Training Budgets Are at a Record High. So Why Is No One Ready to Play?

‍This article is the full text of a contributed column published in the January 2026 special issue of the Korean HR magazine Injae Gyeongyeong (Talent Management).

‍

A Mismatch Between Classroom and Workplace, and a New Productivity Paradox

Looking back on 2025, we have to ask ourselves a painful question. More money than ever before in the history of corporate training went into AI literacy and digital transformation programs. Did a matching level of impact actually reach the business?

Since the start of the 2020s, global companies have poured record budgets into employee training to respond to generative AI and digital transformation. According to a Coursera report, enrollments in generative AI courses worldwide jumped 1,060% in 2023 compared with the year before.

Yet business leaders remain unimpressed. In a 2024 Robert Half survey, 65% of technology leaders, including CIOs and CTOs, reported skills gaps in their departments, and 95% said they struggled to find people with the capabilities they needed. They keep asking the same thing: plenty of people have been trained, so why can’t we find anyone ready to put on a project?

Nobel laureate economist Robert Solow once captured the productivity paradox with the line, “You can see the computer age everywhere but in the productivity statistics.” HRD faces much the same reality today.

‍“AI training is everywhere, except in the business’s performance metrics.”

To fix this, many companies reflexively review their training content. But make no mistake: the heart of today’s dilemma is not content quality. The real problem is that training and talent deployment are structurally disconnected.

‍

Why Supply-Driven HRD Misses the Business Context

HRD has long run on a strictly supply-driven model. The overriding goal was to source good courses, roll them out to employees, and push up completion rates. We have made this model as efficient as possible, but it has a built-in limit: it is out of step with the business context.

HRD supplies standardized programs (e.g., a data analyst course or a leadership course), while the business wants the applied ability to solve its department’s problems right now. Because of this gap, plenty of employees get trained, yet business leaders still feel they have no one to deploy. The bigger problem is that the data needed to fix this mismatch is cut off.

Look more closely, and HRM and HRD see talent through different lenses. Strictly speaking, if the data HRM manages is a record of the past (a resume), the skills data HRD must prove is tomorrow’s potential (a competency profile).

Most companies, however, have not found the words to describe that potential. Recording in the HR system that “this employee completed an intermediate Python course,” for instance, means very little. Business leaders don’t need Python course graduates; they need people who can solve the problem at hand. The data should be defined in concrete skill units, such as “this employee can create data visualizations using generative AI.”

Only when HRD translates the skills gained through training into concrete language like this, and that growth data is integrated into the HR system, can an employee be rediscovered beyond the label of “manager with 10 years of experience” as “an expert ready to join an AI project.”

‍

Use Skills Data to Reconnect Corporate Training and Talent Deployment

To reconnect these broken links, HRD’s role is to make skills data visible. Staffing and internal mobility are, of course, not HRD’s job alone. But HRD knows better than anyone the latent skills employees have. We need to make that data visible and give the whole organization a starting point for moving talent where it’s needed.

Consider a real example. Schneider Electric introduced an open talent market to uncover hidden talent. Instead of job titles, it made employees’ skills visible. As a result, internal talent that had previously gone unnoticed was deployed across departmental boundaries to more than 13,000 projects.

Unilever is another case. An R&D employee with 18 years at the company used skills data to rediscover their aptitude and moved into a role as a D&I (Diversity & Inclusion) manager. That kind of dramatic internal move would not have been possible without skills data.

Recommended Reading
Read Skills-Based HR in Practice: Unilever, IBM, and BMO

When HRD goes beyond training records and proves with data what an employee can do right now, the blocked arteries of the internal talent pipeline open up and the whole organization can move as one.

‍

HRD’s Real Impact Is Proven Outside the Classroom

Change the ROI Standard from Satisfaction to Skill Matching

It’s time for HRD to be honest with itself. A training satisfaction score of 4.5 and a completion rate of 98% are HRD metrics, not business metrics. The real result that executives will accept is the Skill Matching Rate.

We need to prove with data that developing people internally is more cost-effective than hiring from outside. That starts with objective assessment of employees and clear skill definitions, because real internal mobility only becomes possible once that data foundation is in place.

Combine that foundation with long-term, organization-wide development programs, and employees will grow into ready players who can be deployed to projects at the right time. This is how HRD moves past the outdated view of itself as a cost center and creates visible value over the long term.

Start with Small Skills Data, Not a Grand System

HRD can’t change everything on its own, of course. But the change should start with HRD, which holds the most reliable data.

If adopting a large platform or solution feels like too much, pick one key department and start a pilot. This is also the principle Telta stresses most with clients considering skills-based HR. Define the skills the business needs, close the gap through training, and then track whether those employees actually contribute to projects. Connecting even this small amount of data can prime the pump for a much larger change in your talent ecosystem.

In 2026, HRD should not be aiming for a classroom with the perfect curriculum. It should be aiming for the heart of the business, where people who have grown ride the current of data to exactly where they’re needed.

Related Reading

• To Improve Job Training Outcomes, Assess Competencies Before Training Starts
• Why Completion Rates Can’t Measure the Impact of Corporate AI Training

‍