The Capability Gap: Why $400B in Training Isn’t Building the Workforce Organisations Need

Organisations invest over $400 billion a year in employee training (Statista, 2024). Yet across industries, leaders report the same concern: their workforce lacks the capability required to execute strategy.

McKinsey Global Institute (2023) found that 87% of executives either face skills gaps today or expect them in the near future.

This is not a funding problem. And it is not a learning problem. It is a design problem.

The capability gap is the disconnect between what people have learned and what they can actually do consistently and effectively when it matters most in real work.

Training Is Not Capability

This gap begins with how organisations define the problem. Performance shortfalls are commonly interpreted as a lack of training. The response is predictable: more programmes, more content, and more mandated learning.

But training and capability are not the same. Training delivers knowledge. Capability is the ability to apply skills, judgement, and behaviours effectively in real conditions, particularly when it matters most.

Despite rising investment in learning, productivity across many developed economies has remained largely flat (OECD, 2023). While influenced by multiple factors, this trend suggests that training investment alone does not reliably translate into improved workforce performance.

The constraint is not the volume of training, but the organisation's ability to translate learning into performance in real work.

The Illusion of Learning

This misalignment shapes how success is measured. On the surface, learning appears to be working. Courses are completed. Dashboards are full. Certifications are awarded.

Yet performance on the ground often tells a different story.

Corporate learning systems have converged on a simple proxy for success: completion. Courses completed, hours logged, and certifications achieved are easy to track, but reveal little about whether capability has actually improved.

The data makes this disconnect clear. While 93% of organisations track learning activity, fewer than 30% measure its impact on job performance (Association for Talent Development, 2025). Only 5% believe they have the capabilities they need (McKinsey & Company, 2023).

The result is an illusion of progress. Activity increases, visibility improves, but capability remains inconsistent and difficult to verify.

The Capability Gap Is a Design Problem

This is not accidental. It is the outcome of how learning systems are designed.

• Learning is separated from work.
• Success is measured through activity rather than performance.
• Ownership of capability is fragmented across functions.

Capability gaps persist when learning is disconnected from role requirements, when application is not measured, and when reinforcement is inconsistent or absent.

In these environments, adding more training rarely resolves performance issues. It often increases inefficiency by consuming time and resources without improving execution.

Leaders who recognise this shift begin to think in terms of performance systems. Capability is defined in the context of real work, measured against outcomes, and reinforced through application.

Timing Is Critical to Capability

Even relevant learning often arrives at the wrong moment. Too early, and it is forgotten or remains abstract. Too late, and performance has already been impacted.

This gap can be understood as learning latency: the delay between when knowledge is acquired and when it is needed in practice.

As work changes faster, learning latency becomes more costly. Capability failures surface not annually, but in real-time decisions and execution breakdowns.

Capability is not built through exposure to content alone. It is built when learning is closely connected to work, appearing at the point of decision, reinforced through application, and evolving as roles change.

It is not simply what people learn, but when learning becomes available in the context of work that determines whether capability develops.

Skills Are Changing Faster Than Systems Can Respond

Even when training is effective at a point in time, it is becoming obsolete more quickly.

The World Economic Forum (2023) estimates that 44% of workers' core skills will change within five years, accelerated by AI reshaping how work is performed and how rapidly roles evolve.

Traditional learning systems were designed for a slower pace of change. Annual cycles, fixed curricula, and centralised functions made sense when skills evolved gradually.

By the time training reaches the workforce, parts may already be outdated or misaligned with current work demands. In many domains, the lag now exceeds the useful lifespan of the skills being taught.

Organisations that assume training guarantees ongoing proficiency are increasingly exposed to this risk.

From Learning Systems to Capability Systems

Addressing this challenge requires more than adapting training. It requires rethinking the system itself.

Traditional learning systems are designed to deliver content. Capability systems are designed to enable performance.

• Align skills with real work.
• Surface capability gaps through performance data.
• Support decisions at the point of need.
• Reinforce learning through continuous application.

Unlike learning optimisation, capability systems are organised around performance exposure, revealing gaps through work itself rather than relying on periodic assessments or self-reported needs.

In this model, learning is no longer a discrete event. It is embedded within the flow of work.

This points to a fundamentally different approach: learning that is continuous, embedded, and guided by learning intelligence, enabling organisations to connect skills, work, and outcomes, and to adapt as performance evolves.

A Structural Problem Requires a Structural Shift

The capability gap is not a content problem. It is a structural one.

Learning sits outside of work. Performance is measured differently from learning activity. And few systems consistently connect them in practice.

As a result, organisations do what the system allows: they deliver more training, track more activity, and report more progress without materially improving capability.

This is not a failure of effort. It is the outcome of design.

Closing the gap requires a shift from learning as an activity to capability as a system, connecting skills, work, and performance through learning intelligence.

Because the question is no longer whether organisations are investing in learning, but whether that learning shows up where and when it matters.

That is the capability gap.

About Perpetua Foundry

Perpetua Foundry is transforming how organisations develop their people in the age of AI.

We believe the challenge is not a lack of training, but a lack of learning intelligence: the ability to continuously connect skills, work, and performance as roles and decisions evolve.

We are building Perpetua Intelligence, an AI-native learning intelligence system designed to help organisations move beyond training activity and start building capability in real time.