Measuring Training Results: Skills, Performance and Realistic Evaluation

Training evaluation is most useful when it helps an employer decide what to continue, change or support next. A high satisfaction score can tell you that people liked an event, but it does not by itself establish that they learned the required skill or used it at work.

This guide offers a proportionate approach for UK employers. It uses fictional examples and avoids promising a universal return on investment. The aim is to gather credible evidence connected with the original need while recognising other influences on performance.

Start with the decision you need to make

Ask what the evaluation will inform. Are you deciding whether to repeat a course, change its content, provide more practice or remove a workplace obstacle? A clear decision helps you choose evidence that matters and avoid collecting data simply because a platform makes it available.

CIPD’s public evaluation overview links effective evaluation with identified performance gaps and business objectives. Use our training needs analysis guide to recover the original problem before selecting measures. If the intended improvement was never defined, clarify it rather than inventing a success story afterwards.

Record a useful starting point

Choose evidence from before the learning that relates to the task. This could be a short work sample, observation or review of a defined process. Explain how it was selected and its limitations. A few examples can guide improvement without pretending to be a precise measure of the entire organisation.

For a fictional handover-note programme, a manager might review ten notes using the same five required elements. Record which elements are missing and whether the staff had access to the current template. The exercise may reveal a training need, a process issue or both.

Keep the purpose transparent to staff and collect only information needed for it. Avoid turning a development exercise into unexpected individual surveillance. Use appropriate privacy and employment processes where personal information is involved, and explain how the evidence will be used.

Distinguish experience, learning and application

Ask learners about the experience where that helps improve delivery: whether instructions were clear, access worked and practice was useful. Separately, check whether they can perform the intended task through an appropriate exercise. Later, examine whether they can apply it in their actual work.

These questions are connected but not interchangeable. Someone may enjoy a session yet need more practice. Someone may pass an exercise but lack permission or resources to use the method at work. Understanding the difference helps you choose the right response instead of automatically buying more training.

Choose a small number of meaningful measures

Prefer a few measures that answer the decision question. If the goal is complete handover notes, assess completeness using a clear checklist. Do not substitute course attendance because it is easier to count. Attendance can describe participation, but it cannot establish that the task is performed well.

Define each measure so another person could use it consistently. State the sample, period and criteria. For example, “notes containing all five agreed elements” is clearer than “better communication”. Check that the measure does not encourage staff to optimise a number while neglecting the real purpose.

Use our provider brief guide to agree evidence requirements before delivery. This allows the trainer to include appropriate practice and feedback, and makes the employer’s follow-up responsibilities visible from the start.

Interpret change without overstating causation

Suppose a fictional sample improves from six complete notes out of ten to eight out of ten. Report the sample sizes and the observed change. Do not automatically claim that training caused a company-wide improvement or calculate a financial return from those twenty observations alone.

Ask what else changed: the template, workload, staffing, software or manager attention. Check whether the samples are reasonably comparable. A later review can help determine whether the improvement persists, but it still needs to be interpreted in context.

Where a financial estimate is useful, separate measured values from assumptions. Explain the cost included and how any benefit was estimated. Avoid presenting a precise return figure when the underlying time savings, attribution or future usage remain uncertain.

Investigate barriers to using the learning

Speak with learners and managers about the first attempts at work. Ask for examples of where the new approach helped and where it was difficult to use. Look for missing resources, conflicting procedures or lack of practice before concluding that the course failed.

Our staff development plan guide helps turn that discussion into actions. A manager might arrange another supervised task, correct a template or clarify an escalation route. The response should follow the evidence rather than repeat the original course by default.

Close the loop with a practical decision

Summarise the original objective, evidence gathered, limitations and next action. State who will act and when the result will be reviewed. Share useful findings with the provider where appropriate so that future delivery can improve.

Keep the evaluation proportionate. A short internal workshop may need a simple task check and manager conversation, while a large programme may justify a more structured approach. The test is whether the evidence supports a better decision, not how many pages the report contains.

1. Define. Name the task, starting evidence and decision.
2. Check. Examine learning, application and relevant context.
3. Act. Change support or delivery and review the result.

Common questions

Is a satisfaction survey enough?

It can help assess the experience, but it does not answer every question. Add evidence that matches the intended skill or work outcome. Keep the survey short and use the results to improve something specific rather than treating the score as proof of impact.

Do we need a financial ROI calculation?

Not for every programme. Start with the decision and available evidence. If a financial estimate is useful, state assumptions and attribution limits clearly. A modest credible account of improvement is better than a precise-looking return based on unsupported guesses.

How soon should we review application?

Choose a point when staff have had a realistic chance to use the skill. A task performed daily and one used quarterly need different timing. Agree the review in advance and account for whether relevant work was actually available.

What if performance does not improve?

Investigate the reason before choosing a response. Check learning, practice opportunities, resources and process barriers. More training may help, but the issue may also be that staff cannot apply the method under the current working conditions.

Can a small sample still be useful?

Yes, when its purpose and limits are clear. A few work examples can identify questions or guide coaching. Report the sample size and avoid presenting it as definitive evidence about every employee or the whole organisation.

Who should collect the evidence?

Choose someone who understands the task and can apply the criteria consistently. Clarify the purpose and appropriate handling of personal information. Involve learners and managers in interpreting findings so that the numbers are considered alongside real working conditions.

How do we avoid measuring too much?

Link each measure to a decision. If you cannot explain how the information will change an action, reconsider collecting it. Use existing appropriate evidence where possible and keep the burden on learners and managers proportionate.

What should the final evaluation note contain?

State the objective, evidence, observed result, limitations and agreed next step. Identify responsibility and review timing. The note should make it easy to act, rather than simply document that a training event took place.

Sources and further reading

• CIPD: evaluation, impact and transfer, public overview
• CIPD: learning needs, public overview