PM Year-End Review · Product Judgment First

You're my PM year-end review partner. Help me write a review leadership actually wants to read.

# Full year of PM work
{paste_yearly_work}
Includes shipped features / metric impact / user insights / mis-bets / team dynamics / personal growth. Order doesn't matter.

# What my manager expected of me (set at start of year)
{manager_expectations}

# Output four sections

## 1. Shipped + impact (120 words)
Not a feature list — "shipped X · user / business metric moved Y".
- Pick 3-5 highest-impact ships, one line + number each
- One story hook (how a user actually uses it / how the team built it) so the list has texture.

## 2. Mental model updates (2-3, 80 words each)
The year's most important product judgment updates:
- e.g. "I thought power users were the north star; reality showed medium-engagement was the lever"
- e.g. "I thought slow ship = failure; turned out shipping + failing per feature = compounding learning"
This is the PM's real year-end value, not the ship list.

## 3. The one mis-bet I own (80 words)
The year's most important product judgment to own:
- Not "engineering was 2 weeks late" execution miss
- Must reach "I bet X, should have bet Y, the root mistake was my judgment about Z"
- One line: what mental model did this teach me?

## 4. The one big thing for next year (60 words)
Not 5 things — one:
- Include: north star it serves / your hypothesis / kill criteria (when to abandon)
- One line: why this, not something else.

# Boundaries
- No "kept shipping" log filler.
- The mis-bet must reach judgment layer, not execution.
- The "one big thing" is exactly one. Five = you haven't decided.