Skip to main content
UPSIDE GMS
Training

Measuring training that actually works

Four levels of evaluation, and why most programmes never get past the first one.

Published
27 Feb 2025
Category
Training
Duration
6 min read

Kirkpatrick's model is nearly seventy years old and still widely cited, but most organisations implement only its first level: did participants enjoy it. That measures delivery quality, not effect.

The second level costs almost nothing

A structured pre- and post-assessment against defined learning outcomes is inexpensive and immediately converts a training programme from an activity into an intervention with a result.

Behaviour is the honest measure

Level three asks whether practice changed in the workplace, which requires observation some weeks after delivery. It is more effort, and it is the level at which weak programmes are exposed.

Tie it to an operational indicator

Level four connects the programme to a business result. It is rarely attributable cleanly, but naming the indicator in advance — rework rate, incident frequency, approval cycle time — imposes a useful discipline on the design of the programme itself.

27 Feb 2025 — 6 min read Back to Insights

Tell us what you are trying to build

Send a brief outline of the project, programme or capability gap. You will hear back from the discipline lead responsible — not a sales desk.

Quick contact