In 1885, Hermann Ebbinghaus sat alone in a room and memorised nonsense syllables until he could recite them perfectly. Then he waited. And waited. And tested himself again. What he discovered — painstakingly, before spreadsheets and LMS dashboards existed — changed everything we should know about learning. And yet, 140 years later, most corporate eLearning ignores it entirely.

The forgetting curve is not a theory. It is one of the most replicated findings in all of psychology. Without active reinforcement, people forget approximately 50% of new information within an hour, 70% within 24 hours, and nearly 90% within a week. The curve is steep, ruthless, and completely predictable.

Here is the uncomfortable question: what does your training programme do on day two?

The One-and-Done Trap

Most corporate learning is built on a model that cognitive science dismantled a century ago: exposure equals learning. Attend the course. Watch the video. Pass the quiz. Move on. We measure completion because completion is easy to count. We do not measure what learners actually remember three weeks later because that number would make L&D teams uncomfortable.

I have audited dozens of compliance programmes with completion rates above 90%. When I request time-on-task data, I consistently find that learners are spending 2–4 minutes on modules designed to take 12–15. They are clicking through. The forgetting curve never even gets a chance to operate — because meaningful encoding never happened in the first place.

But here is the cruel irony: the courses where learners actually engage and genuinely learn something are the ones hit hardest by the forgetting curve. Because the curve requires something to forget. If learning occurred, forgetting will follow. And if you have built nothing to counteract it, you have invested in a temporary illusion of capability.

What the Science Actually Says

The research on what defeats the forgetting curve is not ambiguous. There are two interventions with overwhelming evidence: spaced practice and retrieval practice. They are not complicated. They are just underused.

Spaced practice means distributing learning over time instead of front-loading it. Instead of one 60-minute module, three 20-minute sessions spread across two weeks. Instead of a two-day onboarding bootcamp, a 90-day journey of bite-sized encounters. The spacing effect — the phenomenon where distributed practice dramatically outperforms massed practice — has been demonstrated across subjects, ages, languages, and cultures since Ebbinghaus himself identified it.

Retrieval practice — also called the testing effect — means making learners recall information rather than re-reading it. This feels counterintuitive. Re-reading feels like studying. But re-reading is passive. Retrieval is active. When you force your brain to reconstruct a memory, you strengthen the neural pathway. When you re-read, you are just recognising words on a page.

A landmark study by Roediger and Karpicke (2006) showed that students who took three practice tests retained 80% of material after a week. Students who re-read the material three times retained only 40%. Same time invested. Double the retention. The difference was retrieval.

How Your LMS Is Actively Working Against You

Most LMS platforms are designed for delivery, not retention. They track when a learner opened a module, how long they kept the tab active, and whether they clicked the final slide. They do not track whether the learner can actually perform the behaviour the training was meant to produce.

Worse, many LMS platforms actively discourage the interventions that defeat the forgetting curve. Spaced practice requires re-engaging learners multiple times — which is harder to schedule, harder to mandate, and harder to report on than a single completion. Retrieval practice requires assessment design that goes beyond multiple choice recognition — which takes more time to build and is often seen as punitive rather than pedagogical.

The result is a system optimised for the metric that is easy to capture (completion) rather than the outcome that actually matters (retention and transfer).

What You Can Do About It

You do not need to rebuild your entire L&D infrastructure to fight the forgetting curve. Start with three things.

First, add a follow-up moment. Whatever you train today, schedule a five-minute retrieval activity for one week later. A short quiz. Three reflection questions sent via email. A scenario prompt pushed through your LMS. This single intervention can more than double retention at almost no additional development cost.

Second, redesign your assessments as learning tools. Stop treating quizzes as gates to completion. Use them as retrieval opportunities. Space them throughout the experience. Make them slightly challenging — the research on desirable difficulty shows that making retrieval effortful (without making it impossible) produces stronger long-term retention than easy tests.

Third, activate the manager. The most powerful reinforcement mechanism available in any organisation is not a push notification. It is a manager who asks their direct report: "You completed that training last week — what are you doing differently?" A single conversation within 48 hours of training can dramatically extend the learning's effective lifespan. Build manager guides. Make it easy for them to have that conversation.

The Hard Truth

The forgetting curve is not your enemy. It is a feature of how human memory works — a system designed to discard information that is not repeatedly confirmed as important. The curve is doing exactly what it should do. It is pruning the irrelevant to make space for the essential.

Your job, as a learning professional, is to send the signal that this information is worth keeping. You send that signal through repetition, through retrieval, through application in the real world. A one-time course sends no such signal. It says: here is some information. And then it says nothing else.

Ebbinghaus figured this out in 1885. We have had 140 years to act on it. The question is not whether the forgetting curve is real. The question is what you are going to do about it on day two.

"The curve is doing exactly what it should. Your job is to send the signal that this information is worth keeping."