How to measure whether training actually worked
Most organisations measure training by completion rates. X% of employees completed the module. X number of people attended the workshop. Training delivered - box ticked.
This tells you almost nothing useful.
Completion rates measure whether people showed up. They say nothing about whether anyone learned anything, whether behaviour changed, or whether the business problem the training was supposed to solve actually got solved.
Here is how to measure something that actually matters.
---
The four levels worth thinking about.
Donald Kirkpatrick's model has been around since the 1950s and it is still useful because it is simple and honest about what different types of measurement actually tell you.
Level 1 is reaction - did people like the training? Useful for improving delivery but a poor indicator of effectiveness. People can love a training session and learn nothing from it.
Level 2 is learning - did people actually acquire the knowledge or skill? Tests, assessments and practical demonstrations measure this. More useful than reaction scores.
Level 3 is behaviour - are people doing things differently back on the job? This is where it gets harder. You need to observe or measure what people do after the training, not just during it.
Level 4 is results - did the training contribute to a measurable business outcome? Reduced errors, faster onboarding, improved satisfaction scores, higher adoption rates. This is what most organisations say they want to measure and almost none actually do.
---
Why most organisations stop at level 1.
Level 1 is easy. Send a survey after the training, collect the scores, report them upward.
Levels 3 and 4 are harder. They require follow-up. They require baseline data before the training happens. They require someone to care about the outcome weeks or months after the training is done.
They also require honesty about the results. If the training did not change behaviour, level 3 measurement will show that. Not everyone wants to know.
---
What good measurement looks like in practice.
Before the training: agree on what success looks like. Not "people will understand the process" but "error rate on this task will drop by X% within 30 days" or "users will be independently completing this workflow within two weeks of training."
After the training: check whether that happened. Not by asking people if they feel more confident - by looking at what they are actually doing.
This does not have to be complicated. For many training programmes a simple manager check-in at 30 days - are people applying what they learned? - is more valuable than a sophisticated measurement framework nobody uses.
---
The point of measuring is not to justify the training budget.
The point is to find out what is working and what is not - and get better.
Training that gets measured honestly improves over time. Training that only gets measured on completion rates stays the same regardless of whether it is working.
That is the real cost of not measuring properly.
Want to work together?
Let’s start a conversation.