You just finished something. A project wrapped, a system launched, a mess got untangled, a difficult quarter ended. Before it fades, ask yourself three questions. They take about five minutes, and they are the difference between evidence and a vague warm memory.
There is decent evidence that the five minutes are worth more than they feel like. In a study run by researchers at Harvard Business School, HEC Paris, and UNC, people who spent a short period reflecting on what they had just learned went on to perform 18% better than those who spent the same time on additional practice. Writing down what happened is not overhead on top of the work. It is part of how the work turns into something you keep.
1. Where, and at what scale?
Not "did it go well." Where did this happen, and how big was it? How many locations, users, teams, sites, countries? How many people were affected by the thing you changed?
Scale is the first detail memory throws away, and it is the detail that separates "managed implementation" from "managed implementation across 19 locations supporting 84 users." Both sentences are true. Only one of them tells a stranger anything.
2. What was actually on the line?
What would have happened if this had gone wrong? What budget was riding on it, what deadline, what commitment to a customer, what risk to the business? A $600,000 implementation and a side experiment are different accomplishments even when the activities look identical.
Stakes are how a reader judges weight. Work with real consequences, stated plainly, needs no adjectives. That is the quiet reason strong résumés read as calm: the facts are doing the persuading.
3. What changed, and how do you know? Not how it felt. How do you know?
This is the question almost everyone stops short of. Things "went well." The team "was happy." The rollout "was a success." Felt how? Measured by what?
Ninety-two percent adoption in the first quarter is an answer. So is "cycle time fell from 23 days to 14," and so is "passed the audit with zero critical findings." Even "we honestly never measured it, but the complaint volume visibly dropped" is worth writing down, because it is true, and because it tells future-you where the measurement gap was.
If you can't answer this one yet, that is worth knowing too. Sometimes the measurement arrives a month after the project ends. Leave a note to come back for it. Numbers you capture at the source are numbers you never have to reconstruct under deadline pressure, or worse, guess at.
Five minutes, while it's still true
Answer those three questions in writing, anywhere durable, the same week the work ends. A notebook works. A document works.
The point is not the medium. The point is the timing. Memory decay is not a character flaw; it is the documented default. When researchers replicated Ebbinghaus's classic forgetting-curve experiment in 2015, they found the same steep early drop he described in 1885: most of what is lost goes in the first day or so, before you have any sense that it is going.
Every one of these details is easiest to capture at the moment the work happens and hardest to reconstruct years later, and the moment you'll want them most, mid-job-search, mid-review, mid-interview, is exactly when they're hardest to recover.
This is the habit Standing Proof exists to hold: the same three questions, asked while the answers still exist. Try it on something you've done, or read the method first.