Kinetiq
Workforce Trends

McKinsey’s Reskilling Research: Why 87% of Companies Face a Skills Gap

K

Kinetiq Team

McKinsey’s Reskilling Research: Why 87% of Companies Face a Skills Gap

Eighty-seven percent of companies say they currently have, or expect to have, skills gaps within the next few years. That figure comes from McKinsey Global Institute’s research on reskilling and the future of work, and it represents something more fundamental than a hiring problem. When nearly nine out of ten organizations acknowledge the same capability shortfall, the issue is not that talent is scarce. The issue is that the systems designed to build talent are not working.

McKinsey’s research goes further. An estimated 375 million workers globally may need to switch occupational categories by 2030. The skills in highest demand are not purely technical. Social-emotional skills and technology skills are growing fastest, often in combination. The organizations that close this gap will not be the ones that spend the most on training. They will be the ones that build systems for applying new capabilities in real work.

What the Research Shows

The Skills Gap Is Nearly Universal

The 87% figure is striking because of its breadth. This is not concentrated in a single industry or region. Companies across sectors, from technology to manufacturing to professional services, report the same fundamental problem: the capabilities their workforce has today do not match the capabilities their strategy requires. And the gap is widening, not closing.

Traditional responses to skills gaps follow a predictable pattern. Organizations identify the deficit, purchase training programs, deliver those programs to employees, and then wonder why the gap persists. The problem is that training delivery and capability building are not the same thing. As Skills-Based Hiring in 2026 details, organizations are increasingly recognizing that credentials and course completion do not reliably predict on-the-job performance.

375 Million Workers May Need to Switch Categories

McKinsey estimates that approximately 375 million workers globally may need to switch occupational categories by 2030. This is not a marginal adjustment. It represents a fundamental restructuring of how work is organized and what skills are valued. The shift is being driven by automation, AI adoption, and changing business models that reward adaptability over routine expertise.

This scale of transition cannot be managed through traditional training programs alone. It requires systematic approaches to capability development that connect learning to application in real time. Workers who need to shift categories need more than new knowledge. They need new systems for how they approach work, make decisions, and collaborate across functions.

Social-Emotional and Technology Skills Are Growing Fastest

The skills in greatest demand are not what most training catalogs emphasize. McKinsey’s research identifies social-emotional skills (communication, collaboration, adaptability, leadership) and technology skills (particularly AI literacy and data fluency) as the fastest-growing categories. As The AI Literacy Requirement explores, most organizations are not prepared for the pace of this shift.

The intersection of these two categories is particularly important. The most valuable emerging capabilities are not purely technical or purely interpersonal. They live at the overlap: the ability to use technology tools effectively within a team context, to communicate clearly about AI-assisted decisions, to collaborate asynchronously using digital systems. These compound skills cannot be developed through single-topic training modules. They require integrated systems that build multiple capabilities simultaneously through applied practice.

The Training-to-Performance Gap Persists

Perhaps the most consequential finding is the persistent gap between training investment and performance outcomes. Organizations are spending more on learning and development than ever before, yet the skills gap continues to widen. This paradox points directly to a design problem. The issue is not the volume of training. It is the architecture of how that training connects (or fails to connect) to daily work.

Research consistently shows that knowledge retention from traditional training drops by 70-90% within days if the learning is not applied in context. The implication is clear: the 87% skills gap is not primarily a knowledge problem. It is a transfer problem. Organizations have the content. What they lack is the system for turning content into capability.

Why This Matters for Teams

Skills gaps do not manifest as abstract organizational deficits. They show up in specific, daily team failures. A team that lacks structured decision-making skills does not just “make bad decisions.” It experiences slower project timelines, more rework, interpersonal friction, and eventually, turnover. A team that lacks AI literacy does not just miss efficiency gains. It develops informal, inconsistent practices that create compliance risks and quality problems.

The team level is where skills gaps become operational costs. And it is where the solution must be implemented. Gallup’s engagement data shows that 23% global engagement and $8.8 trillion in lost productivity correlate directly with the conditions that skills gaps produce: unclear expectations, poor communication, weak collaboration systems.

When teams lack shared frameworks for common challenges (how to prioritize, how to hand off work, how to run a productive meeting, how to integrate AI tools), every interaction carries unnecessary friction. That friction compounds. Over weeks and months, it produces the disengagement, burnout, and turnover that organizations then try to solve with more training, completing a cycle that never addresses the root cause.

The Gap the Data Reveals

McKinsey’s research correctly identifies the scale and urgency of the skills challenge. Where it provides less guidance is on the specific mechanisms that close the gap at the team and individual level. The recommendations tend toward macro-level strategies: invest more in reskilling, create learning cultures, develop public-private partnerships for workforce development.

These are necessary but insufficient. The missing layer is the operational design that connects learning to practice. What does a team actually do differently on Tuesday morning because someone completed a training module on Friday? Without an explicit answer to that question, training remains an event rather than a system.

The transfer problem has a specific architecture. Effective capability building requires three elements that most training programs omit. First, a framework that gives the learner a repeatable structure for the skill. Second, immediate application in real work within the learner’s actual context. Third, feedback loops that refine the application over time. When any of these elements is missing, the training-to-performance gap remains.

This is consistent with what Deloitte’s Human Capital Trends research finds: 93% of leaders agree that skills-based approaches improve talent outcomes, but only 20% are making real progress on implementation. The knowing-doing gap in capability development mirrors the skills gap itself. Organizations understand what is needed. They lack the systems to execute it.

What This Looks Like in Practice

Closing skills gaps requires a fundamentally different approach to capability building. Instead of training programs that deliver knowledge in isolation, effective systems embed skill development into the operational rhythm of real work.

Consider how this works for a skill like “clear asynchronous communication,” one of the social-emotional capabilities McKinsey identifies as high demand. A traditional approach would offer a workshop on writing better messages. An applied systems approach would build a team protocol for async updates that specifies structure, frequency, and feedback mechanisms. The protocol itself becomes the learning tool, and every use of it is a practice repetition that deepens the skill.

KINETIQ’s framework treats every operational system as a capability development opportunity. Priority-setting rhythms build decision-making skills. Handoff protocols build communication clarity. Feedback systems build the social-emotional skills that McKinsey’s research identifies as critical. The skill is not taught separately from the work. It is built into the system that produces the work.

This integrated approach addresses the three elements that traditional training misses. The framework is embedded in the operational system. The application is immediate because the system runs in real work. The feedback loop is built into the rhythm of the system itself. No separate “learning transfer” initiative is required because transfer is the design, not an afterthought.

The World Economic Forum’s Future of Jobs data projects that 44% of core skills will change within five years. LinkedIn’s Workplace Learning Report shows that AI skills demand is surging while learning engagement remains a persistent challenge. Both point to the same conclusion: the velocity of skills change requires a fundamentally different architecture for capability building, one that is continuous, applied, and embedded in how teams actually operate.

McKinsey finds that 87% of companies have or expect skills gaps, yet organizations continue to rely on training models with 70-90% knowledge loss within days. The issue is not the volume of training. It is the architecture of how learning connects to daily work.

The 87% skills gap will not be closed by more training. It will be closed by better systems. Organizations that build operational infrastructure for continuous, applied capability development will outperform those that continue to treat training as a separate, periodic event. The research is clear on this point. The question is which organizations will redesign their systems accordingly.

Related Reading

Share this article:
K

Written by

Kinetiq Team

Contributing writer at Kinetiq, covering topics in cybersecurity, compliance, and professional development.