Designing mentor notes that analysts actually reopen

Mentor notes fail when they read like marketing blurbs. We ask reviewers to attach one concrete timestamp, one suggested rewrite, and one explicit non-goal for the next attempt. That triplet keeps feedback scannable after midnight handoffs.

Cohorts told us they reopened notes more often when uncertainty was named. Phrases like “low confidence on host lineage” outperform vague “needs work” tags because the next analyst knows where to dig first.

We also cap note length. Long essays signal heroics; short memos signal respect for shift time. The cap forces mentors to pick the highest leverage edit instead of documenting everything they noticed.

Finally, we archive exemplar notes (anonymized) inside the cohort workspace so newer mentors calibrate tone quickly. The library is small on purpose—quality over volume—and rotates every quarter as scenarios refresh.