A simple yearly review ritual worth repeating
Most annual reviews are goal-setting with extra steps. This one reads the year you actually had first, which changes what you conclude from it.
The chapters you are told you will never forget are the ones you retain least. A low-effort way to keep a record while you are too busy to keep one.
A new child. A move to another country. Caring for a parent. Starting the company. Leaving the marriage.
These are the periods people most want a record of, and they are reliably the periods when journalling collapses. The reason is simple: the same upheaval that makes the time worth recording is what removes the capacity to record it.
So the usual advice — write daily, write at length, be reflective — is close to useless here. What follows is the version that survives contact with an actual difficult year.
The expectation is that something this significant will stay vivid. It does not, and the way it fails is specific.
Intense periods compress. A year of early parenthood becomes four or five scenes and a general impression of exhaustion. The texture — what you actually worried about in month three, what was funny in week two, the thing that turned out not to matter — goes almost completely.
Memory also revises toward the ending. Once you know how it turned out, you cannot reconstruct not knowing. A move that worked becomes a story about a good decision. The six weeks when you were sure it was a mistake get quietly edited out, and those six weeks are the most useful part of the record.
This is why the oral-history people are so insistent about recording close to the event. Projects like StoryCorps exist because testimony collected at the time differs from testimony collected later, in ways the teller cannot detect.
A journal is the only record written before you know the ending.
The target during a hard chapter is not a good entry. It is any entry.
Three sentences is a complete entry. It is not a lesser version of a real one. In four years, three sentences from a Tuesday in month two is worth more than the thoughtful essay you did not write.
Write badly on purpose. No structure, no conclusion, no attempt at meaning. Meaning is a later job and attempting it now is one of the main reasons people stop.
Skip freely. A journal with gaps is a journal. A missed fortnight is not a failure and does not need an apology entry — which is, reliably, the entry that precedes quitting altogether.
Write at the same moment every day, not the same time. Attached to an existing fixed point — after the last feed, on the train, once the house is quiet. During upheaval, clock time is unreliable and sequence is not.
Most people, given three sentences, write how they feel. That is the least useful content, because feelings are the part memory keeps in outline and the part that is least legible later without context.
Better, in rough order:
Then feelings, if there is room.
The test is whether a stranger could follow it. Not because one will, but because in four years you will be close enough to a stranger that the same standard applies.
If the chapter has a recognisable beginning, spend twenty minutes once, at the start, on a single entry covering:
That entry will be worth more than the next fifty combined, because it is the only one written in genuine ignorance of the outcome. Everything after it is partly contaminated by knowing a bit more.
If the chapter has already started, write it now anyway. Late is fine. The point is to capture the not-knowing before it is gone.
During a chaotic period, an archive can be lost without anyone noticing.
If it has already ended up scattered — a notes app, some voice memos, a paper notebook — that is normal and recoverable. Pulling it into one archive is the job, and dated text is all it takes.
Two warnings, both learned the hard way.
Not too soon. Reading an intense period while still inside it, or just out of it, is usually unproductive. The distance that makes rereading valuable has not arrived. A year is a reasonable minimum.
Not chronologically. Fifty entries from a hard year read in order are repetitive and heavy, and most people stop a tenth of the way in. Read one thread instead — one person, one worry, one question — front to back. It is finishable and it shows change, which the chronological read flattens.
The finding that tends to land hardest is what stopped. A concern that occupied every entry for two months and then vanished without your noticing is direct evidence about how the current ones will age — and that is not available from anywhere else.
When the chapter closes, there is one entry worth writing deliberately: what you now know that you did not know at the start.
Write it before rereading anything. Then read the archive, and write a second one. The gap between those two entries is the most honest measure you will ever get of how much the period actually changed you, and how much your memory had already smoothed it.
That pair of entries is also the thing you will reread in a decade, far more often than the daily ones. The dailies are the evidence. The pair is the finding.
If you are in the middle of something now, this is roughly what the whole practice is for.
Most annual reviews are goal-setting with extra steps. This one reads the year you actually had first, which changes what you conclude from it.
Day One exports in three formats and only two of them are useful outside Day One. Here is which to pick, how to get it, and what survives the move.