Most executive teams have already agreed that reskilling matters. Far fewer have agreed who pays for it, and that is where the program quietly stalls.
How is reskilling funded and scaled inside a large organization? In practice there are three answers: a centralized learning and development budget, business-unit budgets that carry the cost against their own P&L, or a transformation program that absorbs the training spend inside a wider investment case. Most companies use some combination of the three. Very few have decided on purpose which one they are using.
That sounds like a technicality. It is not. The funding model determines who owns the outcome, how long the program survives, and what happens to it the first time the forecast tightens. A reskilling initiative funded from a central cost center behaves very differently under pressure than one funded inside a business case with a return attached to it.
The numbers make the stakes concrete. The Association for Talent Development's 2026 State of the Industry report, published in May 2026 with 2025 data, puts average direct learning expenditure at $846 per employee, down from $1,254 the year before. Formal learning hours rose over the same period, to 16.7 per employee. Companies are asking their learning functions to deliver more with roughly a third less money. That is a funding design problem before it is a training problem.
How Is Reskilling Funded and Scaled in Large Enterprises?
Each of the three models carries a different set of incentives. Understanding what each one actually buys is the starting point for any credible budget conversation.
The centralized L&D budget. A single pot, owned by the learning function, allocated across the organization. It protects standards, keeps quality consistent across markets, and makes it possible to build capabilities that no single business unit would fund on its own. Its weakness is structural: it sits in the cost center column, it has no revenue line to defend it, and it is visible to finance as a discrete number that can be reduced without touching an operating plan. Central budgets are the easiest to build and the easiest to cut.
Business-unit funding. Each division pays for the reskilling of its own people, either directly or through an internal charge from L&D. This model produces ownership. When a business unit leader signs the invoice, the training tends to be relevant, attendance tends to be enforced, and managers tend to care whether the skill shows up in the work. The trade-off is fragmentation. Cross-cutting capabilities such as AI fluency, data literacy, or cross-border leadership get underfunded because no single unit captures the full return, and short-horizon P&L pressure pushes spend toward immediate operational needs.
Transformation-funded reskilling. The training cost is written into the business case for something else: an ERP migration, a plant automation program, an AI deployment, a market entry. This is the most resilient of the three, because the spend is defended by the return on the wider project rather than by the learning function. It is also the most time-bound. When the project closes, the funding closes with it, and the capability often stops being maintained.
The organizations that scale reskilling successfully do not choose one model. They run a deliberate mix: a central floor that guarantees a minimum standard everywhere, business-unit funding for role-specific capability, and transformation budgets for the surges. What matters is that the mix is designed rather than inherited.
Building the Corporate Reskilling ROI Case Finance Will Approve
The reason reskilling budgets are vulnerable is rarely that executives doubt the value. It is that the learning function cannot demonstrate it in the terms finance uses.
LinkedIn's Workplace Learning Report found that 67% of learning and development leaders struggle to demonstrate training impact to their executives, and only 8% describe themselves as highly confident in their ability to measure the business impact of learning. That is not a measurement problem. It is a negotiating position, and it is a weak one.
The correction is procedural, not analytical. Three things change the conversation.
Set the baseline before the program starts, not after. The most common reason a reskilling ROI case fails is that nobody recorded the "before" state. Error rates, cycle times, internal fill rates, time to competence in a role, cost of vacancy in critical positions. All of these are available in systems the company already runs. Capturing them in advance costs almost nothing and converts a qualitative argument into a quantitative one.
Model the cost of not doing it. Finance is comfortable with counterfactuals. The relevant comparison is not training spend against zero. It is training spend against the fully loaded cost of buying the same capability externally, plus the cost of the vacancy while the search runs, plus the risk premium on a hire who does not know the business. In markets where the talent is scarce and the employer brand is unproven, that comparison usually favors building.
Attach the case to a decision that is already being made. A reskilling proposal that arrives on its own competes with everything else in the budget cycle. The same proposal attached to an automation program, a market entry, or a succession gap arrives with a sponsor and a deadline.
McKinsey research indicates that fewer than 40% of companies have a clear reskilling strategy, which means most organizations are opening roles their current workforce is not equipped to fill. The same research finds that organizations applying people and learning analytics effectively achieve productivity gains of up to 25%. The gap between those two figures is where the ROI case lives.
What Scaling Actually Means, and What It Doesn't
Scaling is the word that causes the most confusion in these conversations, because it is usually read as a volume metric. More learners, more hours, more content in the catalog.
The ATD data is instructive here. Per-employee spend fell by a third in 2025 while formal learning hours rose 22%. More hours, less money. Cheaper content libraries and AI-assisted content development can explain part of that, but the pattern should give any CFO pause: the organization is consuming more training and paying less for it, and nobody can say whether capability improved.
Scaling a reskilling program means something narrower and harder. It means the same capability reaches the same standard across more of the organization, in more markets, without the quality degrading as the population grows. Three design choices determine whether that happens.
Cohorts, not catalogs. Programs that move groups of people through a defined capability together, with a shared timeline and shared work, embed faster than self-serve libraries. A catalog scales distribution. A cohort scales capability.
Manager involvement as a funded requirement, not a hope. Line managers determine whether a new skill gets used. If the manager's time on coaching and application is not built into the program design and protected in the operating plan, the training becomes an event rather than a change.
Protected time. Half of employees report that workload leaves no room for development. Any scaling plan that does not account for capacity is a plan to generate completion data.
We wrote about why most corporate training programs fail to move the needle in Reskilling at Scale: The Corporate Training Imperative. The funding mechanics described here are what turn the principles in that piece into something a finance committee will actually sign.
