Churn risk is the likelihood that a client account will leave, which a predictor expresses as a churn probability. In the Client Success Loop, a churn probability of 0.6 triggers the escalation SOP. The CSM and an executive sponsor respond in 4 hours, a root-cause call follows in 24 hours, and a mitigation plan follows in 72 hours.
What Churn Risk Means in the Client Success Loop
The Client Success Loop treats churn risk as a measured number rather than a feeling about an account. A churn predictor assigns each client a churn probability, and the escalation SOP reads that number directly. The method sits inside the VWCG OS, where the Client Success Loop module holds the instrument, routes, and playbooks for client success.
Escalation and Churn Response Activation is the playbook that turns that number into a governed response. It registers the predictor, sets its activation mode, codifies the escalation SOP, runs the override log, and test-fires the path. Each step carries a published rule, and the sections below take them in the order the playbook lists them.
Account-level churn risk differs from the churn rate a business reports across its whole client book. A guide to measuring churn rate covers that portfolio metric, which sits beside the account-level flag described here.
The same module publishes a first-value retention split alongside the churn threshold for every account. Clients reaching first-value inside the window retain at 85% or better, and outside it at 40% or lower. That split belongs to a separate playbook, First-Value Acceleration, rather than to escalation work.
Where Escalation Sits in the Diagnostic
The Client Success Loop Diagnostic scores each dimension on a 0 to 5 scale against written anchors. The composite score is a weighted sum, with equal weights of 0.20 per dimension across the instrument. Churn Predictor Governance is the dimension that measures how the business governs its churn predictor.
That dimension checks registration, activation mode, overrides, and audits against the published standard. The predictor must sit in the AI Register within AI Deployment Canvas, with model and business owners named. Its activation mode must match the AERI tier from Intelligent Foundations, the module that holds that score. Overrides must be documented with the 20% rule tracked, and quarterly bias and drift audits must run.
At the low anchor, a business has no predictor, or it deployed a predictor on technical readiness alone. That low anchor also describes a predictor that ignores AERI state when it starts an escalation. At the high anchor, tiered activation matches AERI, overrides follow the 20% rule, and quarterly audits come back clean. The high anchor also requires the escalation SOP to fire on 0.6 within its time bounds.
A red reading on Churn Predictor Governance fires the escalation playbook inside the Client Success Loop. That red reading also sends AI Deployment Canvas a route to register or re-register the predictor. The same route covers the QBR generator, and both tools need both owners named in the register.
Because this route fires from a scored dimension, it starts as a proposal rather than an order. Leadership confirms it at the next monthly governance review or the Core Execution review, whichever comes first. The playbook’s stated duration starts at confirmation, not at the date the dimension turned red.
What Opens the Escalation Playbook
Several conditions open Escalation and Churn Response Activation, and any of them is enough to start a run. A red reading on Churn Predictor Governance opens it, as the diagnostic section above describes. A missing or unmet escalation SOP opens it as well, even when the dimension has not turned red.
The playbook also opens when a predictor operates outside the AI Register in AI Deployment Canvas. An override rate above 20% opens it too, which ties the override log directly to the trigger list.
If a trigger arrives while the playbook is already open, it is logged against the open run. Leaders therefore track the open run for each playbook rather than a stack of parallel responses to similar signals. Every run follows the same lifecycle, which the closing section of this guide lays out step by step.
As an outside lens, a RACI chart can help a leader see who already touches each trigger today. That chart sits outside the method, because the playbook names its own Business Owner and Model Owner directly.
Registering the Predictor With Both Owners
The playbook registers the churn predictor in the AI Register within AI Deployment Canvas as its opening step. Registration names a Model Owner and a Business Owner for the predictor in that register. For this playbook, the VP of Customer Success serves as Business Owner for the churn predictor.
A data and machine learning Model Owner works alongside the VP of Customer Success on the registration. The playbook outcome lists the predictor as registered with both owners, alongside quarterly audits on the schedule. Those audits cover bias and drift, matching what the Churn Predictor Governance dimension asks the business to run.
Registered audits continue as standing work once the escalation playbook closes its run for the owner. Under Retention Loop Monitoring, the standing playbook, quarterly audit results from AI Deployment Canvas go on file.
Not sure where the gap in your customer or sales process sits? See how an operations review works →
Matching Activation Mode to the AERI Score
The playbook then sets the activation mode from the current AERI score held in Intelligent Foundations. That AERI score runs on a 0 to 100 scale, and the published bands decide how much the predictor may act. Predictor mode follows the AERI band directly, and that rule now runs as an Intelligent Foundations route.
Below 40, the predictor runs human-in-the-loop, and the model never starts the escalation SOP by itself. A person decides at that level, and the QBR generator also runs in advisory mode only.
From 40 to 70, each churn flag starts the escalation SOP once a person confirms it. That supervised band also carries a monthly accuracy review under Retention Loop Monitoring. Above 70, the churn predictor fires the escalation SOP automatically without a confirming step.
