The 2026 ISTF barometer contains a figure that should give every digital learning manager pause: only 5 % of programmes without tutoring reach more than 80 % completion. With support in place, that rate climbs to 51 %.
Ten times better, for identical content. The conclusion is hard to sidestep: what makes people finish a course is not the quality of the modules, it is the fact that someone comes back to the learner.
The trouble is that human tutoring does not scale. A tutor can follow thirty learners, not three thousand.
- Where learners actually drop off
- Why email is no longer enough for this follow-up
- What Web Push brings to digital learning
- The scenarios that lift completion
- Cadence matters as much as the message
- Measuring beyond the completion rate
- The mistakes that cost you learners
- What to take away
- Frequently asked questions
Where learners actually drop off
Drop-off almost never happens during a session. It happens between two sessions.
A learner finishes their Monday evening module, satisfied, sincerely intending to pick it up again on Wednesday. On Wednesday a meeting overruns. On Thursday they vaguely think about it. By Sunday they have lost the thread of what they had learned and push it to next week. Three weeks later, they are not coming back.
This decline follows a curve known since Ebbinghaus: without reactivation, a significant share of what has just been learned fades within days. The learner who returns after two weeks does not pick up where they left off, they feel they are starting over. That feeling is precisely what discourages them.
The tutor’s role, in the programmes posting 51 % completion, is not to explain the content. It is to remind the learner it exists, at the right moment, before the thread snaps.
Why email is no longer enough for this follow-up
Email follow-up remains the default reflex, and it is becoming less and less effective. A training reminder lands in a work inbox already receiving dozens of more urgent messages. It is opened the next day, or not at all, and above all it is read at a moment when the learner is not available to study.
The problem is not the open rate, it is the gap between the moment of reading and the moment of action. A reminder that arrives in the middle of a meeting produces nothing.
What Web Push brings to digital learning
The web push notification appears directly on the learner’s screen, without them having to open anything. Three characteristics make it particularly well suited to online training.
It works on desktop, where people actually study
Unlike the app notification, designed for mobile, Web Push also displays on desktop. Yet the bulk of professional e-learning is consumed on a computer, often on the very machine where the learner works. The notification therefore arrives exactly where resuming is one click away.
It requires no app and no personal data
The learner accepts notifications in one click from their browser, with no app to install and no personal address to hand over. For an LMS platform or a training provider, that removes the sign-up friction that usually weighs down follow-up channels. The format is detailed on our Web Push page.
It triggers on real behaviour
Effective follow-up is not calendar-based, it is conditional: it goes out because the learner has done nothing for four days, not because it happens to be Tuesday. Wired into LMS events, the notification becomes the delivery arm of a marketing automation logic applied to pedagogy.
The scenarios that lift completion
- The resume reminder after three days of inactivity, carrying the exact title of the next module rather than a generic message
- First-module onboarding: a learner who has not started within 48 hours of enrolling probably never will
- The live session or virtual classroom reminder, one hour ahead, which follows the same mechanic as cutting no-shows at your events
- The deadline prompt: end of programme, certification cut-off, expiry of a funded access
- Progress recognition: module completed, badge earned, milestone passed, which sustains momentum rather than merely reacting to its loss
- The consolidation quiz at day seven, reactivating memory just as the forgetting curve starts to bite
Cadence matters as much as the message
Over-frequent follow-up produces the exact opposite of the intended effect. The learner unsubscribes in one click, and there is no way to re-engage them afterwards: unlike an email address, a lost push subscription is lost for good.
Two notifications a week is a reasonable ceiling on an active programme. The useful rule is simple: every send must carry information the learner could not have anticipated. A prompt about the next module, yes. A message repeating what they already know, no.
The logic is close to what applies when reducing churn on a subscription: what makes a user leave is almost never a single message too many, it is an accumulation of prompts without value.
Measuring beyond the completion rate
Completion rate is the sector’s most quoted metric, and its most contested. A learner who takes three modules out of ten because those were exactly the ones they needed is not a pedagogical failure, even though the dashboard counts them as one.
Three metrics give a more honest read on the effect of follow-ups: average time between two sessions, which should shorten; the resume rate after a notification, which measures the channel’s effect directly; and the share of learners still active at day 30, a better predictor of final completion than any start-of-course indicator.
To isolate the programme’s real effect, the most reliable method is to deliberately exclude one cohort from the follow-ups and compare its progress with that of the notified learners.
The mistakes that cost you learners
The first is the generic message. “Resume your course” triggers nothing. “You have 12 minutes left on the Conflict Management module” works, because the effort asked for is quantified and looks achievable.
The second is the wrong moment. A prompt sent at 9 am, as the learner opens their working day, will be swept aside. Real resume windows sit rather in late morning, early afternoon and early evening.
The third is following up only with those who have stalled. Regular learners drop off too, simply later. A progress acknowledgement sent to someone moving along well costs one message and secures a completion.
What to take away
The gap between 5 % and 51 % completion is not decided by content quality, it is decided by presence. The programmes that work are those where something comes back to the learner before they have disengaged.
Web Push does not replace a tutor, it industrialises the one thing a tutor does that a platform does not: going to the learner rather than waiting for them.
Frequently asked questions
How do you reduce course drop-off with push notifications?
By following up with the learner before they disengage, not after. A notification triggered after three or four days of inactivity, naming the exact next module and the time remaining, brings the learner back while the thread is still intact. The 2026 ISTF barometer shows that support raises the share of programmes exceeding 80 % completion from 5 % to 51 %.
Why is Web Push more effective than email for re-engaging a learner?
Because it appears directly on screen, without going through a saturated inbox. Above all, it works on desktop, where most professional e-learning is consumed: the prompt therefore arrives exactly where resuming is one click away, rather than in a space checked at a moment when the learner is unavailable.
Do you need a mobile app to send notifications to learners?
No. Web Push works from the browser, on desktop as well as mobile. The learner accepts notifications in one click, with no download and no personal address to share, which removes the main friction LMS platforms face on their follow-up channels.
How often can you follow up without driving learners away?
Two notifications a week is a reasonable ceiling on an active programme. The decisive criterion is not the number but the value: every send must carry information the learner could not have anticipated. A lost push subscription cannot be recovered, unlike an email address.
Which metrics should you track beyond the completion rate?
Average time between two sessions, which should shorten, the resume rate after a notification, which measures the channel’s effect directly, and the share of learners still active at thirty days, the best predictor of final completion. To isolate the real effect of follow-ups, compare a notified cohort against a control cohort excluded from them.
