How-to: Draft a Performance Review as a Manager
A performance review draft is not a surprise reveal. It is a written synthesis of what you already coached in 1:1s, project debriefs, and same-day feedback. Good drafts are specific, evidence-backed, and fair to the full review window. Weak drafts recycle adjectives, over-weight the last two weeks, or bury the real message until the live meeting.
Time box: 60–90 minutes per review draft once your evidence folder is ready. Block a quiet hour. Do not write ten reviews in one exhausted evening if you can stagger them across a week.
This sits in the New Manager hub and Director / VP hub. Pair with how HR works for your company’s rating scale, calibration rules, and system of record. For compensation language, stay inside policy and use finance for non-finance managers only to understand budget context, not to invent promises.
When to use / when not
Use when:
– Your company has opened the review window and you are the rating manager
– You have (or can gather) artifacts from the full period: goals, 1:1 notes, peer input, delivery outcomes
– You will hold a live conversation after the draft, not only submit a form
– Calibration or skip-level review may follow, so clarity matters
Do not use when:
– You have not observed the person enough and are tempted to invent a narrative (ask for an extension or co-rater guidance)
– The topic is an active employee-relations case that HR owns
– You want the written review to replace a PIP or formal warning (those are separate processes)
– You are writing under anger from a single recent incident
Education only: this is a drafting runbook, not legal advice. Follow your handbook and HR partner.
Prep
Build a private evidence folder before you open the form:
- Goals / OKRs for the period and status
- Outcomes: shipped work, metrics you already track, customer or stakeholder notes (no invented numbers)
- Feedback already given: 1:1 themes, praise, course-corrects
- Peer or cross-functional input if your process allows it (see asking peers in a separate how-to)
- Growth edges discussed earlier, not brand-new surprises
- Company rubric: rating definitions, competency language, character limits
Also skim how departments work together if their role is heavily cross-functional so you do not judge only inside your silo.
Steps
- Re-read the rubric (10 min). Write the rating definitions in your own words so you do not drift into “gut feel.”
- Timeline the period (15 min). List major projects and moments chronologically. Mark wins, misses, and context (scope changes, headcount, blockers).
- Draft strengths with evidence (20 min). For each strength: situation, behavior, impact. Prefer observable work over personality labels.
- Draft development areas with evidence (20 min). Same structure. Tie each to a forward ask for the next period.
- Write the summary narrative (15 min). Three to five short paragraphs: overall story, strengths, growth, next goals. Match the form’s tone (formal vs. conversational).
- Propose next-period goals (10 min). Few, measurable where possible, owned by the employee with your support.
- Self-edit for fairness (10 min). Check recency bias, similarity bias, and whether remote or quiet contributors got equal evidence weight.
- Submit per policy, then schedule the live talk. The draft is not done until the conversation is on the calendar.
Script (what to say / send)
Invite to the review conversation:
I’ve submitted your review draft in [system]. Let’s meet [day/time] to walk through strengths, growth areas, and goals for next period. Please bring anything you want me to consider; this is a dialogue, not a reading of a verdict.
Opening live:
“I’ll share how I see the period, then I want your view. We’ll end with goals. Nothing here should be brand new if we’ve used our 1:1s well; if something feels new, call it out.”
Strengths beat:
“On [project], when you [behavior], the impact was [outcome]. That’s the pattern I want us to protect.”
Growth beat:
“On [situation], the gap was [behavior vs. expectation]. Going forward I need [specific ask]. Here’s how I’ll support: [support].”
If they disagree on facts:
“Help me correct the record. What context am I missing, and what evidence should we add to the file or the next cycle?”
Close:
“I’ll update [system] with clarifications we agreed. Next 1:1 we’ll break goals into monthly checkpoints.”
Output artifact
Manager review draft packet (private until submitted):
- Evidence log (dated bullets)
- Strengths section (3 examples max that are vivid)
- Development section (1–3 focus areas, each with a next ask)
- Overall narrative aligned to the rating you will defend in calibration
- Proposed goals for next period
- Talking points for the live meeting (one page)
Keep a clean version in the HR system of record. Do not store sensitive commentary in random shared drives.
Watch-outs
- Recency bias. The last crisis is not the whole year. Force a full-period timeline.
- Surprise theater. If the growth area never appeared in 1:1s, you failed the process before you wrote the form.
- Vague praise. “Great attitude” without examples helps neither promotion cases nor the employee.
- Stack ranking in prose. Do not write comparative insults about peers in someone’s file.
- Compensation improvisation. Do not promise raises, levels, or bonuses the budget and HR process do not support.
- Copy-paste reviews. People can tell. Calibration committees can tell.
- Ignoring cross-team impact. Partner with how sales works or other function hubs when the role’s outcomes live outside your team.