Published 9 September 2026 · Updated 10 October 2026 · By Marie Castelli, Partner, Mentivis
Training metrics: the 7 KPIs every L&D team should track
Seven metrics allow a learning and development (L&D) function to manage the effect of training rather than its volume: the on-the-job application rate at 90 days, measured progress before and after training, coverage of critical skills, cost per participant who reached the objective, adjusted completion rate, learners' recommendation score and the share of budget aligned with strategic priorities. Their thresholds are set by each organisation, case by case.
Pillar: Skills measurement: methods, tools and metrics
Why are volume metrics not enough?
Training hours, number of trainees, budget spent: these metrics measure training effort. They are necessary for management and reporting, but they do not say whether training had an effect.
The Kirkpatrick model distinguishes four levels of evaluation: reaction, learning, behaviour at work, results. The ROI Institute's ROI Methodology adds a fifth level, return on investment. The seven metrics below cover these levels, which avoids steering from a single angle.
Overview
| Metric | What it measures | Level (Kirkpatrick / ROI Institute) |
|---|---|---|
| 1. Application rate at 90 days | Use on the job | 3. Behaviour / application |
| 2. Before/after progress | Acquisition | 2. Learning |
| 3. Critical skills coverage | Skills risk | 2 and 3, aggregated |
| 4. Cost per participant who reached the objective | Efficiency | Cost relative to level 2 |
| 5. Adjusted completion rate | Real engagement | Activity |
| 6. Recommendation score | Perceived quality | 1. Reaction |
| 7. Strategic alignment of budget | Portfolio relevance | Steering |
The 7 metrics in detail
1. Application rate at 90 days.
Definition: share of participants who use on the job what they learned, about three months after training.
Measurement: short questionnaire to the participant and their manager (three to five questions on specific behaviours), ideally with a concrete example.
Limit: self-reported; it becomes more reliable when the manager confirms.
2. Measured progress before and after training.
Definition: difference between the level assessed before training and the level assessed after, on the targeted skills.
Measurement: the same criterion-referenced assessment as pre-test and post-test, then repeated a few weeks later to measure retention. See measuring skills acquisition.
Limit: an immediate post-test overestimates retention.
3. Critical skills coverage rate.
Definition: share of the people concerned who reach the required level on skills considered critical for the business.
Measurement: cross-reference assessment results with the skills map or a skills matrix.
Value: it reveals skills held by only one or two people.
4. Cost per participant who reached the objective.
Definition: full cost of a programme (design, purchase, time spent) divided by the number of participants who reached the target level in the post-test.
Measurement: requires metric 2.
Value: a programme that is cheap per hour but has no measurable effect can turn out more expensive, relative to results, than a more expensive and effective one.
5. Adjusted completion rate.
Definition: share of learning paths completed with real engagement, excluding "mechanical" completions (abnormally short sessions, no interaction, fast-forwarding).
Measurement: analysis of learning platform logs, with written and stable exclusion rules.
Limit: it remains an activity metric; it does not replace metric 2.
6. Learner recommendation score (NPS).
Definition: on a scale of 0 to 10, the likelihood of recommending the training to a colleague. According to Bain & Company, the Net Promoter Score is calculated by subtracting the percentage of detractors (scores of 0 to 6) from the percentage of promoters (scores of 9 or 10).
Limit: it measures perceived quality (level 1). A well-liked course is not necessarily effective; this score should be read alongside metrics 1 and 2. No benchmark threshold is agreed for training: compare your programmes with one another and over time instead.
7. Strategic alignment of the portfolio.
Definition: share of the training budget spent on programmes directly linked to strategic priorities, as distinct from mandatory and skills-maintenance training.
Measurement: classify each programme in the skills development plan according to its link with strategic priorities. In France, in companies with at least 300 employees, the negotiation on GEPP (strategic workforce and career-path planning) covers in particular the three-year training orientations (Article L2242-20 of the Labour Code), which provides a classification framework.
Limit: mandatory training is not "non-strategic"; it belongs to a separate category.
How should targets be set?
There is no public standard setting thresholds for these metrics. An organisation may decide, for example, that a majority of participants should apply what they learned in high-stakes programmes, or that a minimum share of the budget should be aligned with its priorities: these choices are its own and are not benchmarks. The most reliable approach is to measure a first period, then set a progress target from that baseline.
Where to start?
Start with two metrics: before/after progress (2) and application rate (1), on three or four high-stakes programmes. These feed the calculation of training plan ROI. The others can be added once data collection is stable.