In the 1880s, Hermann Ebbinghaus memorised thousands of nonsense syllables, tested himself at intervals, and drew the result: a curve that drops off a cliff. Memory fades fast at first — much of new material is gone within a day — then the fade slows, leaving a small stable residue. A 2015 replication ran his experiment again, 130 years later, and got essentially the same curve.
The exact numbers vary with material and person (meaningful material fades slower than nonsense syllables), but the shape is the finding: steep, then flat. Whatever you learned this morning is disappearing fastest right now.
Two honest conclusions
1. Forgetting is a feature. Your brain is aggressively discarding what it predicts you won't need — it's compression, not failure. The problem is that the prediction is often wrong: the client's exact words, the flash of an idea, the reason behind a decision.
2. Willpower is not a strategy. "I'll remember that" is a bet against a century of replicated data. You need a mechanism.
The two mechanisms that beat the curve
Review it — at spaced intervals. Each retrieval flattens the curve; spacing the retrievals flattens it durably. That's spaced repetition, and it's the right tool for things that must live in your head — languages, exam material, fundamentals.
Write it down — within seconds. For everything that just needs to be available (decisions, ideas, details), notes are the cheaper mechanism: an external memory with a flat curve. But the curve dictates the design — capture must happen in seconds, before the steep part does its work. This is the entire reason we harp on one-tap capture in meeting notes and everywhere else: the note you take now beats the perfect note in ten minutes, because in ten minutes half the texture is gone.
The broader science of what helps memory — retrieval practice, attention, offloading — lives in memory & focus.