bar-raiser-interviewing · git:20260728.fac4544 · 2026-07-28 · sha256 ec290a798fbe98d3

bar-raiser-interviewing git:20260728.fac4544A

Immutable. This exact content is served forever at /api/v1/blob/ec290a798fbe98d3.

---
name: bar-raiser-interviewing
description: Interview as an Amazon-style bar raiser who holds an independent veto and judges each candidate against the long-term hiring bar. Use when you sit on an interview loop as the neutral interviewer accountable for the bar, not for filling the team's open seat.
---

# Bar raiser interviewing

A bar raiser is an interviewer with no stake in filling the seat and one job:
protect the company's long-term hiring bar against the hiring manager's
short-term need. The role exists because a team under headcount pressure will
rationalize a weak yes, and the person best placed to catch that is someone the
vacancy does not hurt. Strip out the independent veto and every loop drifts
toward the median of whoever happened to be available.

## Method

1. **Confirm your independence before the loop starts.** You are not on the
   hiring team, you carry none of its headcount pressure, and you own the bar
   across future hires, not this quarter's gap. If the hiring manager can
   overrule you, you are a reviewer, not a bar raiser.
2. **Assign competencies across the loop so nothing is tested twice or not at
   all.** Map each interviewer to two or three Leadership Principles or role
   competencies, split behavioral coverage, and remove overlap. A loop where
   three people probe the same strength has three blind spots.
3. **Collect written feedback independently before the debrief.** Each
   interviewer submits Inclined or Not Inclined with evidence before hearing
   anyone else. Feedback written after the room talks is anchored feedback, and
   groupthink launders a weak candidate into a hire.
4. **Demand evidence, not impressions.** Push interviewers to STAR structure:
   Situation, Task, Action, Result, and separate the candidate's "I" from the
   team's "we". "Strong presence" is a vibe; "reduced p99 latency 40% by
   rewriting the cache path" is a signal.
5. **Apply the raises-the-bar test explicitly.** The question is not "would I
   work with them" but "is this person better than half of current employees at
   this level in this role". A hire who lowers the median is a no even when the
   team is desperate.
6. **Run the debrief to surface dissent, not bury it.** Ask the
   lowest-confidence interviewer first, make each vote defend its evidence,
   and block the hiring manager's urgency or the loudest voice from carrying
   the room. Consensus reached by fatigue is not consensus.
7. **Exercise the veto when the bar is not met, and write down why.** An
   independent no, recorded with its reasoning, is a legitimate and durable
   outcome. Reopening the search costs weeks; a mis-hire costs the team for
   years.

## Litmus tests

- Could you defend the decision to someone who joins the company in two years
  and reads only the written feedback?
- Did any yes rest on team fit or urgency rather than evidence of the
  competency?
- Would this candidate raise the median of the current team at this level?

## Boundaries

The bar raiser judges the hire against the long-term bar, not sourcing
diversity, comp negotiation, or the hiring manager's read on day-to-day team
fit. This is an Amazon-shaped ritual: adapt the veto and the leveling rubric to
your org's convention, and never let it stand in for reference checks or the
manager's own call on scope.