B5SALES & REVENUE

Renewal and churn-risk early warning

Continuously watches signals across the customer base — usage patterns, support ticket trends, NPS responses, payment behavior, key contact changes — and flags accounts where renewal risk is rising. Outputs a ranked list to the customer success team with the evidence behind each flag, so the team can intervene early rather than discovering the problem 30 days before renewal. The pattern's value comes from catching the slow churn signals (declining logins, champion leaving) months before they show up in renewal conversations.

WHERE THIS FITS

BUSINESS SHAPES

B2B servicesProduct company

VOLUME THRESHOLD

Below 50 active customer accounts a month, the payback rarely earns the build. Patterns this shape reliably pay back at 500+.

REQUIREMENTS · 5 REQUIRED, 2 OPTIONAL

Requirements describe capabilities the pattern needs in your environment, not the vendors you must buy. Any system that fills a requirement satisfies it — that’s what makes the catalog portable across the long tail of SMB tooling.

account_record_store
REQUIREDREAD + WRITErequest
Customer accounts being monitored, including the renewal date and current commercial status.
DATA SHAPE
Account records with contract dates, renewal date, ARR/MRR, plan, owner. Pattern writes a risk score and reason codes.
COMMONLY FILLED BY
- CRM with account records and renewal dates
- customer success platform
- subscription billing system as the source of truth for accounts
product_usage_signal
REQUIREDREADbatch
How customers are actually using the product. The single most predictive churn signal in most SaaS businesses.
DATA SHAPE
Per-account usage metrics over time: active users, key actions, feature adoption, session frequency. Aggregated or detailed.
COMMONLY FILLED BY
- product analytics platform
- data warehouse with usage tables
- internal usage metrics dashboard
- (for non-product businesses: project hours logged, deliverables shipped, etc.)
support_interaction_signal
REQUIREDREADbatch
Support volume, sentiment, and resolution patterns per account. Rising tickets or unresolved issues are churn precursors.
DATA SHAPE
Per-account support metrics: ticket volume trend, resolution time, sentiment of recent tickets, repeat-issue rate.
COMMONLY FILLED BY
- support system metrics
- aggregated data from A1/A2 patterns if live
- customer success notes archive
contact_health_signal
REQUIREDREADrequest
Who's still at the account and engaged. Key contacts leaving is one of the strongest churn predictors.
DATA SHAPE
Per-account contact list with role, last interaction, engagement level. Changes flagged when a key contact disappears.
COMMONLY FILLED BY
- CRM with up-to-date contact records
- data from B2 hygiene pattern flagging contact changes
- support system records of who's been opening tickets
payment_behavior_signal
RECOMMENDEDREADbatch
Late payments, disputes, downgrades. Commercial signals that hint at intent.
DATA SHAPE
Per-account payment events: invoice age, dispute flags, plan changes.
IF MISSING
Pattern still produces good risk signals but misses commercial-side cues. Recommend at least monthly visibility.
COMMONLY FILLED BY
- billing system with payment status
- accounts receivable system
- finance team's monthly account health report
csm_alert_destination
REQUIREDWRITEevent
Where the customer success team sees flagged accounts. Must be in their normal working surface, not a separate dashboard.
DATA SHAPE
Per-account risk flag with score, reason codes, evidence, suggested intervention. Appears in CSM's daily view.
COMMONLY FILLED BY
- field on the account record in CRM with sorting and filtering
- dedicated risk view in the customer success platform
- Slack alerts for accounts crossing thresholds
intervention_outcome_loop
RECOMMENDEDWRITEbatch
Feedback on whether interventions worked, used to tune signal weights over time.
DATA SHAPE
Per-flagged-account outcome: did they renew, did intervention happen, what was the action taken.
IF MISSING
Pattern produces flags but can't learn which signals predict actual churn. Strongly recommend closing the loop.
COMMONLY FILLED BY
- monthly churn retrospective with outcome tagging
- renewal outcome data fed back from billing
- structured CSM playbook outcomes

RUNTIME FLOW · 8 STEPS

01
On a regular cadence (daily for top accounts, weekly for others), compute the risk score per account
account_record_store
02
Read product usage signals and detect trend changes (declining usage, dormant features, drop in active users)
product_usage_signal
03
Read support signals (volume trend, sentiment, unresolved issues)
support_interaction_signal
04
Check contact health: are key champions still there, are new contacts being added, is engagement broad or narrowing
contact_health_signal
05
Read payment behavior signals if available
payment_behavior_signal
DECISION Skip if payment_behavior_signal not filled.
06
Combine signals into a risk score with reason codes citing the specific evidence
07
Write the score back to the account record; surface flagged accounts to the CSM alert destination
account_record_storecsm_alert_destination
DECISION Only alert on threshold crossings and large score changes.
08
Capture intervention outcomes after the fact and feed back to tune signal weights
intervention_outcome_loop

EMISSIONS · 2

Structured outputs this pattern produces. Other patterns and client systems can subscribe to them, which is how the catalog composes over time.

risk_score_stream
Per-account risk score updated continuously, with reason codes.
CONSUMED BY
- customer success dashboards
- renewal forecasting
- executive account reviews
churn_signal_correlation
Which specific signals actually predict churn in this customer base, refined over time from outcome data.
CONSUMED BY
- model tuning
- product team understanding what drives retention