The annual review
Rees Calder · 20 April 2026 · 5 min read
Most advice about self-improvement is about inputs. Set goals, read books, wake up earlier, take cold showers. The input side is crowded.
The output side is almost empty. Very few people do a structured review of the year just past. Almost no one does it well. This is strange, because the research on reflection is unambiguous: people who reflect get better at whatever they're reflecting on. People who don't, don't.
This is the Net Positive annual review. One hour, once a year, using a structure that's been borrowed from management, therapy, and the implementation-intention literature. You will get more out of this one hour than you will out of most books you read this year.
Why reflection works
Three findings from the research.
Self-reflection outperforms additional experience on skill acquisition. Di Stefano, Gino, Pisano and Staats (HBS Working Paper, 2014, published in Mgmt Sci 2023) ran a controlled study where participants either did more of the task or spent 15 minutes reflecting on what they'd learned. The reflection group outperformed by ~23% on subsequent tests. This is a robust result, replicated across education, medicine, and sport. Reflection is a multiplier on experience, not a substitute for it.
Calibration improves with feedback. Dunning-Kruger-style miscalibration (we overestimate how well we know things, especially in domains we're inexperienced in) shrinks when you force yourself to produce predictions and then check them against reality. The year-end review is one of the cleanest natural moments to do this.
Implementation intentions beat goal-setting. Peter Gollwitzer's 1999 paper on implementation intentions, and the 200+ replications since, find that "when X happens, I will do Y" statements outperform generic goals by 2-3x on completion rates. The annual review is where you form these for next year, based on what actually happened this year.
The pattern: reflection + calibration + implementation intentions is worth more than any additional input you could cram in.
The four-part structure
Spend 15 minutes on each. 60 minutes total. One sitting. Pen and paper if you can, though digital is fine.
1. What actually happened (15 min)
Open your calendar from January 1 to today. Scroll through it. Write down the 10 most important things that happened this year. Mix the good and the bad. Include events, conversations, decisions, things you made, things you ended.
The point is to counter reconstructive memory. What you remember about a year is heavily biased toward the last 60 days and toward anything emotionally vivid. The calendar pulls you back to the actual texture. You will almost always find you've underrated several quieter wins and overrated the most recent drama.
2. What worked, what didn't (15 min)
Two columns. On the left, three things that worked better than expected (in work, relationships, health, anything). On the right, three things that didn't. Don't philosophise about them yet. Just note them.
Most people's lists come out heavier on the didn't-work side. That's the negativity bias, not reality. Force yourself to write three things that worked. If you can't, you've had a bad year, but you've also been under-attributing.
3. The predictions check (15 min)
If you did last year's annual review (you didn't, but you will next year), read it now. Mark which of your predictions came true and which didn't. Most will feel surprising.
For a first-time reviewer: write out three predictions you would have made for this year a year ago, and mark them against what actually happened. You will be humbled, and that humility is the point. Most of us are badly calibrated about our own lives.
This is the step most people skip. It is also the step that most improves future-you. Calibration is skill. Skill needs practice.
4. Implementation intentions for next year (15 min)
Now, and only now, write three specific implementation intentions for the coming year. Format: "When X happens (trigger), I will do Y (action)."
Bad: "I want to exercise more." Good: "When Monday, Wednesday, Friday at 6:45am arrives, I will change into running clothes and go outside."
Bad: "I want to be more present with my family." Good: "When I walk through my front door after work, I will leave my phone on the hall table for the first hour."
Three. Not ten. The research shows that people who pick three implementation intentions hit ~60% of them. People who pick ten hit under 10%. Constraint is the mechanism.
Where most annual reviews go wrong
Three failure modes to avoid.
The inspirational journal entry. People write what they wish they'd done, or what they think sounds good, instead of what actually happened. The honest version is more useful than the aspirational one. If you had a bad year, say that. Your future self will read this, not your LinkedIn network.
The vague goal trap. "Be happier. Be healthier. Be more focused." These are wishes, not plans. The research on implementation intentions is clear: specificity and triggers are where the lift comes from. If you can't write it as a trigger-action statement, it won't happen.
Skipping the prediction check. The single most educational part of an annual review is marking last year's predictions. If you don't build the habit of predicting and checking, your calibration stays flat forever.
What the research on end-of-history illusion tells you
One last thing. Quoidbach, Gilbert and Wilson (Science, 2013) documented the "end of history illusion": at any given age, people believe they've changed a lot in the past ten years and will change very little in the next ten. They are always wrong about the second part.
You will change more in the next twelve months than you expect. Your preferences, beliefs, interests, and plans will all shift. The annual review is a way to capture the before-state so that next year's you has a fair read of what actually changed. Without it, you'll just tell yourself a story.
One hour, once a year
That's the whole protocol. One hour. Once a year. Four parts. Three implementation intentions. Read it back in 12 months.
The evidence supporting this specific kind of structured reflection is stronger than the evidence supporting most self-improvement advice. The cost is the lowest. The compounding, across decades, is meaningful.
Set the hour. Actually do the review.
Sources used: Di Stefano, Gino, Pisano & Staats, Making Experience Count: The Role of Reflection in Individual Learning, Management Science (2023, WP 2014), Gollwitzer, Implementation Intentions, American Psychologist (1999), Quoidbach, Gilbert & Wilson, The End of History Illusion, Science (2013), Dunning, The Dunning-Kruger Effect, Advances in Experimental Social Psychology (2011), Locke & Latham, New Directions in Goal-Setting Theory, Current Directions in Psychological Science (2006). Full links in the planning doc.