How to Measure Training Transfer: Metrics, Questions and Examples

Oran Cohen

October 17, 2026

17min read

A practical L&D measurement guide

See the practice.
Report the limits.

To measure training transfer, check whether people apply a specific skill in real work after a genuine opportunity. Record the attempt, its evidence and the support needed next. Keep missing updates visible.

Use the worked examples ↓Copy a report outline ↗

Illustrative workplace scene of colleagues reviewing evidence and deciding their next step

Illustrative image. The examples below use fictional data, not customer results.

Start with the question

What is training transfer?

Training transfer means applying learning in the workplace. Learning transfer and transfer of learning are also used for this application; the wider concept includes other settings. Course completion shows participation. A transfer review asks what happened afterwards.

How to measure whether training sticks

BEFORE THE SESSION

Define the action

Choose one observable behaviour and what a useful example would look like.

AFTER AN OPPORTUNITY

Ask what happened

Separate an attempt from an intention, no opportunity and no update.

AT THE REVIEW

Decide the support

Review evidence, gaps and manager support before agreeing what to continue or change.

This first-month approach records practice and follow-through. It does not establish lasting habit formation or prove that training caused a business outcome. Thirty days is a planning window, not a universal evaluation standard.

Choose evidence for the question

Four questions.
Different evidence.

Was the session useful?

Use session feedback to improve relevance and delivery. Perception alone does not establish workplace application.

Can someone demonstrate the skill?

Use an appropriate assessment or observed exercise. A training exercise may differ from normal work.

Did they use it at work?

Review a specific example, work artefact or suitable observation. One example does not establish consistent use.

Did business results improve?

Use a suitable outcome evaluation with a credible comparison. Other changes may explain an apparent improvement.

The CDC’s evaluation guidance distinguishes learning from learning transfer and recommends follow-up after a chance to apply the learning. Choose timing around the topic, resources and work.

Define a starting example and protect sensitive details

After feedback training, choose an action such as preparing one specific observation before a project milestone discussion. Agree what counts as an attempt and what evidence is appropriate. Use the smallest useful anonymised example; do not collect private conversation details or recordings just because they are available.

Record a starting example where possible. If it is recalled after training, label it retrospective. Agree access and confidentiality before collecting evidence. Use participant IDs in the tracker.

Record the evidence source separately from practice status

The kit uses self-report, observation, both or none. An intention or confidence rating is not an attempt. A work document may show preparation without showing conversation quality; an observer’s account may be partial or biased. State those limits.

Use consistent definitions

Missing information
is not a result.

No updateNo usable report has been received. Opportunity and attempts are unknown.
No opportunityThe person reports no relevant chance to practise.
Not attemptedAn opportunity occurred, but the person did not try the action.
Tried onceThe person reports one attempt.
Repeated practiceThe person reports at least two attempts during the review period.

These are the kit’s operational definitions. Repeated practice does not establish automaticity or durable behaviour change.

Choose the denominator

Show the count
behind the percentage.

Update coverage

Usable updates ÷ participants

How complete is the picture? A usable update includes no opportunity, not attempted, tried once or repeated practice.

Reported attempts in the cohort

People reporting an attempt ÷ participants

How many participants reported trying the action at least once?

Attempts among reported opportunities

People reporting an attempt ÷ people reporting an opportunity

Among respondents who had a chance, how many reported trying?

Repeated practice and reviews

Report the counts separately

Count people reporting two or more attempts, and people with a completed review conversation. These measure different parts of follow-through.

Missing updates are excluded from both parts of the opportunity-based attempt rate because opportunity is unknown. If nobody reports an opportunity, show the rate as unavailable. Always report counts alongside percentages; the respondent-only rate cannot represent non-respondents.

Two fictional worked examples

Keep the measures distinct.

The kit example counts reported attempts. The evaluation worksheet example counts reviewed work examples meeting agreed criteria. They answer different questions.

One person in each practice status.

Five participants: one no update, one no opportunity, one not attempted, one tried once and one repeated practice. Two review conversations are complete.

4/5Usable updates · 80%

2/5Reported an attempt · 40% of the cohort

2/3Attempted among reported opportunities · about 67%

Two of five participants reported trying the action. Three reported a practice opportunity, of whom two tried it. One reported no opportunity and one has not supplied an update. Two review conversations are complete.

Calling this “67% adoption” would hide missing information and overstate the measure. One person reports repeated practice; neither that count nor the attempt rate proves a lasting habit.

Both examples are fictional. The five-person counts match the kit example; the twenty-person example matches the evaluation worksheet readme.

Turn the review into a decision

Use low counts
as a question to investigate.

Many missing updates

Check whether the request, channel and reporting effort are workable. Do not assume nobody practised.

Few opportunities to practise

Reconsider timing, access to relevant work or the chosen behaviour. Extend the review window if needed.

Opportunities but few attempts

Ask about confidence, resources, competing work and manager support. A spreadsheet cannot identify the cause on its own.

Attempts with weak or unclear evidence

Improve the example prompt or agree a suitable observation. Record the limits rather than upgrading an unverified report.

Practice reported but reviews skipped

Make review ownership and timing explicit. Agree an owner and next review date.

Copy-ready reporting outline

Make the next decision clear.

Include the objective, review period, coverage, counts, evidence, limits and support decision. Replace every bracketed field before sharing.

Programme and workplace action: [programme and observable action]Review window: [start and end of review window]Coverage: [usable updates] of [cohort size]Opportunity: [reported opportunities]; [no opportunity]; [missing updates]Practice: [tried once]; [repeated practice]; [not attempted]Evidence: [self-report, observation, work examples and limits]Reviews completed: [number]What remains unknown: [gaps, retrospective starting points and alternative explanations]Next decision: [continue, adjust or stop; support action; owner; next review]

If people have had few opportunities, extend the window. If practice is reported, plan a later check of continued use and quality. A first-month scorecard starts an evaluation; it does not replace longer-term assessment or a suitable business-outcome study.

Kirkpatrick Level 3

Ask about use
in the work environment.

Level 3 concerns behaviour and its support in the work environment. A satisfaction score, confidence rating or intention does not establish that behaviour occurred. Use specific, proportionate examples and keep participant accounts separate from observations.

↗Choose Level 3 evaluation questionsCopy participant and manager prompts, then use the observation checklist.↗Organise the evidencePreview blank and example evaluation worksheets before downloading.↗Assign the follow-throughUse the checklist, owners and four adaptable messages.

GWork’s prompts and worksheets are practical aids, not official Kirkpatrick questionnaires or validated assessments. Read the official framework. For action-level formulas and a separate quality check, see the behaviour-KPI example.

Start with one session

Run your first review
with the free kit.

Use the action plan, manager briefing, follow-up messages and Excel tracker to organise practice before reporting it.

Get the free kit →

For help coordinating a cohort, discuss your learning programme with GWork.

How to Measure Training Transfer: Metrics, Questions and Examples

See the practice.
Report the limits.

What is training transfer?