More activity is not the same as better work performance. Start with a relevant task, identify where people struggle and test one improvement to the training unit.
Choose a task that matters at work
Identify a decision or procedure the training is meant to support. Ask a small group to attempt a representative task before revising the lesson. Look for unclear instructions, missing prerequisites or examples that do not match their work.
Do not assume low participation is caused only by uninteresting content. Time, access, workload and relevance may also need investigation.
Change one part of the unit
Add a worked example, clarify a question or improve feedback on a common error. Give people a chance to apply the explanation, not just acknowledge that they viewed it.
If you use AI assistance, keep subject review and an escalation route to a responsible person. Do not assume a generated answer is accurate or that a tool automatically delivers a personalised curriculum.
Keep the measures separate
Record attendance, completed practice and performance on the relevant task as different observations. Ask for feedback about the course’s usefulness and access barriers.
A comparison between two groups may also reflect different starting knowledge or working conditions. Describe those limits when reporting a result.
Make the next revision manageable
Assign an owner for each correction and retain the previous version for reference. Avoid adding badges, reminders or media unless they address an observed problem.
Review the analytics workflow and feedback features against the measures you need. Confirm what the actual configuration captures. The aim is a useful, reviewable training process, not a software promise of higher engagement or productivity.
