TL;DR. Two hours, three passes. Pass one is retrieval: rebuild the year from records before you judge any of it, because your memory of twelve months is mostly a memory of the last six weeks. Pass two is about ten questions, grouped into what happened, what changed in me, and what I am still carrying. Pass three is deciding one thing, not twelve. Then put a date in the calendar to reopen it, or it becomes a document nobody reads.
Every December I sit down to think about the year and find I can only really remember November.
That is not a memory failure, it is how recall works. Recent material is loud and easy to reach, and the rest of the year is filed somewhere you cannot get to by sitting quietly and trying harder. Which is why most annual reviews turn out to be a review of the last quarter with a sentence about January glued on top.
The fix is unglamorous. You go and fetch the year from records first, before you allow yourself an opinion about it.
I build a habit tracker in the evenings, so I have more raw material about my own year than most people do, and I still get this wrong whenever I skip the first pass. Here is the version that works: two hours, three passes, doable in one evening or spread across three.
Why "what went well, what went badly" is a weak frame
Start with the frame, because most templates hand you the wrong one.
"What went well and what went badly" sounds neutral and is not. It is a verdict rather than a question. It sorts the year into two bins by outcome, and outcomes are contaminated with luck: a launch that landed and a launch that did not can be the same work in a different month. Sorting by result teaches you almost nothing you can repeat.
It also invites performance. Nobody writes an honest "went badly" list on the first pass. You write the socially acceptable version, the one you would say out loud at a dinner.
Better questions ask about behaviour and cost. What did I do repeatedly. What did I keep renegotiating. What did I pay for and not use. Those have answers you can act on, and they are harder to flatter yourself with.
Pass one: retrieval, about 45 minutes
This is the pass everyone skips, and skipping it is why the exercise so often feels thin.
The rule for this pass: collect, do not judge. No scoring, no adjectives, no conclusions. You are a clerk assembling a file, not a critic.
Work through the sources in roughly this order:
- Your calendar, month by month, start to finish. It is the single best record you own because you did not write it to impress anyone.
- Your camera roll, filtered by month. Photos are timestamped and they remember trips, people and ordinary weekends that your narrative dropped.
- Bank and card statements. Money is an honest record of attention. Twelve months of it will show you a subscription you forgot, a habit you deny having, and at least one month that was clearly harder than you remember.
- Sent messages and sent email. Not the inbox, the sent folder. What you initiated is the real signal.
- Anything you tracked. Habits, workouts, books, commits, work tickets, invoices, a journal. If you kept a time audit at any point this year, that week is worth more than a month of impressions.
- Your notes app. Old drafts and abandoned lists are a map of what you meant to do.
Write it as a plain timeline, one to three lines per month. Twelve rows. That is the whole deliverable of pass one.
Two things happen when you do this properly. Months you had written off turn out to contain something. And the month you were about to build the whole review around shrinks back to its actual size.
Pass two: about ten questions, roughly 45 minutes
Now you are allowed opinions, because you have a document to have them about.
Keep it near ten. A forty-question template does not produce more insight, it produces shorter answers.
Group A: what happened.
- What did I actually spend this year doing? Answer from the timeline, not from memory.
- What is on the timeline that I had already forgotten?
- What did I finish?
- What did I start and abandon, and what did the abandoning cost me?
- Who did I actually spend the year with? Not who matters in principle. Who is in the photos and the messages.
Group B: what changed in me.
- What am I noticeably better at than twelve months ago, including things nobody would applaud?
- What did I keep renegotiating with myself all year? Repeated renegotiation almost always means the commitment was badly sized, not that you are weak.
- What did I do this year that last December's version of me would not have done?
Group C: what I want to stop carrying.
- What is finished but still taking up room? Old projects, an identity that expired, a grudge, a subscription, a plan you have stopped believing in.
- What did I say yes to out of habit, guilt, or who I used to be, rather than because I wanted it?
Question 9 and question 10 are the ones people quietly skip. They are also the two most likely to change your next year, because they free capacity rather than adding obligations.
Pass three: decide one thing, about 30 minutes
Here is where reviews go to die. You reach the end feeling clear-headed and productive, and you write down twelve things.
Twelve is zero. Pick one.
The format that survives contact with February is a decision, not an aspiration: one thing you will do, when you will do it, and what it displaces. That last part is not optional. Your calendar was full this year too, so anything new arrives at the expense of something specific. Name the something.
If you genuinely cannot get below three, rank them and admit that two of them are wishes.
Doing it in the tool
If you would rather be walked through it than stare at a blank page, the year in review generator is the guided version of pass two and pass three. Seven steps, nothing to sign up for, everything saved in your own browser.
The steps are: rate seven areas of your year from 1 to 10 (health, work, relationships, money, growth, fun, home), build a highlight reel of up to ten moments, answer four prompts about wins, record up to seven lessons as a pair of what did not work and the takeaway, name up to eight people who mattered, list up to seven pieces of unfinished business, then set a theme word and three goals for next year.
At the end it produces an on-page summary, a downloadable PNG summary card, a plain-text version you can copy into whatever you keep, and a print view. It autosaves as you type. The year switcher offers this year and last year, so it works in late December and in January without arguing with you about which year you mean.
One honest note about using it. The tool opens on the ratings step, and a gut rating is exactly the thing that recency distorts. Do your retrieval pass first, on paper or in a note, then come back and rate. Rating a year you have just reconstructed produces different numbers than rating a year you are trying to recall, and the difference is the entire point of this article.
If you want the present-tense companion to this exercise rather than the backwards-looking one, a life audit asks what is true right now and forces a keep, change or stop verdict on each finding.
What an annual review is actually good for
I am not going to tell you that reviewing your year makes the next one go better. I have no evidence for that and neither does anybody selling you a template.
What it is good for is narrower and still worth two hours:
- It corrects the recency distortion. You end the year with a more accurate picture of it than you started the evening with. That is the reliable payoff.
- It makes the year retrievable. In four years, this file is the only thing standing between you and a decade that reads as a blur. Seeing your life laid out in weeks makes that argument better than I can.
- It surfaces patterns invisible at week scale. A weekly review cannot see that you have restarted the same project every March for three years.
- It produces one decision, which is more than most Decembers produce.
What it will not do is create motivation, replace a plan, or make a hard thing easier. It is a reflection exercise. Judge it as one.
Afterwards, or it was just a nice evening
Three small moves, all of which take less time than the review did.
Put the decision where you will trip over it. A habit, a recurring block, a deadline with a date. Not a line in a document.
Set a reopen date. Put a calendar event around June with a link to the file. Half the value of a review arrives six months later, when you find out which version of you was writing.
Keep them in one place, in the same shape, every year. Year one is mildly interesting. Year three, read side by side, is genuinely useful, and it is the only part of this that compounds.
The thirty-minute version
If this is December 29 and you are tired: do pass one only. Twelve rows, one to three lines each, straight from the calendar and the camera roll.
Then answer a single question underneath it: what do I want more of next year, and what does it come out of?
That is a real annual review. Everything above is the same exercise with more room.
Last updated: December 2026.
Written by Eusebiu, the solo founder building Loggd in public. I build it in the evenings around contract work and a small daughter, so my own timeline rows are short and my December conclusions are usually some version of "do fewer things, finish them". I share what the aggregate data keeps showing on Threads.
Make next year retrievable. Loggd keeps your habits, goals and check-ins in one place, with a contribution grid that treats a quiet week as a quiet week, so next December you review evidence instead of impressions. Start free.
