## Why It Matters

The worst part of review season is reconstructing six months of work from memory — recency bias wins and Q1's biggest ship gets forgotten. Goals with check-ins fix this at the cost of thirty seconds a week: each check-in is a dated breadcrumb of progress that's sitting there waiting when the review cycle opens.

In Kit, goals aren't a separate OKR ceremony — they're deliberately lightweight and they feed directly into [AI-drafted reviews](/docs/performance-ai-drafted-reviews). The drafter reads your goals, their status, and your latest check-in notes as first-class evidence. A well-tended goal list is the difference between an AI draft that cites your actual year and one that has nothing to work with.

## Setting Goals

Go to [Performance > Goals](/performance/goals). Each goal is a title ("Ship the billing revamp"), an optional description, and an optional due date — press **Add goal** and it's live. Goals belong to you; this is a personal working surface, not a top-down assignment tree.

A goal moves through four statuses:

| Status | Meaning |
|--------|---------|
| Active | In progress — the default |
| Achieved | Done — say so, it's review evidence |
| Missed | Didn't land; the check-in trail preserves the why |
| Dropped | Deprioritized deliberately — distinct from missed |

> [!TIP]
> Mark goals **Missed** or **Dropped** honestly instead of deleting them. A goal dropped because priorities shifted is a normal, defensible review narrative; a silently vanished goal is nothing. Removing a goal deletes its check-ins with it.

Goal descriptions and check-in notes are encrypted at rest, like all review content.

## Progress Check-ins

Press **Check in** on any goal to record two things: a progress percentage (0–100) and a short note — the prompt is "What moved since last time?". The goal's displayed progress is simply the latest check-in's value, and the latest note is what the AI drafter quotes.

Do and don't:

| Do | Don't |
|----|-------|
| Check in when something *moved* — a ship, a milestone, a decision | Check in on a fixed schedule with "still going" |
| Name concrete artifacts ("migration ran clean in prod") | Write vibes ("good progress this week") |
| Keep 3–7 active goals | Track every task — that's your issue tracker's job |
| Update progress down when scope grew | Let a stale 80% imply something it doesn't |

## How Goals Feed Review Evidence

When a reviewer (or you, on a self-review) presses **Draft with AI**, the drafter gathers your most recent goals with their status, progress, and latest check-in note, alongside the previous evaluation, your role expectations, and your own notes. Goals give the draft its skeleton: which themes mattered this period and what actually shipped against them.

Goals are evidence, never metrics. Nothing in Kit scores you on goal counts or completion rates, no cross-employee goal comparison surface exists, and goal data appears in a review only through the drafter's narrative — which a human reviewer then edits and owns.

> [!NOTE]
> Goals are personal and continuous — they don't belong to a cycle and don't reset when one finalizes. Prune achieved and dropped goals at cycle boundaries so the list entering the next period reflects the next period.

## Quick Checklist

- [ ] Add 3–7 goals in [Performance > Goals](/performance/goals) that describe your next quarter
- [ ] Check in whenever something material moves — one sentence is enough
- [ ] Resolve each goal to Achieved / Missed / Dropped before your cycle's due date
- [ ] On your self-review, pair goals with **Your own notes** for the work goals don't capture

## Next Steps

- [AI-Drafted Reviews](/docs/performance-ai-drafted-reviews) — see exactly how goal evidence appears in a draft
- [Writing Reviews](/docs/performance-writing-reviews) — the form your goal trail is preparing you for