Kirkpatrick's four-level evaluation model is taught in every L&D certification and used in almost no real HR reports. This guide skips the academic framework and shows you four metrics that work in practice — what to measure, when, and how to report it in a quarterly review.
Why post-training "happy sheets" don't count
Day-0 satisfaction surveys ("rate the trainer / venue / materials") measure how the participants felt walking out the door. They do not measure outcome change, and they correlate weakly with whether the training delivered any business value. Use them for trainer quality control, not for ROI.
Metric 1 — Behaviour-change observation at Day 60
Sixty days after training, the manager observes whether the participant is doing the new behaviour in their day-to-day work. For sales training: are reps using the new qualification questions in discovery calls? For leadership training: are managers running weekly 1:1s with a written agenda? This is the highest-signal metric and the one most often skipped.
How to set it up
- Before training: agree 3-5 specific observable behaviours with the manager
- Day 60: manager scores each behaviour 0/1/2 (not visible / partial / consistent) for each participant
- Report the average score per cohort and the trend month-on-month
Metric 2 — Lagging business indicator tied to the training outcome
Tie one lagging indicator to the training outcome — and only one. For sales training: revenue per rep, deal velocity, or pipeline conversion at the targeted stage. For leadership training: skip-level engagement scores, voluntary turnover in the trained manager's team, or 1:1 completion rate from calendar data. Resist the urge to measure five things; one lagging indicator with clear baseline and 90-day window is more credible than five with noisy data.
Metric 3 — Application rate of HRDF levy spend
For HRDF-registered employers, this is the metric finance directors actually ask about: of the annual levy contribution, what percentage is being claimed back through training? Levy utilisation below 60% is wasted; above 90% is exemplary. Track quarterly and report against the annual levy contribution estimate (see /tools/hrdf-calculator).
Metric 4 — Cost per behaviour-changed participant
Take the all-in training cost (provider fee + materials + venue + F&B + participant time at internal day rate) and divide by the number of participants who scored 2 on the Day-60 behaviour observation (consistent). This is your real cost per outcome. A cheap training that delivers behaviour change in 1 in 10 participants is more expensive than a premium training that delivers behaviour change in 7 in 10.
What a quarterly L&D review looks like with these four metrics
A one-page report per programme: (1) Day-60 behaviour score chart, (2) lagging indicator pre-training baseline vs. 90-day post-training, (3) levy utilisation year-to-date, (4) cost per behaviour-changed participant by programme. That is enough to defend the L&D budget and to make renewal-or-replace decisions on each provider you work with.
How FindTraining helps
When sourcing on findtraining.com, ask each shortlisted provider which of the four metrics above they will help you set up. Providers who include behaviour-change observation in their delivery (rather than just leaving you to do it) outperform on Metric 4 by a wide margin.