Measuring knowledge retention: 10 KPIs that show what employees really remember
Measuring knowledge retention: 10 KPIs that show what employees really remember
Evaluating a training with the question 'did you like the training?' is easy, but it says little about what employees actually learned and retained. If you want to know whether a training has a long-term effect, you have to look further. Measuring knowledge retention gives organisations insight into what employees still know, can retrieve, and can apply over time. Here are ten KPIs that can help.
1. Immediate knowledge score
Measure knowledge right after the training, for example with 25 questions and an average score of 86%. This gives you a baseline, but it does not yet say how much knowledge is retained later. That is why you need follow-up measurements.
2. Retention score after 30 days
Ask the same or similar core questions after thirty days. If the score goes from 86% right after the training to 74% after 30 days, that is a 12-percentage-point change, and that is far more informative than an evaluation score alone.
3. Retention score after 90 days
Thirty days is interesting, but for many trainings ninety days is even more interesting: does the knowledge stay available in the longer term too? A 90-day measurement can therefore be part of a training programme.
4. Retrieval accuracy
Measure how many questions someone can answer correctly without aids. This distinguishes recognition from actual recall: if someone answers 8 out of 10 questions correctly, the retrieval accuracy is 80%. See also why retrieval practice is so effective.
5. Application score
Having knowledge is not the same as using it. So give scenarios, for example: 'an employee spots a potential safety incident, which three actions does he take?' That way you measure not just factual knowledge, but application.
6. Retention decay
An interesting KPI is the difference between the score right after the training and later, for example:
| Moment | Score |
|---|---|
| Immediately | 90% |
| 30 days | 79% |
| 60 days | 76% |
| 90 days | 74% |
This gives you a retention curve that tells you far more than a single end score.
Not all knowledge fades at the same rate. So analyse it by topic, for example:
| Topic | 30-day score |
|---|---|
| Procedure A | 91% |
| Procedure B | 84% |
| Procedure C | 62% |
| Procedure D | 88% |
Procedure C then deserves extra attention. This makes learning more adaptive, see also how often you should actually repeat knowledge.
8. Recovery rate
What happens after someone gives a wrong answer? Bring the same knowledge back later and measure how many participants get it right the second time. This shows the effect of feedback and repetition.
9. Participation in retention activities
Measure things like open rate, participation, completion, number of questions answered, and time between activities. These are process KPIs: they don't directly tell you whether knowledge is retained, but they can explain why a programme works or doesn't.
10. Real-world impact
The most interesting KPI, ultimately, is: did the learned behaviour actually change? Depending on the training, you could look at fewer mistakes, fewer incidents, better procedure compliance, higher sales results, faster onboarding, or fewer escalations. Not every change can be fully attributed to the training, but looking at knowledge, behaviour, and business results together gives you a far more complete picture, see also how to measure the return of a training.
A simple knowledge retention dashboard
An organisation could, for example, work with a dashboard that shows the score right after the training alongside the score after 30, 60, and 90 days, plus a separate application score. That creates a far more interesting conversation: where are we losing knowledge, which topics are vulnerable, where do we need to repeat, and has behaviour actually changed?
Why this matters for training providers
A training provider can use this to expand its service. Instead of just 'here is the evaluation of the training', the provider can report: 'this is what participants knew right away, what they still know after 30 days, and where extra attention is needed'. That makes training more measurable, and measurability makes the value of training more visible. Stiqqo's retention score per participant is built exactly on this principle.
Conclusion
Measuring knowledge retention doesn't have to be complicated. Start with three moments: right after the training, after 30 days, and after 90 days. Then measure knowledge, recall, application, and behaviour. That shifts training from 'did they attend the course?' to 'what is actually still there three months later?'