Training evaluation is the process of collecting and analysing evidence to determine whether a training program achieved its goals and what impact it had on the business.
Why it matters#
A short post-course survey is not evaluation. Real evaluation asks whether learners improved, whether they changed their behaviour on the job, and whether the business results the training was meant to drive actually moved. Without that evidence, you cannot defend the program, improve it, or secure future funding. Evaluation is also how you build credibility as a designer — it shows that you care whether training works, not just whether it was delivered.
Key terms#
- Formative evaluation. Evaluates the program during development — materials, exercises, and instructional design. A pilot run is a common example. See formative assessment.
- Summative evaluation. Evaluates the program after delivery — whether it improved business results. See summative assessment.
- Data analysis. Interpreting collected data to draw actionable conclusions. It answers why a result happened, not just what happened.
- Data presentation. Communicating findings to stakeholders — for example, showing before-and-after performance metrics.
- Return on investment (ROI). A financial measure of training value:
(project gain − project cost) / project cost. See Phillips ROI Methodology.
The evaluation process#
- Identify customer expectations. Meet the sponsor at the start of the project. Establish who the customer is, why they want to evaluate, and what they need to know.
- Select appropriate strategies. Decide what to measure, which model(s) to use, how to capture the data, and what success looks like.
- Gain support for the evaluation plan. Summarise the strategy in a clear document covering: project description, measurement goals, evaluation strategy, data collection plan, team members, and a communication plan.
- Manage data collection. Apply consistent methods, monitor collection, and document the data.
- Analyse data. Start with research questions, find data that answers them, and drill down to understand why results occurred.
- Apply learning analytics. Tie the evaluation to organisational goals and needs.
- Make recommendations. Answer whether goals were achieved, what worked well, what should be improved, and what lessons apply elsewhere. Do not hide failures — provide context and a path forward.
Key facts#
- Start measuring at the end of the design phase, not after delivery. Deciding what to measure after the course is built means you may not have collected the right data. Define success criteria before development begins.
- A single post-course survey measures satisfaction, not learning. Reaction data (how learners felt) is the easiest to collect and the least useful on its own. It needs to be paired with learning, behaviour, and results data.
- Isolate training impact from other factors. Performance improvements may be caused by process changes, management behaviour, or market conditions — not the training. Good evaluation designs controls for this.
- Evaluation builds credibility and secures funding. Sponsors need evidence that training works before committing future resources. Without data, the program competes on goodwill rather than results.
- Common models provide a structured approach. The Kirkpatrick Model is the most widely used. The Phillips ROI Methodology extends it with a financial layer. The Brinkerhoff’s Success Case Method takes a different approach by investigating what separates high and low performers.
When to use it#
- At the start of any training project — define the evaluation plan before design begins
- When a sponsor asks whether training is working or worth the cost
- When post-course performance data is not matching expectations