Which hiring practice was quietly born in 1970s symphonies and still boosts equity today? We piloted blinded reviews — masking names and schools in Greenhouse — for initial screens last spring and saw pass-through for Black and Latine candidates rise 11% with no drop in quality signals. Who else has tried this and what tweak made it stick?
We made it stick by enforcing a “no peeking until the score is in” rule in Greenhouse — reviewers can’t open resumes or LinkedIn until they submit the structured scorecard; we allow an unblind override for rare roles via a quick Slack approval, which kept speed reasonable. Did you also delay unblinding until the HM submits?
Requiring a one-line evidence note for each rubric score before unblinding made it stick for us — kept reviewers honest and we saw the same “no drop in quality signals.” We kept speed by capping first-pass reviews at 7 minutes; did you mask employers too or just schools?
Mask referrals too — we saw bias creep back whenever a rec came in. In Greenhouse we added a single “one-minute work sample” question to the app and forced review by that score before any resume; pass-through bumped about 9% and hiring bar stayed flat. Works best for writing/analyst roles, so we swap in a 5–7 min scenario for others.
Quick example: in Greenhouse we hid locations and employment dates alongside names/schools for the first pass, and it mirrored your 11% bump for Black and Latine candidates. Only caveat: senior roles needed a manual second-look on borderline scores; , bias drift shows up as “gravitas” talk. @OP did you try hiding location or dates, or was that too heavy for the spring pilot?