A company sends a new cohort through its onboarding program and the program gets good reviews. The survey at the end says the content was clear and the facilitator was engaging, and the score on the closing quiz is high. Six months in, the work the cohort is doing looks about like the work last year’s cohort did before they went through the same program. Something between the training and the work isn’t connecting, and nobody has a clear view of what.
Companies invest heavily in training and almost nothing in measuring whether the training landed. The spending on content and the people who deliver it is large and visible. The spending on measurement is small and hidden, and most of what gets called measurement is a survey that the learner fills out while the facilitator is still in the room, asking how the session went. The answer to that question doesn’t tell anyone whether the learner came out able to do something they couldn’t do before.
The content and the measurement are different jobs. The content job is to put the material in front of the learner in a form that’s easy to follow. The measurement job is to check, after enough time has passed for the memory of the session to fade, whether the learner can do the thing the content was teaching in the real work. Companies staff the first job heavily and the second job barely at all, because the first is legible and the second is work whose payoff only shows up if you build the measurement, which most companies haven’t.
The better edtech businesses ran into this problem years ago, because when the paying customer is also the learner, failure shows up faster. A learner who doesn’t come out of a course able to do the thing it promised stops paying, and the platform has a strong reason to build the loop that tells it where the course is failing. That forcing function pulled measurement into the product in a way corporate training rarely has.
Corporate training rarely has such a forcing function. The paying customer is the employer and the learner is the user, and the gap between the two is where the measurement falls through. The employer doesn’t see whether the training landed because they’re not the one doing the work, and the learner has no strong reason to report that the training didn’t land, because the relationship with the employer isn’t transactional in the way a subscription is. The training gets checked on delivery and never on outcome.
A working measurement for a corporate training program is taken on the job, not in the training. It’s a read, done some weeks after the session, of whether the work the training was meant to improve has improved. The read is designed around a specific thing the learner should now be able to do, and it’s designed before the training runs, so the training can be built around the thing the read is going to check. A training whose measurement is the survey at the end of the session is a training with no read, because the survey is asking about the session rather than the work.
The operator lesson from edtech is that a training without a measurement is a content library, and a content library isn’t a training. The distinction sounds pedantic until you look at the budget and see that most of what a company spends on learning is going to the library rather than the training. A program that runs on measurement looks different from one that doesn’t: fewer sessions, narrower content, longer checks afterwards, and honest conversations with the business about what the training is meant to change. A program that runs on delivery alone looks impressive in the catalog and makes little difference in the work.
Companies haven’t made the shift yet because the shift is expensive and the current arrangement works for everyone who runs it. L&D runs delivery and reports on what’s easy to count: sessions delivered and the scores they got afterwards. Tying training to work outcomes makes L&D accountable for something harder to manage, and makes the business responsible for describing the outcome in enough detail that it can be measured, which is a meeting nobody wants to have. Both sides hold the current shape in place, and the training budget goes on buying content that nobody is looking at the results of.