How to Prove Leadership Training Works
Completion and satisfaction don't prove managers can do the job. A clinical psychologist on the missing rung, and what to report instead.
Home › Resources › Proving leadership training works
The short answer
Most leadership programs can show who attended and whether they liked it. Few can show whether managers can actually handle the conversations the training was for. The missing evidence sits between learning and on-the-job behaviour: practice held to a standard. Measure that, add one business signal you already collect, and you can report capability rather than completion.
"Where's the business case?"
Two HR leaders said versions of the same thing to me within a week of each other. The first, after looking at how she reports leadership development to her executive team, asked: where's the business case? Not the attendance numbers, which she already had. The evidence that anything changed.
The second put it more simply. She wanted to know where the gaps were, and whether anything was closing them. Without that, she said, she had nothing to act on.
Neither of them doubted that training matters. They doubted that they could show it. In my clinical work, a person telling me they understand something and that person being able to do it when it counts are two very different pieces of evidence. Most leadership reporting only collects the first.
What counts as proof: the Kirkpatrick levels
Direct answer: the most widely used training evaluation model has four levels: Reaction, Learning, Behaviour and Results. Most leadership programs report the first two. The third is the one that matters.
Kirkpatrick's four levels
| Level | Question it answers | Typical evidence | How often it's measured |
|---|---|---|---|
| 1. Reaction | Did participants find it useful? | Satisfaction survey ("happy sheet") | Almost always |
| 2. Learning | Did they gain the knowledge or skill? | Quiz, knowledge check, completion | Usually |
| 3. Behaviour | Do they apply it on the job? | Observation, follow-up survey, 360 | Rarely |
| 4. Results | Did business outcomes change? | Engagement, turnover, escalations, performance | Occasionally, with weak attribution |
Level 3 is skipped for understandable reasons. The behaviour that matters most for managers, a performance conversation or a pay "no", happens in private, rarely and unpredictably. You can't send an observer into a one-on-one without changing it. So L&D reports what it can see, and the organisation infers the rest.
Completion tells you someone was there. It doesn't tell you they can do it.
Why completion and satisfaction aren't proof
Direct answer: completion and satisfaction measure exposure to training, not the ability to use it under pressure.
This matters because managers have an outsized effect on how people experience work. Gallup estimates that managers account for at least 70% of the variance in employee engagement scores across business units [1]. If the manager is the lever, then "attended the program" is a weak thing to know about the lever.
There's also a well-documented gap between learning in a program and doing it afterwards. Beer, Finnström and Schrader argued in Harvard Business Review that people often learn in the program, then return to a system that doesn't support the new behaviour, and it fades [2]. The classroom wasn't the constraint. The day after was.
And feedback itself isn't automatically helpful. Kluger and DeNisi's meta-analysis found that feedback interventions improved performance on average but made it worse in more than a third of cases, particularly when the feedback drew attention to the person rather than the task [3]. A competency rating delivered months after a program is a comment on the person. It's a weak tool for changing what someone says in a room.
A program can have excellent completion, glowing satisfaction scores and no change in how managers handle hard conversations. Those numbers aren't wrong. They're just answering a different question.
Why managers don't use what they learn
Direct answer: usually because of timing, relevance and safety, not motivation.
Why the learning doesn't show up at work
| Barrier | What it looks like | What helps |
|---|---|---|
| Timing | The workshop was in March; the hard conversation is in September | Support at the moment of need, shortly before the conversation |
| Relevance | Generic scenarios that don't match their team, their person or their situation | Practise the conversation they actually have |
| No safe first attempt | The first real try is with the employee | Private rehearsal with a counterpart who pushes back |
| Nobody senior goes first | It reads as remedial, or optional | A senior leader who practises visibly and says so |
| Fear of being watched | Managers suspect HR will see their attempts | A written promise that practice isn't an HR file |
| The system doesn't change | Back at work, nothing expects or rewards the new behaviour | Expectations set in the role, followed up by the manager's own leader |
The People leaders we speak with keep making the same point: managers won't practise every day, but many will practise once, right before the conversation that matters. Adoption improves when training is designed around that moment rather than around a completion deadline.
The proof ladder: the missing rung
Direct answer: between "learnt it" and "did it on the job" there's a rung most programs skip: practised it to a standard. It's measurable, private and close enough to the real conversation to be a useful leading indicator.
A five-rung ladder, bottom to top: Attended, Learnt, Practised to a standard (highlighted as the missing rung), Used it in the real conversation, Outcome moved.
- Attended. They were there. (Kirkpatrick Level 1 territory)
- Learnt. They can describe the framework. (Level 2)
- Practised to a standard. They said it out loud, in the specific conversation, against a counterpart who pushed back, and met clear criteria for that conversation, retrying until they did. This is the missing rung.
- Used it in the real conversation. They applied it on the job. (Level 3)
- Outcome moved. Engagement, escalations, retention or performance shifted. (Level 4)
Rung 3 isn't the same as rung 4. A rehearsal isn't a recording of Monday, and it shouldn't be sold as one. But it's much closer to rung 4 than a quiz is, and unlike rung 4 you can measure it without anyone sitting in the room.
Look at your current leadership program report. For each number on it, write which rung it measures. If everything is on rungs 1 and 2, you know where the gap is.