Where Public Co-Funding Fits, Particularly in Europe
For companies operating in the European Union, there is a funding layer that most mid-market organizations underuse.
The Council Recommendation on human capital, adopted in March 2026, is explicit that public spending alone cannot meet the scale of Europe's reskilling need. Job-related training is predominantly financed by employers, close to 90% of all job-related adult learning. At the same time, one in three companies in the EU provide no training at all to their staff, with cost cited as one of the main obstacles alongside workload and time. Cedefop estimates that 128 million adults across Europe need upskilling or reskilling.
Against that backdrop, the European Union is deploying more than €150 billion between 2021 and 2027 through instruments including ESF+, the Recovery and Resilience Facility, Erasmus+, ERDF, the Just Transition Fund, and InvestEU. National mechanisms sit alongside these, and in Italy the interprofessional funds remain a familiar route for financing structured training.
Two cautions apply. First, co-funding is a cash-flow instrument, not a strategy. Programs designed backward from an available grant tend to train the wrong people in the wrong capability at the right price. Design the program against the business need, then find the funding that fits it. Second, the administrative load is real, and it belongs in the cost model. A grant that requires six months of reporting to recover 30% of the spend may still be worth it, but only if that is a conscious calculation.
For companies operating on both sides of the Atlantic, the asymmetry matters. The US has no comparable structure at federal level. Support is fragmented across state workforce boards and tax treatment, which means a global program often has to be funded on different logic in different markets, even when the capability being built is identical.
The Most Common Mistake: Measuring Completion Instead of Capability
The single most damaging habit in corporate reskilling is reporting completion rates to an executive audience.
It is understandable. Completion is easy to capture, it is always available, and it looks like progress. A slide showing 94% completion across 3,000 employees reads as a well-run program. But it answers a question nobody at the executive table is asking. Completion tells you that people finished. It says nothing about whether anyone can now do something they could not do before.
The damage is not that the metric is uninformative. It is that it makes the budget indefensible. When the forecast tightens and the CFO asks what the organization got for its learning spend, a completion figure invites the obvious response: if the measure of success is attendance, the program can run at lower cost. Completion data is an argument for cutting the budget, delivered by the function that owns it.
The alternative is a small set of measures the business already recognizes. Time to capability, meaning how long it takes a participant to reach the required standard in the role. Internal fill rate for the positions the program was designed to feed. Reduction in time to fill and in vacancy cost for critical roles. Manager-assessed application of the skill at 90 days. None of these are exotic, and all of them survive contact with a finance review.
In our own search work across 40+ markets, the pattern is consistent: the companies that can tell us precisely which capabilities they intend to build internally and which they will buy externally are the ones with the shortest, cleanest hiring processes. Clarity about the reskilling plan is what makes the external search brief precise.
Protecting the Budget When the Forecast Tightens
Reskilling is frequently among the first lines cut in a downturn, and the reason is rarely conviction. It is that the spend is annual, discretionary, and centrally held, which makes it structurally easy to remove.
Three governance choices change that exposure. Commit multi-year, not annually, so that reducing the program requires reopening a decision rather than declining to renew one. Anchor the program to a strategic initiative with its own board sponsor, so that cutting it visibly delays something the board has committed to. And place accountability with a business leader rather than with the learning function, so that the person defending the spend is the person who owns the outcome it supports.
The last point connects to a broader shift in how the people function operates at senior level, which we examined in The Evolving CHRO Role in 2026. A CHRO who arrives at the budget conversation with a capability plan tied to the operating plan is negotiating from a different position than one presenting a training calendar.
ATD data suggests around three-quarters of organizations expect to increase AI-related training spending in the next fiscal year. That money will be allocated. The question is whether it is allocated inside a structure that can demonstrate what it produced, or inside one that will have to justify itself again in twelve months with the same evidence it has today.
Key Takeaways
- There are three real funding models for reskilling: central L&D budget, business-unit funding, and transformation-funded programs. Most companies run a mix by accident rather than by design.
- ATD's 2026 State of the Industry puts direct learning spend at $846 per employee for 2025, down from $1,254, while learning hours rose. More training, less money, and no clear view of capability gained.
- The corporate reskilling ROI case fails on process, not analysis. Baselines set before the program starts and a costed counterfactual do more for approval than any measurement framework.
- Scaling means the same capability reaching the same standard across more of the organization. Cohorts, funded manager involvement, and protected time make that possible. A larger catalog does not.
- Completion rates are the most common and most damaging metric in corporate reskilling. They give finance an argument for spending less, delivered by the function that owns the budget.
Future Manager World works with CHROs and senior leadership teams across 40+ markets. Talk to our team.
Frequently Asked Questions
How is reskilling funded in most large companies?
Through three models: a centralized learning and development budget, business-unit budgets charged against their own P&L, or a transformation program that absorbs training cost inside a wider investment case. Most organizations use a mix of all three.
How much do companies spend on reskilling per employee?
ATD's 2026 State of the Industry report puts average direct learning expenditure at $846 per employee for 2025, down from $1,254 the previous year, while formal learning hours per employee rose to 16.7.
How do you measure corporate reskilling ROI?
Set baselines before the program starts, then track time to capability, internal fill rate, reduction in time to fill, and manager-assessed application at 90 days. Completion rates measure attendance, not capability.
Can EU companies get public funding for reskilling?
Yes. The EU is deploying over €150 billion between 2021 and 2027 through ESF+, the Recovery and Resilience Facility, Erasmus+, ERDF, the Just Transition Fund, and InvestEU, alongside national instruments. Design the program against the business need first, then find the funding that fits.
.jpg)
