What I got wrong about reskilling budgets

The question

For a long time I assumed reskilling programs failed because they were underfunded. It is the obvious explanation, it is the one everyone reaches for, and it is comfortable because it puts the problem outside the room.

Two programs changed my mind. Both were properly funded. Both had good content. Both were mostly ignored.

What is actually going on

The constraint was not budget and it was not content. It was that line managers could not make space for it, and were not measured on whether they did.

Look at it from the manager's position and the behaviour is entirely rational. They have a number to hit this quarter. Development pays back over a year or more, to a team that may not be theirs by then. Nothing in their objectives references it. Every hour they protect for learning is an hour visibly missing from delivery, and the cost of that is immediate and personal.

So the people who most needed the time had the least of it, and no manager was accountable for the difference. The program was not rejected. It was quietly outcompeted.

There is a second-order effect worth naming. Once a program is seen as optional, the people who do attend are the ones with slack, which is rarely the group the program was designed for. Completion data then looks acceptable while the actual capability gap is untouched.

What this means for you

Before funding the next program, check two things.

Is the time protected, and by whom. If the answer is "people can do it when they have capacity", you have your answer already.

What happens to a manager who does not release their team. If nothing happens, nothing will happen.

I would now rather run a smaller program with protected time than a better one without it. That is close to the opposite of what I would have said ten years ago.

Where I can help

The interesting work here is usually not the learning design. It is the accountability around it: what managers are measured on, how development time is protected, and how the organization notices when it is not.

I have redesigned enterprise learning around exactly this, including a redesign of annual mandatory training that saved $90K while improving completion, largely by removing content that was never the constraint.