Run a retention strategy diagnostic and prioritise interventions

You are a retention strategist. Retention is the metric that actually pays the bills. Run mine forensically.

Deliver as a numbered list: (1) THE RETENTION CURVE READ: month 1, 3, 6, 12 - flat curve, slow decay, or bleeding, (2) THE SHAPE DIAGNOSIS: is the curve shaped like 'bad activation' (huge M1 drop), 'no second-value moment' (steady drop after M1), 'no habit loop' (slow leak from M3), or 'segment-mismatch' (cohort-dependent), (3) THE COHORT SLICE: retention by acquisition channel, by tier, by use case, (4) THE QUALITATIVE LAYER: the top 3 reasons given in cancellation surveys, the top 3 quotes from churned customers, (5) THE 3 INTERVENTIONS prioritised by ease vs impact, (6) for each intervention: what we change, how we measure impact, when we revisit, (7) THE 'DO NOT FIX YET' list - retention drops that look fixable but actually point to a deeper positioning issue (don't fix retention with ops if positioning is wrong), (8) THE WIN-BACK STRATEGY: the cadence and offer for churned customers (and the rule against win-back spam), (9) THE BRAND-RETENTION INTERPLAY: the parts of brand that drive retention (community, content, founder presence) and how to compound them, (10) THE METRIC TO WATCH WEEKLY that tells us retention is improving before the curve confirms it. Plain English.

Product: {{product}}
Retention curve: {{curve}}
Top 3 churn reasons from survey: {{churn_reasons}}
The cohort that retains best: {{best_cohort}}