Training metrics are the measures an organization uses to judge whether a training program was delivered well, whether people learned, whether they changed how they work, and whether the business got a result. The useful ones sit at the behavior and results end of that chain; most dashboards stop at attendance and satisfaction.
This guide lists the training metrics worth tracking, groups them by what they actually tell you, and explains how to pick a small set that a CEO or CFO will take seriously.
Why most training metrics fail to persuade anyone
The typical L&D report counts completions, hours, and smile-sheet scores. Those numbers describe activity. They do not show that anything changed at work. The gap is well documented: the 2024 LEADx Leadership Development Benchmark Report found a median estimate that only 40% of leadership training is applied on the job, and only 39% of L&D teams measure behavior change at all.
So the problem is rarely a lack of data. It is that the data tracked is the easiest to collect, not the most informative. A short list of behavior and outcome metrics, agreed before the program starts, beats a long list of activity metrics gathered afterward.
The four categories of training metrics
The most widely used structure is the Kirkpatrick model, which sorts evaluation into four levels: reaction, learning, behavior, and results. It works as a filing system for metrics too.
| Category | Example metrics | What it tells you | What it cannot tell you |
|---|---|---|---|
| Delivery and reaction | Enrollment, completion rate, attendance, relevance rating, net promoter score | Whether people showed up and found it worth their time | Whether anything changed after the session |
| Learning | Pre and post assessment scores, skill demonstrations, confidence ratings | Whether knowledge or skill increased | Whether people use it under real pressure |
| Behavior | Manager-observed behavior ratings, 360 feedback change, commitments kept, frequency of target behaviors | Whether the work itself changed | The business value of that change |
| Results | Retention in trained teams, time to productivity, quality, cycle time, engagement scores, internal promotion rate | Whether the organization got the outcome it paid for | On its own, how much of the change came from training |
Use at least one metric from each row, but put the weight on the bottom two. A program with great reaction scores and no behavior change is a pleasant event, not a development investment.
Behavior metrics that are practical to collect
Behavior is where most programs go unmeasured, usually because people assume it requires an expensive study. It does not. A few practical options:
- Manager observation checklists. Define three to five observable behaviors the program targets (for example, “states a clear owner and due date at the end of every meeting”). Managers rate frequency at 30, 60, and 90 days.
- Repeated 360 feedback. A short pulse on the same items before the program and again within a year. A meta-analysis by Smither, London and Reilly (2005) found improvement after multisource feedback was larger when the feedback was repeated.
- Commitment tracking. Count the commitments a participant makes in the program and how many are kept, renegotiated openly, or dropped. This is the most direct measure of accountability there is.
- Work artifacts. Meeting notes, project plans, and one-on-one agendas show whether new habits stuck.
Results metrics: pick the ones the program can plausibly move
The fastest way to lose credibility is to claim that a two-day workshop raised company revenue. Choose results the participants actually influence. For a first-line manager program, that may be team turnover, absenteeism, and engagement. For an executive program, it may be decision speed on cross-functional issues or the share of strategic initiatives delivered on schedule.
Engagement is a common choice because the link to business outcomes is well studied. Gallup’s 2024 Q12 meta-analysis reports that business units in the top quartile of engagement show 23% higher profitability and 78% lower absenteeism than those in the bottom quartile. That does not mean training caused any engagement change, but it gives a defensible reason to track it.
If leadership wants a financial return, see our guide to leadership development ROI, which covers how to isolate the program’s effect before putting a number on it.
How to choose your training metrics in five steps
- Start with the business problem. Write one sentence: “We are running this program because…” The metric should measure that problem.
- Name the behaviors that would fix it. Three to five, observable, specific.
- Set a baseline before the program. Without one, every after number is a guess.
- Pick a comparison. A group that has not yet gone through the program, or the same group’s prior trend.
- Agree on the metrics with the sponsor in advance. A metric chosen afterward looks like a search for good news.
For leadership programs specifically, our article on leadership training effectiveness goes deeper on which indicators separate real development from activity.
The metric most programs never track: agreements
DorWay’s view is that organizations run as a Network of Agreements™. Missed results, slow decisions, and low trust usually trace back to commitments that were vague from the start: no clear owner, no date, no shared definition of done. Training that works shows up as better agreements. People make fewer, clearer commitments, keep more of them, and renegotiate early instead of letting them quietly slip.
That gives you a measurable target. Track the percentage of commitments with a named owner and date, the share kept on time, and the share renegotiated before the due date. The free PACT™ Leadership Assessment gives a quick look at these patterns for an individual leader, and the eight leadership practices describe the behaviors behind them.
FAQ
What are the most important training metrics?
Behavior change and business results. Completion and satisfaction are worth recording, but they only show the program was delivered. Pick at least one observable behavior metric and one outcome metric the participants can plausibly influence.
How do you measure training effectiveness with metrics?
Set a baseline before the program, track the same metrics at 30, 60, and 90 days afterward, and compare against a group that has not been trained yet. The comparison is what lets you attribute change to the program.
What is a good training completion rate?
There is no universal benchmark worth relying on, and completion says little about impact. Treat a low completion rate as a design or relevance signal, then focus measurement effort on whether completers changed how they work.
How many training metrics should we track?
Usually four to six: one or two for reaction and learning, two for behavior, and one or two for results. More than that and nobody reads the report.
