early-churn-downgrade-signal-detection

Early churn signals are detected using a three-tier model: Tier 1 at deal close, Tier 2 across 30–90 days of onboarding behavior, and Tier 3 across 90–120 days pre-renewal. This approach flags risk upstream to give CS teams more runway for intervention.

To set up Tier 1 churn signals, ask the skill to run a Tier 1 assessment for the account and specify the desired tier1_mode as either rule or cohort. Cohort mode requires sufficient historical data for comparative analysis.

The best way to route escalation flags is to generate a Markdown risk-flag assessment with tier classification, confidence levels, and intervention recommendations. Every Tier 2 and Tier 3 flag includes an escalation path and named owner for handoff.

Yes, you can run a portfolio scan for downgrade risk. The skill emits Markdown output in either single-account or portfolio-scan formats and applies optional cohort-mode comparisons when sufficient historical data exists across the portfolio.

Cohort-mode comparisons in churn detection require sufficient historical account data to exist for analysis. The skill validates this prerequisite during pre-flight configuration before applying cohort-mode to Tier 1 or subsequent behavioral tiers.

Rule-mode for early churn signal detection lacks the comparative behavioral context provided by cohort-mode. When sufficient historical data exists, cohort-mode offers deeper analytical framing for Tier 1 classification than rule-mode.