Training is a nice-to-have until it makes money

I wrote most of this by hand, in the Radcliffe Camera, in May last year. No laptop, no deck, no plan to publish it. Just a pen and the question I could not put down: is what we are building actually the biggest thing we could be building?

I still do not have a clean answer. But the argument I wrote that afternoon has held up better than most of what I have thought since, so here it is roughly as it came out.

Nobody loses sleep over the training budget

Ask a COO what keeps them up at night. Training will not be on the list.

That is not a criticism of L&D. It is a description of where training sits in most organizations. It is a cost line defended every budget cycle, run by a team everyone likes and nobody consults before a decision gets made. When the year gets hard, it is among the first things cut, and the business carries on more or less as before.

That last part is the uncomfortable bit. You can remove most corporate training tomorrow and the quarter still closes.

We built something that makes training faster and cheaper to produce. That is real, and teams pay us for it. But cheaper training is an argument about a cost line. It is not an argument about the business.

Companies care about two things

Revenue up, costs down. Almost everything else is a proxy for one of those.

Training reduces costs, so it clears the second bar. But it clears it the way a better supplier contract clears it: worth doing, not worth a meeting. Nobody has ever been promoted for producing a course more cheaply.

If your entire pitch is efficiency, you are competing to be somebody's smallest problem. You will win some of those fights. You will never be the reason a company changes how it operates.

The problem underneath the training

Here is what I kept circling back to that afternoon.

Nobody wants training. They want the job done correctly.

Training is a mechanism people reach for when there is a gap between how work is being done and how the company needs it done. A new policy lands and nobody follows it. A product changes and half the sales team keeps describing the old version. Someone joins and takes four months to become genuinely useful.

In every one of those cases, the real problem is a gap in the work. A course is one response to it. It has never been obvious that it is the best response. It has only ever been the one that was available.

That reframing changes what you think you are building. Not a faster way to produce courses, but something that sits closer to the work itself, notices where it is going wrong, and closes the gap in whatever way actually closes it. Sometimes that is a lesson. Often it is not.

Being a platform is fine for now

We are a platform today, and I have made my peace with that.

You do not get to skip the step where you build the useful, unglamorous thing people will pay for. Every company I admire spent years being something smaller and more boring than what it became. Nobody had a script. They just did not die, and they kept finding the next use case.

But the direction matters more than the current shape. The question worth chasing is not how to produce training faster. It is what happens when the system that trains your workforce is also the system that understands how the work is going, and can tell the difference between somebody who needs a lesson and a process that needs fixing.

We are not there. As far as I can tell, nobody is. But that is the version worth building, and it is a long way from generating courses.

Written in the Radcliffe Camera, Oxford, May 2025. Lightly edited.