What "to a standard" means
A standard is specific to the conversation. For a performance feedback conversation, it might include:
- Was the behaviour described specifically, not as a trait?
- Was the impact made clear?
- Was the manager direct and caring?
- Did they invite the employee's view?
- Did they agree a next step with a date?
- Did they avoid burying the message in praise?
That's very different from scoring tone, pace or filler words. It answers whether the conversation would have worked, not just how the manager sounded.
For a worked example, see how to give performance feedback.
What to report to your executive team
Direct answer: report practice, capability and one business signal, with honest limits. Not completion alone.
One-page report to your exec
| Section | What to include | Rung |
|---|---|---|
| The problem | One sentence: which conversations we needed managers to handle better, and why | — |
| Reach | How many people leaders took part; which conversations they practised | 1–2 |
| Capability | Share of practice conversations that reached the agreed standard; how results moved between first attempt and retry | 3 |
| Common gaps | The two or three most frequent gaps across the group (for example, "next steps rarely include a date") | 3 |
| Manager voice | Short pulse: "I felt more prepared for the real conversation" | 3–4 |
| Business signal | One measure you already collect (manager-related HR escalations, exit-interview themes or an engagement item about managers), with a note on attribution | 4 |
| What we'll do next | Expand, adjust or stop, and why | — |
Template. Select the table and copy it. No email.
Be honest about attribution. A twelve-week change in escalations or engagement has many causes. Present it as a signal alongside the capability evidence, not as proof that training alone caused it. Executives trust a report more when it states its own limits.
For how to set these measures before a rollout, see implementing AI roleplay for managers.
Where Evro fits
Direct answer: Evro gives HR the missing rung: practice on the real conversation, scored to a standard, without anyone watching the manager.
A manager picks the conversation they're about to have, prepares with evidence-based frames, roleplays it with an AI counterpart who pushes back, is scored against a weighted rubric for that conversation, and retries until they're at the line.
The organisation gets evidence that critical conversations were practised to a standard, and where the common gaps are. It doesn't get transcripts, and it doesn't get a ranking of who failed.
Practice isn't an HR file. Evro doesn't send session content or transcripts to HR. The live conversation stays private too; Evro doesn't listen to it. What you can see is whether managers can prepare for it, and whether practice moved them.
For the questions to ask any vendor before you buy, see 8 questions to ask any AI roleplay vendor.
Frequently asked questions
How do you measure the effectiveness of leadership training?
Go beyond attendance and satisfaction. Measure whether managers can perform the specific behaviours the training targeted, ideally by having them practise real conversations against a clear standard, then track one on-the-job or business signal you already collect, such as manager-related HR escalations or engagement items about managers.
What are the four Kirkpatrick levels?
Level 1 Reaction (did participants find it useful), Level 2 Learning (did they gain knowledge or skill), Level 3 Behaviour (do they apply it on the job) and Level 4 Results (did business outcomes change). Most leadership programs report Levels 1 and 2; Level 3 is the one that matters most and is most often skipped.
Why aren't completion rates proof that training worked?
Completion shows someone finished the content. It does not show they can use it under pressure, in a real conversation, with a person who pushes back. A program can have excellent completion and satisfaction while manager behaviour stays the same.
Why do managers not use what they learn in leadership training?
Common reasons are timing (the training comes months before the conversation), relevance (generic scenarios rather than their real ones), a lack of safe practice (the first real attempt is with the employee) and a system that does not support the new behaviour once they return to work.
How can HR prove manager capability without watching managers?
Use practice held to a standard as a leading indicator. Managers rehearse the specific conversation privately and are scored against criteria for it; HR sees aggregate results, such as which conversations reach the standard and where common gaps are, without seeing what any individual said.
What is Evro?
Evro is an AI platform for people leaders in Australian organisations. Managers prepare for a real conversation, such as performance feedback, rehearse it with an AI counterpart, are scored against a weighted rubric for that conversation and retry until they reach the standard. Practice sessions are not an HR file.
A last thought
Leadership development has run on faith for a long time: spend on the program, wait for engagement and hope the two are connected. That's not because L&D teams are doing it badly. It's because the moment that matters has always been out of sight.
We still can't sit in the room, and we shouldn't. But we can know whether a manager could handle the conversation before they walked in. That might be enough to stop hoping.
Sources
- Gallup. Why great managers are so rare. Managers account for at least 70% of the variance in employee engagement scores across business units. gallup.com/workplace/231593/why-great-managers-rare
- Beer, M., Finnström, M. and Schrader, D. (2016). Why leadership training fails — and what to do about it. Harvard Business Review. hbr.org/2016/10/why-leadership-training-fails-and-what-to-do-about-it
- Kluger, A. N. and DeNisi, A. (1996). The effects of feedback interventions on performance: a historical review, a meta-analysis, and a preliminary feedback intervention theory. Psychological Bulletin, 119(2), 254–284. psycnet.apa.org/record/1996-02773-003
- Kirkpatrick's four levels of training evaluation (Reaction, Learning, Behaviour, Results) are the common frame used throughout. Back to the table.
About Evro. Evro helps people leaders in Australian organisations rehearse the conversations that matter (performance feedback, coaching high performers, pay and promotion conversations, team conflict) and practise them to a standard before they happen. Managers prepare, roleplay with an AI counterpart, are scored against a rubric for that conversation and retry until they're ready. Practice sessions are not an HR file.