Recorded video is the cheapest part of digital learning and the part least connected to whether anyone learns anything. Learners stop at a particular module for a particular reason, and that point is visible in the data.
This guide covers how to find the stopping point, what the people who stopped will tell you, and why fixing one module usually moves completion further than producing a new course. Most of the budget in a learning project goes on production, and most of the failure happens somewhere else entirely, which is the uncomfortable part of looking at the data honestly.
Key takeaways
- Find the module where people stop: Learners quit at a specific point for a specific reason, and more video on top of it adds cost only.
- Ask ten who stopped: Their answers are about length, timing or a missing example rather than production quality.
- Feedback delay decides learning: The gap between a misunderstanding and finding out about it is usually the cheapest thing to improve.
- Count the chasing: If an administrator spends their week chasing stalled learners, that is what the platform is actually for.
Look for the business signal behind the technology
The signal is in the completion curve rather than in the feedback form. Learners stop at a specific module for a specific reason, and that point is visible in the data long before anybody says anything about the course.
Low completion is not always a problem. People take what they need and leave, and a learner who watched the three modules relevant to their job has succeeded by any sensible measure. Separate deliberate partial use from the drop-off that happens at one blocked step.
Contact ten people who stopped and ask what they were doing at the time. The answers are usually about length, timing or a missing example rather than about production quality, which is where most of the budget normally goes.
Use four lenses to understand the opportunity
Learner
Describe who the learner is, when in the day they study, and what interrupts them on a typical attempt. A course designed for a quiet evening and attempted in a lunch break fails for reasons no amount of production value will fix.
Content
Decide the smallest useful unit of content, how long it runs, and whether it can be resumed on a different device. Learners rarely finish anything in one sitting, and a module that loses progress on a switch from laptop to phone loses the learner.
Feedback
Work out how someone finds out they have misunderstood something and how long that takes. Feedback delay is the difference between a course that teaches and a course that is watched, and it is usually the cheapest thing on this list to improve.
Operations
Define what an instructor or administrator does each week and how much of it is chasing people who have stalled. If that work is manual, the platform has a job to do that has nothing to do with playing video.
Assess the current state without buying software first
Plot completion module by module for one existing course. You do not need a platform to do it, and the shape of the line tells you whether the problem is at the start, at one particular module, or spread evenly, which point at three different fixes.
Decide what useful measurement looks like
Measure completion by cohort rather than overall, so you can tell whether a change helped. A single percentage mixes people who joined for different reasons in different months and cannot show improvement.
Pair completion with whether anything changed afterwards. A course finished and forgotten has produced a certificate, and if the business commissioned it to change how work is done, that is the figure that matters.
When software is likely to help
Platform work pays where the operational load is real: chasing stalled learners, managing cohorts, issuing certificates, reporting to a regulator. Those are repeated administrative tasks and software removes them reliably.
It is the wrong investment when the content itself is the problem. A better player does not fix a module that assumes knowledge the learner has not got, and no amount of progress tracking makes an unclear explanation clear.
For a comparison with live, synchronous formats and what they cost to run, see Video Call App Development: Features, Risks, and Roadmap.
Start with a bounded improvement
Rework the single module where people stop, release it to a new cohort, and compare their completion at the same point. One fixed blockage usually moves the overall figure more than a new course would, at a fraction of the cost.
Separate a structural constraint from a temporary problem
Check whether the drop-off is structural before rebuilding anything. Does it happen with every cohort, at every time of year, with every instructor? A pattern that only appears in one group is a scheduling or a facilitation problem rather than a content one.
Lay it out in a grid: motivated and obliged learners on one axis, quiet and busy periods on the other. Look at how the learner, the content, the feedback and the operations behave in each square. Compulsory training in a busy month is usually the square that fails.
Where the problem is structural, write the improvement so it can be disproved: completion at module four will rise for the next two cohorts. Then check. A learning project justified by engagement generally is justified by nothing measurable.
Risks to make visible
| Lens | Common risk | Early response |
|---|---|---|
| Learner | The learner has no time to study | Ask when people actually attempt it and match the unit length to that. |
| Content | More content is produced before the blockage is found | Plot completion module by module and fix the stopping point first. |
| Feedback | Misunderstanding surfaces too late | Shorten the loop so a learner finds out within the session. |
| Operations | Chasing stalled learners stays manual | Count the hours it takes each week before deciding what the platform is for. |
A 30-day validation plan: Digital learning experience
Days 1 to 5: find where learners stop. Take one existing course and plot completion module by module. There will be a specific point where people stop, and it is usually a single piece of content or a task that assumes knowledge nobody has. Producing more material before finding that point adds cost without adding completions. Separate the people who never started from the ones who stopped partway, because those two groups have nothing in common and need different fixes.
Days 6 to 15: ask ten learners what happened there. Contact the people who stopped and ask what they were doing at the time and what they would have needed. Their answers are usually about length, timing or a missing example rather than about production quality, which is where most of the budget normally goes. Ask them when they tried to study, since a course designed for an evening and attempted in a lunch break fails for reasons no content change will solve.
Days 16 to 30: change that one module and compare cohorts. Rework the module, release it to a new group, and compare their completion with the previous cohort at the same point. A single fixed blockage usually moves the overall number more than a new course would, and it costs a fraction of one. Hold everything else constant for that comparison, as a redesign released alongside a new reminder email cannot tell you which one worked.
A practical worksheet: Digital learning experience
Recorded video is the cheapest part of digital learning and the part least connected to whether anyone learns. Fill these rows for one course that already exists and is not finishing well.
| Decision area | What to record |
|---|---|
| Learner | Who they are, when in the day they study, and what interrupts them on a typical attempt |
| Content | The smallest useful unit, how long it runs, and whether it can be resumed on a different device |
| Feedback | How a learner finds out they have misunderstood something, and how long that takes |
| Operations | What an instructor or administrator does each week, and how much of it is chasing people who have stalled |
| Measurement | Completion by cohort alongside the point where people stop, which is more useful than an average progress figure |
Find the drop-off point before commissioning more material. Learners usually stop at a specific module for a specific reason, and producing another ten hours of video on top of an unresolved blockage adds cost without adding completions.
Related guides
- A useful companion to this one is Workforce Management App: Mobile, Web, and Admin Scope.
- For the detail behind the next decision, see Workforce Visibility: Benefits, Risks, and Responsible Metrics.
If Why the Digital Learning Experience Is More Than Course Video describes a problem specific enough to need its own system, Discuss Your Industry Platform.
Ali Boran Gazel