Are L&D leaders doing good work but still being treated as a cost centre?

Candice Mitchell asked me this question recently, and the honest answer is often yes. But that is not the most interesting part of the question. The more important question is this: why does the profession keep expecting respect for work that decision-makers cannot clearly see, cannot clearly use, and cannot clearly defend?

Because in many organisations, L&D has helped create the very conditions it now complains about. If your function is mostly reactive – taking orders, producing training, reporting completions – then “cost centre” is not a slur but an accurate description of how the organisation has learned to use you.

That is not a perception problem alone but a role problem. And it is also a measurement problem.

If the only evidence you can produce is activity, attendance, and output, then you are not making a strategic case. You are describing throughput. Busyness and business are not the same thing.

Too much of L&D still behaves like a supply chain: a request comes in, a product goes out, and metrics are counted – time to design, time to supply, completions, and test scores – and everyone pretends that proves value.

It doesn’t.

The deeper issue is that many L&D teams are still positioned as service providers, not capability builders. They are brought in late, asked to solve problems they did not help define, and then judged on results they were never properly empowered to influence. That is how you end up in the familiar cycle: busy, useful, underpowered, and undervalued.

Capacity makes it worse. Teams are stretched thin, which leaves no room for the kind of work that would change the conversation and strategic thinking gets squeezed out by delivery pressure.

Budgets make it worse. When money tightens, organisations don’t usually become more thoughtful about capability; they become more transactional. They cut slack, not dependency.

Sponsor behaviour makes it worse too. If leaders keep asking for courses, L&D will keep supplying courses. If they keep rewarding responsiveness, the function will keep optimising for responsiveness. Then everyone acts surprised when the function is seen as an order-taking operation.

The compliance burden reinforces the same pattern. Mandatory training volume can swamp anything more valuable, and once a function is judged by volume, it starts behaving like a volume business.

Then AI arrives and exposes the weakness in the old model even more clearly. People can now access useful, immediate, and personalised learning elsewhere. So if L&D’s offer is still generic, scheduled, and content-heavy, why would anyone choose it? They won’t, unless the function is doing something more valuable than delivering material.

This is the ultimate challenge. L&D does not need more praise; it needs a harder standard.

If the function wants to be treated as anything other than a cost centre, it has to stop acting like one. That means getting into the room earlier, defining the capability problem properly, linking work to business outcomes, and being willing to say no to activity that does not build anything of value.

So yes, excellent work can be undervalued, but let’s not hide behind that phrase.

Sometimes the work is good.

Sometimes the function is not.

And sometimes the problem is that L&D has mistaken busyness for influence for too long.