The Forgetting Curve: Are We Remembering the Wrong Lesson?
The Forgetting Curve is one of those ideas that has become deeply embedded in the psyche of those of us working in Learning and Development. Most of us have seen the familiar graphic: a steep line showing how quickly people forget what they have learned, often accompanied by a second line demonstrating how repetition, reinforcement or spaced practice can slow that decline. It is a persuasive image, and it has become a convenient way to explain why learners need reminders, refreshers and repeated exposure.
The story behind it, however, is more interesting than the graphic suggests.
The idea originates with Hermann Ebbinghaus, a German psychologist who published his research into memory in 1885. (See a 1913 translation online here.) Ebbinghaus was trying to study memory experimentally, and he wanted to remove as many confounding factors as possible. Meaningful material was messy: people brought prior knowledge, associations, emotional responses and personal experiences to it. So he deliberately created material that had as little meaning as possible.
He constructed thousands of consonant-vowel-consonant syllables and arranged them into random sequences. Think along the lines of DAX, BOK, LEP, ZIF, RUP, NEM, WUG. He would learn sequences like these until he could reproduce them correctly, then return to them after different periods of time to see how much effort was required to relearn them.
The lack of meaning was not a flaw in Ebbinghaus’s method; it was the point. He wanted material that would not be helped along by familiarity or association. As an experiment in memory, that makes sense. But when the results are carried across into workplace learning, the context changes completely.
That is where things become interesting.
Nick Shackleton-Jones, in How People Learn, (3rd edition July 2026) makes the point that by removing meaning and personal significance from the material, Ebbinghaus may also have created exactly the kind of information our minds are least interested in keeping. The nonsense syllables had no obvious relevance, no emotional significance and no practical use. They were simply items to be remembered for the sake of the experiment.
Viewed from that perspective, the results become rather less surprising. Perhaps Ebbinghaus did not simply show that people forget learning quickly. Perhaps he also showed how quickly we forget information that means very little to us.
That distinction is vital for workplace learning.
The people taking part in our programmes are not usually being asked to memorise arbitrary syllables. They are trying to do a job. They have customers to help, systems to use, risks to manage, conversations to handle, decisions to make and problems to solve. They already bring experience, context and priorities with them. Good learning design connects new information to that reality.
And it is that meaning which gives the learning somewhere to stick.
If you think about something you remember particularly well from your own working life, it is unlikely to be because someone showed you the same slide three times. You may remember it because something went wrong, because a customer asked a question you could not answer, because you suddenly understood why a process mattered, or because someone showed you a better way to solve a problem you genuinely had. The memory is attached to significance.
That does not mean every piece of workplace learning needs to be dramatic, entertaining or wrapped in a story. Meaning is not the same as novelty. It is simply about helping people see why something matters, when they might need it and how it connects to what they already know or do.
That creates a very practical design challenge for L&D. Rather than asking only, “How will we help people remember this?”, it may be more useful to ask, “Why would this be worth remembering in the first place?”
Sometimes learning becomes meaningless because we separate the information from the situation in which someone actually needs it. We explain a policy before showing the learner the kind of judgement the policy is there to support. We teach a software process without giving people a realistic task to complete. We present a list of expected behaviours without exploring what those behaviours look like in a difficult conversation. We communicate information because the organisation needs it covered, rather than beginning with what the learner needs to do.
The content may be accurate. It may even be important. But if the learner cannot see its relevance, it can start to feel a little like DAX, BOK, LEP, ZIF, RUP: information with no obvious place to go.
Here are some examples of how to embed something ‘abstract’ into a meaningful context for the learner.
Example 1:
✖ What we DIDN’T do:
We didn’t just say “You need to introduce yourself and your role to clients when you first meet them.”
✔ What we DID do:
We provided an opportunity to imagine they were talking to someone specific, and figure out how they would formulate that introduction in a real conversation.
Example 2:
✖ What we DIDN’T do: We didn’t present a history of the development of AI, to illustrate how it became pervasive.
✔ What we DID do: We provided examples that were relevant to learner’s experiences, to show the kinds of ways AI is already involved in their everyday lives.
Example 3:
✖ What we DIDN’T do: We didn’t tell students that it was important to know how to write an essay.
✔ What we DID do: We got two current students to talk about their experiences in the writing process, and how helpful they would have found these new resources.
As you can see, the natural ‘therefore’ of Ebbinghaus’ findings (namely spaced repetition) is actually not the whole answer. Retrieval practice, reinforcement and opportunities to revisit learning can all be valuable, but they are most effective when there is something meaningful to retrieve in the first place. If someone has learned something useful but has not needed it for six months, a well-timed opportunity to practise it again may be exactly what they need. If the learning never felt relevant in the first place, repeating it may simply give them another opportunity to tune out.
So when people are not retaining workplace learning, it may be worth looking beyond the familiar instinct to add another reminder, refresher or follow-up module. We can look instead at the learning itself. Did people recognise themselves in it? Could they connect it to the work they actually do? Were they making decisions, solving problems or practising something they would genuinely need? Did they understand why it mattered?
The Forgetting Curve has given L&D a useful reminder that memory is fallible. But perhaps Ebbinghaus’s work offers another lesson that is just as valuable: if we want people to remember something, we need to give them something worth remembering.
