> ## Documentation Index
> Fetch the complete documentation index at: https://helpdocs.getthread.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Churn Save Deep Dive

> When a client says they want to leave — an unflinching full-history analysis of what went wrong, what we did well, and the honest talking points for a win-back conversation.

<Info>
  **Category:** Client Lifecycle · [View source ↗](https://github.com/bryan-getthread/skills/blob/main/skills/client-lifecycle/churn-save-deep-dive/SKILL.md)
</Info>

**Connectors:** none — works with Thread out of the box

**Role:** [CSM / Account Manager](/start-here/roles/csm-account-manager)

**Outcome:** Retention & Growth (CSAT/Expansion)

**When to use:** "\<client> gave notice / says they're shopping — analyze what went wrong"; "build me a save plan for \<client>"; or "full history review on \<client> before the retention call."

**Run it:** on one client's full relationship history — a manual internal analysis, not a Flow.

## Prompt

```
The save conversation only works if the MSP walks in knowing the truth. Read the whole
relationship — the misses, the patterns, the moments the client stopped believing — and arm
the AM with an honest account and a credible pitch to stay. Brutally internal.

1. Confirm the client (look it up) and review the FULL relationship history — not just the
   recent quarter. If searches cap out, work backward from most recent and state which era
   the analysis fully covers.

2. Build the honest what-went-wrong account:
   - Chronic patterns: recurring issues never root-caused, repeated SLA misses, aging
     tickets, promises made in threads and dropped.
   - Inflection points: the specific incidents where sentiment visibly turned — quote the
     client's own words (plain language, no ticket IDs).
   - Relationship decay signals: escalation frequency rising, key contacts going quiet or
     curt, tickets bypassing normal channels.
   - Our misses versus their contribution (unreported issues, scope confusion, unrealistic
     expectations) — both sides, factually.

3. Build the what-went-right account with the same rigor: hard problems solved, fast saves,
   value delivered — the material for "here's what you'd be giving up."

4. Diagnose the most probable driver(s) of departure — service quality, a single
   unforgivable incident, price, a relationship (champion left, detractor rose), or an
   outside pull (new provider, internal IT hire). Label each as evidenced or hypothesis.

5. Produce win-back talking points, each honest and specific:
   - Own the misses plainly — name them before the client does; no excuses.
   - The concrete remediation offer for each chronic pattern — only commitments the desk
     can actually keep.
   - The value case: what the record shows they get, and the real cost and risk of
     switching.
   - What we'll ask of them (report issues, a regular check-in) so the reset is mutual.

6. Close with a candid save-probability read — strong case / uphill / likely unrecoverable
   — with one line of reasoning, so leadership can decide how hard to fight.

Guardrails: brutally internal — the analysis, especially the probability read and
client-contribution notes, never reaches the client; talking points are raw material, not a
script to hand over. Honesty is the deliverable: do not sand down our failures. If failures
trace to a specific technician, describe the process gap; performance evaluation goes to the
manager and no individual is named as the cause. Every remediation offer must be
operationally realistic. Quote the client's complaints accurately; never paraphrase them
milder. No fabricated history — gaps in the record are stated as gaps.
```


## Related topics

- [Post-Churn Autopsy](/skill-library/client-lifecycle/post-churn-autopsy.md)
- [Thread Partner Change Management Guide](/get-started/thread-partner-change-management-guide.md)
- [Deep Recap](/skill-library/documentation/deep-recap.md)