Identify Churn Risk Signals in Active Clients

You are an account management lead at a digital marketing agency analyzing client health across an active retainer book.
Agency: [AGENCY_NAME] — [SERVICES]
Client health data (last 90 days):
[CLIENT_HEALTH_DATA]
Include per-client signals such as: reporting call attendance, response time on Slack/email, ad spend trajectory, project delivery slippage, invoice payment timing, leadership/POC changes, and NPS or sentiment notes.
Identify which active clients are showing churn risk signals right now. For each at-risk client, list the specific signals observed, estimate the revenue at risk (MRR x months remaining on contract), and recommend the most likely save play.

- Base every flag on a real signal present in the input — no generic SaaS churn playbook language
- Use agency-relevant signals only: missed reporting calls, dropping ad spend, slipping project delivery, late invoices, POC/leadership change, drop in engagement on shared docs, reduced scope requests
- Rank clients by urgency (red/yellow) AND revenue at risk
- Skip clients with no signals — do not pad the list
- Tie every save play to the specific signal (e.g. missed 2 reporting calls → exec-to-exec check-in, not generic "send a value email")

Markdown table: Client | Tier (Red/Yellow) | Signals Observed | MRR at Risk | Months Left | Suggested Save Play | Owner. Below the table, list the top 3 patterns you see across the book.

Agency: Northpeak Digital — SEO + paid social retainers ($8k–$25k/mo). Health data: BrightHome ($12k MRR, 4 mo left): skipped 2 of last 3 QBRs, ad spend down 40%, new CMO started Apr 1. OakRoot Legal ($8k MRR, 7 mo left): on-time payments, 100% call attendance, asked about adding content. Stellar Fitness ($15k MRR, 2 mo left): invoice 38 days late, scope reduced last month, POC stopped replying in Slack. Voss Apparel ($22k MRR, 9 mo left): all green.