How Do You Manage the Forgetting Curve?

Published 11 August 2026

You can't stop forgetting. The forgetting curve isn't a design flaw to fix once — it's a baseline you design reinforcement around, permanently.

Key Findings

The forgetting curve was first described by psychologist Hermann Ebbinghaus in 1880s experiments where he tested his own memory of nonsense syllables at increasing intervals after learning them.
A 2015 study by Murre and Dros, published in PLOS ONE, successfully replicated Ebbinghaus's original forgetting curve using the same "savings" method, testing recall at intervals from 20 minutes to 31 days after learning.
That replication found the forgetting curve isn't perfectly smooth: it shows a jump upward in retention starting at the 24-hour mark, meaning memory measured a day later behaves somewhat differently than the smooth exponential decay often assumed.
Separately, a 2008 study by Cepeda, Vul, Rohrer, Wixted, and Pashler found that each time material is successfully reviewed, the subsequent rate of forgetting slows, which is the mechanism spaced review relies on to counter the curve rather than just delay it.
QuikAuthor doesn't currently have a spaced repetition scheduling feature. TalentCards (Epignosis) does. Manual spacing of short, revisitable modules can partially substitute for automated interval scheduling, but doesn't adapt per learner the way a dedicated feature does.

What Is the Forgetting Curve and Why Does It Matter for Training?

The forgetting curve is the pattern by which newly learned information is lost over time if it isn't reviewed. Hermann Ebbinghaus first documented it in the 1880s by testing his own recall of memorised nonsense syllables at increasing time intervals, finding that forgetting happens fastest immediately after learning and then slows down. A 2015 replication of his original experiment, using the same method on a different subject, confirmed the same basic pattern holds up under modern scrutiny.

It matters for training because a single training session, with no follow-up, is fighting a curve that starts working against retention immediately. The forgetting curve isn't an argument against one-off training sessions existing at all; it's an argument against treating a single session as if the knowledge it delivered will simply persist afterward without reinforcement.

How Soon After Training Does Forgetting Start?

Based on Ebbinghaus's original work and its modern replication, forgetting begins immediately and is steepest in the earliest period after learning, then slows down over subsequent days. The 2015 replication specifically found a change in the pattern around the 24-hour mark, where retention behaved somewhat differently than a smooth, continuously declining curve would predict. The practical implication is that the first review after training matters disproportionately: reinforcement that happens within the first day or two intervenes while forgetting is at its fastest.

What Reinforcement Schedule Counters the Forgetting Curve Most Effectively?

Schedule the first reinforcement touchpoint within one to two days of the original training, not at the end of the week or month
Use active recall for reinforcement (a question the learner has to answer) rather than passive re-exposure (re-watching or re-reading the same material)
Space subsequent reinforcement sessions further apart as retention improves, rather than repeating on a fixed weekly or monthly schedule regardless of performance
Focus reinforcement on the specific content that was answered incorrectly at the last check, not a full re-run of everything
Treat reinforcement as an ongoing structural part of the training, not a one-time "refresher" scheduled after the fact

Does the Forgetting Curve Apply to Skills as Well as Knowledge?

The original Ebbinghaus research, and its 2015 replication, both measured recall of memorised information (nonsense syllables), which is closer to fact-based knowledge than to physical or judgment-based skill. Skills that involve muscle memory or repeated practice (like a physical technique or a software workflow performed regularly) tend to degrade more slowly than isolated facts, because ongoing use itself acts as informal reinforcement. Facts and knowledge that aren't used regularly in someone's day-to-day work, like a policy detail or a rarely-used procedure, are the category most exposed to the forgetting curve's steepest early decline.

One-Off Training vs. Training With Spaced Reinforcement

Factor	One-off training session	Training with spaced reinforcement
Retention immediately after	High	High
Retention at 1 week	Declining, following the forgetting curve's initial steep drop	Higher, because the first reinforcement touchpoint intervenes during the steepest decline
Retention at 1 month+	Low, without any intervention	Depends on continued spacing, but each successful review slows subsequent forgetting
Cost	Lower upfront (single session)	Higher upfront (ongoing reinforcement design and delivery)
Best fit	Content rarely needed again, or immediately applied and reinforced by regular use	Content needed reliably over time but not used daily

How Do You Measure Knowledge Retention After eLearning?

Test recall a set interval after training (a week is a reasonable first checkpoint), not only immediately at course completion
Use recall or application questions rather than recognition questions (multiple choice with obvious wrong answers tests less than open recall)
Track individual items separately rather than only an overall pass/fail, so you know which specific content is being forgotten fastest
Repeat the check at a second interval (a month, then a quarter) to see whether the decline is slowing, which indicates reinforcement is working
Compare retention for reinforced versus non-reinforced content where possible, to isolate the effect of the reinforcement itself

FAQ

What does Ebbinghaus's forgetting curve actually show?

It shows that newly learned information is forgotten fastest in the period immediately after learning, with the rate of forgetting slowing over subsequent days and weeks. Hermann Ebbinghaus documented this in the 1880s testing his own recall of memorised material, and a 2015 replication using the same method confirmed the basic pattern.

How do you flatten the forgetting curve after a training session?

By introducing reinforcement, specifically active recall rather than passive review, at increasing intervals starting within the first one to two days after training. Each successful recall slows the subsequent rate of forgetting, which is the mechanism that makes spaced reinforcement effective rather than just delaying the same decline.

Why do refresher courses fail to fix the forgetting curve problem?

A single refresher, scheduled once, addresses one point on the curve but doesn't change the underlying pattern: forgetting resumes again afterward. The research on spacing suggests multiple reviews at increasing intervals, not one refresher, is what actually slows the ongoing rate of forgetting.

Can microlearning slow down the forgetting curve better than one-off training?

Microlearning's short, separable units make it structurally easier to deliver reinforcement over time, which is the mechanism that counters the forgetting curve. The benefit comes from the spacing this format enables, not from short content being inherently more memorable on its own.

Does the forgetting curve apply to physical or software skills, not just facts?

Less steeply. The original research measured recall of memorised facts. Skills reinforced by regular use, like a software workflow performed daily, tend to degrade more slowly because ongoing practice functions as informal reinforcement. Facts or procedures used infrequently are the most exposed to the curve's steep early decline.

Where This Gets Harder Than the Theory Suggests

Knowing the forgetting curve exists doesn't make reinforcement easy to actually schedule and deliver in a busy organisation. Manually tracking which learner needs which piece of content reviewed, and when, at scale, is exactly the kind of task that becomes impractical past a handful of people without dedicated software support. If a scheduled, adaptive spaced repetition feature is what you specifically need, TalentCards has one; QuikAuthor's short-module structure supports manual spacing but doesn't yet automate the interval scheduling itself.

Want to see how short, revisitable modules get structured for reinforcement? Start free.