Calibration meetings have a dirty secret: they are not about calibration.
They are about advocacy. Whoever makes the most compelling argument — with the most energy, the most political capital, or the most leverage — wins. The engineer with the most impact doesn't always get the right outcome. The manager who argues best does.
This is fixable. But the fix doesn't happen in the meeting. It happens in the six weeks before it.
The problem with how most managers prep
The standard approach: two days before calibration, managers pull up whatever they can find (Jira, PRs, Slack) and try to reconstruct a case for their engineers. They write a few bullet points. They show up hoping to out-argue the other managers.
This has two failure modes:
Recency bias. The last four weeks dominate the narrative, even if the most important work happened in Q2.
Vagueness. "She led the migration project" is easy to counter. "She led the auth migration, reducing P95 latency by 40% and eliminating our #1 on-call category" is not.
What strong calibration evidence looks like
Every strong calibration case answers three questions:
1. What did this engineer do that was above their current level? (Evidence of next-level contribution) 2. How do we know it mattered? (Quantified outcomes) 3. Who saw it? (Cross-functional reach)
If your calibration brief can't answer all three, you're going in underprepared.
The six-week prep protocol
Weeks 6–4: Activate your ICs
Ask your engineers to share their impact logs with you. If they've been logging in Reme, they can generate a Manager Calibration Brief in one click. If they haven't been logging, this is the forcing function.
Don't write their cases for them. Ask them to write the first draft. You'll add context and defend it — they should know the specifics better than you do.
Weeks 3–2: Build your calibration briefs
For each engineer you're sponsoring, produce a one-pager:
- TL;DR: a two-sentence summary of their highest-impact contribution
- Quantified evidence table: 3–5 outcomes with numbers
- Cross-functional reach: who they worked with, outside their immediate team
- Level argument: one paragraph connecting their work to the next level's expectations
The brief should be short enough to read in 90 seconds and specific enough that no reasonable person could dismiss it.
Week 1: War game the room
Go through each case and ask: *"What's the best argument against this?"* Then answer it.
Most calibration challenges fall into one of four buckets: - "The project was a team effort" → isolate the specific contribution - "The numbers could have happened anyway" → explain the counterfactual - "This is just their job" → compare to the level bar, not the average - "They had a rough patch in Q1" → acknowledge it, explain the growth arc
Preparation beats spontaneity in every calibration room I've seen.
The manager who doesn't have to argue
The best calibration outcome is one where the decision is obvious before anyone speaks.
When your brief is structured, quantified, and cross-referenced, objections don't gain traction. The evidence does the work. You're not advocating — you're presenting.
That's the position you want to be in. And it's built in the weeks before the meeting, not in the meeting itself.
If you want your engineers to help you get there, Reme gives them the habit and the format to make calibration prep automatic. When your team logs their work the right way throughout the year, your brief is already 80% written by the time you need it.