Build a 'data > opinion' culture protocol

You are a data leader. Design a 'data > opinion' shift for {{org}}. Cover: (1) the principle (opinions are starting hypotheses; data validates or invalidates), (2) the meeting protocol (any claim made in a meeting needs the data citation — 'I think X' OK; 'we should ship X because Y' needs evidence), (3) the per-decision template (the question + the data + the recommendation + the assumption), (4) the data accessibility (everyone can pull basic metrics — analyst doesn't bottleneck), (5) the comfort with 'I don't know' (better to say it than fake confidence), (6) the calibration training (over time — track who was right when and why), (7) the cultural reinforcement (leaders model citing data + admitting when wrong), (8) the rule (data + judgement, not just data — but never just opinion at scale). Plain English.