Most technology advisors treat churn like bad luck — until a major client quietly leaves and commissions evaporate. This episode teaches a practical, non-technical approach to building a Client Risk Index: a lightweight churn-prediction system that combines a handful of high-value signals (billing anomalies, support volume, usage decline, payment delays, contract age) with simple scoring and automation. Bob interviews a practitioner who built a repeatable 30-day sprint that flagged at-risk accounts, automated a human-in-the-loop win-back playbook, and turned passive clients into review conversations and upsell opportunities. You’ll get a clear list of signals you can extract from common tools, a step-by-step scoring model you can implement with Sheets and Zapier/Make plus an LLM prompt set for prioritization, and an automation blueprint that assigns outreach, schedules account reviews, and tracks outcomes. Actionable, feasible, and designed for solo advisors and small teams.