Spacing effect
Study spread over time produces more durable memory than the same study massed together.
Holding total study time constant, splitting it across separate sessions produces better long-term retention than one continuous session. The optimal gap scales with how long you need the memory to last: longer retention intervals call for longer gaps. Massed study feels more effective while you are doing it, which is why people choose it.
Massed and spaced study, same total minutes
How it shows up in software
Learning products, language apps, onboarding sequences and compliance training all schedule against this. Outside education products it shapes how a team teaches a feature: one long tour on day one against a short reminder surfaced at three moments of use. The second pattern matches the evidence and the first matches the calendar.
Using it well
- Split product education across sessions. Teach the one thing needed now and surface the next thing on the next visit.
- Set review gaps against how long the knowledge must last. A skill needed in six months needs wider gaps than a skill needed on Friday.
- Attach reminders to the moment of use. A tip shown in the context that needs it is both a spaced repetition and a just-in-time hint.
- Warn learners that massed study feels better than it works, so they do not read spacing as wasted effort.
Where it turns manipulative
- Using spacing as cover for drip campaigns that exist to raise notification volume rather than to teach anything.
- Scheduling reviews the product needs (streaks, daily active use) rather than reviews the learner needs, then calling the cadence pedagogy.
- Applying spaced reminders to a one-off task, where a person who has already done the thing gets asked again.
Where you have seen it
Anki
Review intervals expand or contract per card based on the grade the user gives their own recall.
Duolingo
Previously learned words resurface in later lessons rather than appearing only in the unit that introduced them.
Figma
Shortcut hints appear next to the tool being used rather than in a single tour at first launch.
What the research says
- Ebbinghaus, 1885Well evidenced
Relearning distributed across days required fewer total repetitions than the same relearning concentrated in one day.
Single participant, nonsense syllables.
- Cepeda, Pashler, Vul, Wixted and Rohrer, 2006Well evidenced
A meta-analysis of 254 studies found distributed practice outperformed massed practice, and that the best gap grew with the length of the retention interval.
Grades are a judgement about the evidence, not about how useful the idea is. Plenty of contested effects are still worth knowing, as long as you do not cite them as settled.