churn-analysis

name	churn-analysis
description	Analyse churn risk for a customer or segment — diagnose root cause, design intervention, and define success criteria for retention.
argument-hint	[customer name, segment, or churn pattern to analyse]
user-invocable	true
allowed-tools	Read, Write, Edit, Bash, Glob, Grep

Analyse churn risk for $ARGUMENTS.

Process (sequential — do not skip steps)

Step 1: Signal Identification

What triggered this analysis? Identify all active churn signals:

Signal category	Indicators to check
Usage decline	DAU/MAU trending down, key features unused, session duration decreasing
Engagement drop	Fewer logins, stopped attending meetings, unresponsive to outreach
Relationship deterioration	NPS dropped, champion left, new decision-maker unfamiliar with product
Value gap	Stated goals not met, ROI not demonstrated, competitor evaluation underway
Commercial friction	Payment failures, pricing complaints, downgrade requests, contract negotiation stalling
Support escalation	Increasing ticket volume, repeated unresolved issues, frustrated tone in tickets

Document every active signal with specific evidence (dates, metrics, quotes).

Step 2: Timeline Reconstruction

Build a timeline of the customer's trajectory:

Date	Event	Signal
[date]	[what happened]	[what it indicates]

Look for:

• Inflection point — when did things start changing? What preceded the change?
• Accelerating decline — is the situation getting worse faster?
• Correlated events — did something specific trigger the decline? (product change, champion departure, competitor launch, internal reorg)

Step 3: Root Cause Diagnosis

Churn signals are symptoms. Diagnose the underlying cause:

Root cause category	Questions to answer	Common indicators
Product-market fit	Does the product actually solve their problem?	Never adopted core features, using workarounds, feature requests for basic functionality
Onboarding failure	Did they reach Time to First Value?	Low adoption after 30+ days, never completed setup, no established workflow
Value delivery gap	Are they achieving their stated goals?	Goals not met, ROI not demonstrated, success metrics not defined
Relationship failure	Has the relationship degraded?	Champion left, no executive sponsor, ignored outreach
Product quality	Is the product failing them?	Recurring bugs, performance issues, data loss, broken integrations
Competitive pressure	Is a competitor winning?	Feature comparisons, pricing benchmarks, evaluation signals
Internal change	Did something change on their side?	Reorg, budget cuts, strategy shift, new leadership

Rules:

• Identify the PRIMARY root cause, not just a list of everything wrong
• Distinguish between causes you can address and causes you cannot
• If the root cause is unclear, say so — don't guess

Step 4: Churn Probability Scoring

Factor	Low risk (1)	Medium risk (2)	High risk (3)	Score
Usage trend	Stable or growing	Declining < 4 weeks	Declining > 4 weeks	[1–3]
Engagement	Responsive, attending	Delayed responses	Unresponsive	[1–3]
Sponsor status	Active champion	Champion passive	No champion	[1–3]
Value realisation	Goals being met	Partially met	Not met	[1–3]
Contract timeline	>6 months to renewal	3–6 months	<3 months	[1–3]
Competitive activity	No signals	Casual mentions	Active evaluation	[1–3]

Churn probability:

• 6–9: Low risk — monitor
• 10–13: Medium risk — proactive intervention
• 14–18: High risk — urgent intervention required

Step 5: Intervention Design

Design a specific intervention based on the root cause:

Root cause	Intervention approach	Timeline
Onboarding failure	Restart onboarding — dedicated session, fast-track to value	2 weeks
Value gap	Success planning — define metrics, demonstrate ROI, remove blockers	4 weeks
Relationship failure	Executive engagement — new sponsor identification, business review	2 weeks
Product quality	Engineering escalation — prioritise fixes, dedicated support	Depends on fix
Competitive pressure	Value reinforcement — feature comparison, switching costs, executive meeting	1 week
Internal change	Stakeholder mapping — identify new decision-makers, rebuild business case	3 weeks

Each intervention must have:

• Owner — single person accountable
• First action — what happens within 48 hours
• Success criteria — measurable definition of "saved"
• Checkpoint — when to assess whether the intervention is working
• Escalation path — what happens if the intervention fails

Step 6: Retention Economics

Quantify the business impact to frame urgency:

Metric	Value
ARR at risk	[annual revenue from this customer]
Replacement cost	[cost to acquire equivalent — typically 5–7x retention cost]
Lifetime value remaining	[expected remaining contract value]
Intervention cost	[effort and resources required]
ROI of retention	[value saved vs intervention cost]

Anti-Patterns (NEVER do these)

• Discounting as first response — discounts treat the symptom (price), not the cause (value). If the product doesn't deliver value, a discount just delays churn
• Intervention without root cause — "schedule an executive meeting" without knowing what problem to solve wastes the executive's time and the customer's patience
• Ignoring the timeline — a customer declining for 6 months needs a different intervention than one who dropped suddenly last week
• Assuming you can save everyone — some churn is natural (company closed, product-market misfit, budget eliminated). Focus where intervention can make a difference
• Expansion during churn risk — never upsell an at-risk customer. Fix the health first
• Waiting for the customer to complain — by the time they announce they're leaving, they've already decided. Detect early

Output Format

# Churn Analysis: [customer/segment]

## Risk Summary
- **Churn probability:** [Low / Medium / High]
- **Risk score:** [6–18]
- **ARR at risk:** [$]
- **Time to action:** [Urgent / This week / This month]

## Active Signals
| Signal | Severity | Evidence | Duration |
|---|---|---|---|
| [signal] | [level] | [specific data] | [how long] |

## Timeline
| Date | Event | Significance |
|---|---|---|
| [date] | [event] | [what it indicates] |

## Root Cause
- **Primary:** [root cause with evidence]
- **Contributing:** [secondary factors]
- **Addressable:** [Yes / Partially / No]

## Risk Scoring
| Factor | Score (1–3) | Evidence |
|---|---|---|
| Usage trend | [n] | [detail] |
| Engagement | [n] | [detail] |
| Sponsor status | [n] | [detail] |
| Value realisation | [n] | [detail] |
| Contract timeline | [n] | [detail] |
| Competitive activity | [n] | [detail] |
| **Total** | [6–18] | |

## Intervention Plan
| Action | Owner | Timeline | Success criteria |
|---|---|---|---|
| [first action — within 48h] | [person] | [date] | [measurable outcome] |
| [follow-up actions] | [person] | [date] | [measurable outcome] |

## Retention Economics
| Metric | Value |
|---|---|
| ARR at risk | [$] |
| Replacement cost | [$] |
| Intervention ROI | [ratio] |

## Checkpoint
- **Review date:** [when to assess intervention effectiveness]
- **Success indicators:** [what improvement looks like]
- **Escalation trigger:** [when to escalate if intervention isn't working]

Related Skills

• /customer-success:health-assessment — run a health assessment first to identify at-risk accounts, then do churn analysis on the flagged ones.