Guide

Invisible impact does not count in calibration. That is unfair when the work is real. It is still true. Build a habit of leaving artifacts so promotion committees and your manager can see what happened.

Helping everyone with everything feels like ownership. Past a point it is the opposite: you become a single point of failure and never create the artifacts of higher scope. Teach, document, and hand off.

What to leave behind

Metrics before and after. Screenshot the dashboard. Record the baseline the week you start. After launch, write one paragraph: what changed, by how much, over what window, with what caveats.

Short write-ups. A launch email, a postmortem section you owned, a design doc decision log, a "how we cut latency" note in the team wiki. Future you will paste these into a brag doc.

Demos. Ten minutes in team meeting beats a thousand Slack messages. Show the before/after, the failure mode you prevented, or the tool others can use.

Name collaborators accurately. "I led X with Y and Z" is stronger than either erasing others or disappearing into "we." Committees distrust both glory-hogging and false modesty that hides your role.

Keep a running brag doc: dated bullets with links. Update it weekly for five minutes. Promo season is the wrong time to reconstruct a year from memory.

In interviews and self-reviews, describe work in scope language: "I owned X end to end" / "I aligned three teams on Y" / "I moved metric Z from A to B." That vocabulary matches how serious ladders are read.

Traps that look like hard work

Not all hard work grows scope.

TrapLooks likeWhy it stalls growthBetter move
Ticket heroAlways available for interruptsNever builds project muscleBatch interrupts; protect project time
Refactor foreverBeautifying code with no user or metric outcomeHard to defend in calibrationPair refactors with a shipping goal
Meeting spongeInvited everywhere, decides nothingCalendar full, impact thinDecline or attend with a decision agenda
Secret ownerFix production quietly, tell no oneTrust without evidenceShort write-up after each class of fix
Scope hoardingStart five projects, finish noneReputation for unreliabilityOne in flight; park the rest

Common mistakes

  • Reconstructing a year in October. Five minutes a week in the brag doc.
  • Erasing yourself into "we" or hogging credit. Name the role and the collaborators.
  • Fixing production forever with no write-up. Trust without evidence still fails calibration.
  • A full calendar of meetings with no decisions. Influence needs artifacts.

Check yourself

  1. You quietly fixed a recurring Sev-2 class. What do you do this week? A short write-up with the before/after and who else to thank. Secret ownership does not count.

  2. Why update a brag doc weekly instead of at promo time? Promo season is too late to reconstruct a year from memory. Dated links are the evidence.

Key takeaways

  • Impact that is not written down, measured, or demoed often does not exist in calibration.
  • Screenshot baselines, write one paragraph after launch, and demo in ten minutes.
  • Name collaborators accurately.
  • Avoid ticket-hero and invisible-savior traps; teach and document so the work scales without you.

Further learning

StaffEng (Will Larson)Staff Engineer: Leadership beyond the management trackStaff Engineer: Leadership beyond the management track

Back to the map on growing your scope.