Training budgets require justification in almost every organization. But when it comes time to prove what the training was worth, most L&D teams fall back on completion rates and satisfaction scores — neither of which tells you much about actual impact. Here's a more useful framework.
Why the usual metrics don't work
Completion rates measure compliance, not learning. A team can complete 100% of a training and come out no more capable than when they started. Satisfaction scores measure how people felt about the experience, which correlates loosely with learning and not at all with behavior change.
The Kirkpatrick model, which most training programs reference, has four levels: reaction, learning, behavior, results. Most programs measure level one (reaction) and stop there. Levels three and four — behavior change and business results — are harder to measure, which is exactly why they tend to get skipped.
A practical approach to measurement
Before training starts, identify two or three specific behaviors that should change as a result of the training. Not "employees will understand the new system" but "employees will correctly categorize expense reports without supervisor review." Specific, observable, tied to a real work output.
After training, wait four to six weeks and check whether those behaviors have changed. This requires manager observation or output data, not a survey. It's more work, but it's the only way to know if the training actually transferred.
When a business outcome is measurable
Sometimes you can connect training directly to a metric. If you train customer service reps on a new handling protocol and first-call resolution improves by eight points the following month, you have a signal. It's not controlled proof, but it's directionally useful. Document these connections when they're available — they're the language finance teams respond to.
Most training won't have clean causal links to financial metrics. That's fine. The goal is better evidence, not perfect evidence. Even showing that behaviors changed — level three — is more defensible than completion rates.