Did the Training Work, or Was It Just a Good Lunch?

"You know how you run a good training? Make sure the broodjes croquette at lunch are nice."

Someone said that to me in a meeting a few years ago, and he meant it. It sounds cynical. The sad part is that he was right: back then, that's pretty much how training got judged.

Level 1: did people like it?

Anyone in L&D will recognise that as Kirkpatrick level 1: did people like it? For a long time, that's where most measurement stopped. You invest a lot of money, and all you really know is that the lunch went down well.

Level 2, with hard data

Meta-Skills simulations will get you to level 2, with hard data. We can show our clients how people perform: across the whole group, per person, and where the growth is. Do people actually get better after five, six, seven rounds? We can show that.

Levels 3 and 4

But when I talk to clients, the question doesn't stop there. You don't train people to score well in a simulation. You train them to behave differently in real conversations. That's level 3. And in the end, you want level 4: more revenue, less churn, happier employees.

Everybody talks about levels 3 and 4 and wants to prove them. Very few actually manage it. Most just put it on a slide.

Looking at the work itself

To really see behaviour change on the work floor, you'd have to look at the work itself: meeting transcripts, calls, emails. That raises the question of how much of your (work) life you'd be willing to share.

So here's that question with a different payoff: would you share your team's meeting transcripts if it finally proved your training worked?