is most important to know. The answer is not a new nine-box grid or a more sophisticated algorithm. It is a richer executive discipline: define the future capabilities strategy requires, expose people to meaningful tests, observe how quickly they learn, examine the quality of their judgement and reward those who can adapt without losing accountability or purpose.

“Past performance deserves a voice in talent decisions. But it should no longer have the casting vote”.