Routes that set AI mode in any module count as safety routes under the published routing rules. Safety routes tighten on the first reading and loosen only after an independent re-score clears the edge. When routes set the same parameter differently, the most restrictive value wins for safety routes. A color change also requires a margin past the band edge or a sustained reading across consecutive scorings.
Consider a business whose AERI score sits below 40 while its predictor already flags accounts. That predictor could still flag clients, but a person would decide every escalation before the SOP started.
Codifying the 4, 24, and 72-Hour Bounds
The playbook codifies the escalation SOP in the SOP Codex with the published time bounds for every flag. A churn probability of 0.6 triggers the escalation SOP for that specific client account. The CSM and an executive sponsor respond within 4 hours of the flag reaching the account. A root-cause call follows within 24 hours, and a mitigation plan follows within 72 hours.
Capital planning can also receive a flag when the client ties to a funded project. A churn probability of 0.6 on a client tied to a funded project opens a hold review in Agile Capital Allocation. That review covers the project’s next tranche at the next Forum, and release still follows the existing evidence rule.
Revenue concentration runs on a separate route that sits beside the escalation SOP in the module. Any single client at or above 15% of trailing-12-month revenue becomes a valuation input for Exit and Acquisition Layer. That route fires from a measured value, so it acts on the reading without waiting for confirmation.
As an outside lens, unit economics can help a leader read that concentration in commercial terms. The published route still sets the rule, and the lens adds no step to the escalation SOP.
Running the Override Log and the 20% Rule
The playbook runs the override log weekly as part of its standard work for the churn predictor. Under the published rule, CSM overrides above 20% of flags mean retrain the model or coach the CSM. The rule names both responses without stating which response applies in any given case.
An override rate above 20% also counts as a trigger that opens this playbook again. The outcome of the playbook requires an override log running with the 20% rule in place. At the high anchor of the dimension, overrides stay documented and the 20% rule stays enforced.
Retention Loop Monitoring tracks the override rate as part of its standing duties after a run closes. That standing playbook also reviews weekly recalculation output for new amber and red client accounts.
As an outside lens, a fishbone diagram can help a team sort override reasons into groups. The method does not prescribe that diagram, and the 20% rule remains the governing test for overrides.
Test-Firing the Path and Closing the Run
The playbook test-fires the escalation path on a live or simulated flag as its final listed step. Its outcome requires that the test-fired escalation met every time bound in the SOP Codex. A test that misses a bound leaves the outcome unmet until the path meets every bound.
The full outcome has several parts that leaders can check directly against the playbook. Both owners appear in the register beside the predictor, and the activation mode matches the AERI tier. The test-fired escalation met every time bound, and the override log runs with the 20% rule. Audits sit on the schedule, which completes the published outcome for the playbook.
The VP of Customer Success owns the playbook as Business Owner, with a data and machine learning Model Owner. The stated duration is 4 weeks, counted from the confirmation of the route that opened the run. Every playbook moves through open, running, met or missed, escalated, and closed as its lifecycle.
An outcome met, with an independent re-score at amber or better, closes the run for the owner. Someone other than the owner re-scores the dimension before the run can close on that basis. A missed outcome at the end of 4 weeks escalates to the owner’s manager. It also goes onto the next Core Execution quarter planning agenda for leadership review.
When the outcome is met, any standing duty the playbook set up continues under Retention Loop Monitoring. That standing playbook runs a 30-day pulse check for each amber dimension of the diagnostic.
Build the process your team should actually be measured against. Book a 30-minute operations review →
Frequently Asked Questions
What does churn risk mean in the Client Success Loop?
Churn risk is the likelihood that a client account will leave, which a predictor expresses as a churn probability. A churn probability of 0.6 triggers the escalation SOP. The Escalation and Churn Response Activation playbook governs that predictor, its activation mode, the SOP time bounds, and the weekly override log.
What are the time bounds in the escalation SOP?
After a churn probability of 0.6, the CSM and an executive sponsor respond within 4 hours. A root-cause call follows within 24 hours, and a mitigation plan follows within 72 hours. The playbook codifies these bounds in the SOP Codex and test-fires the path on a live or simulated flag.
What opens the Escalation and Churn Response Activation playbook?
A red reading on Churn Predictor Governance opens the playbook. A missing or unmet escalation SOP opens it too. So does a predictor outside the AI Register in AI Deployment Canvas, or an override rate above 20%. A red reading from the diagnostic starts as a proposal until leadership confirms it.
How does the AERI score set the predictor’s activation mode?
The AERI score from Intelligent Foundations runs from 0 to 100. Below 40, the predictor runs human-in-the-loop and a person decides every escalation. From 40 to 70, each churn flag starts the escalation SOP once a person confirms it. Above 70, the predictor fires the escalation SOP automatically.
What does the 20% rule say about CSM overrides?
CSM overrides above 20% of flags mean retrain the model or coach the CSM. The playbook runs the override log weekly with the 20% rule in place. An override rate above 20% also triggers the playbook. Retention Loop Monitoring keeps tracking the override rate as a standing duty after the run closes.
How does a service business owner begin escalation and churn response work?
A service business owner can begin with a review from Sales Roadmaps through the VWCG OS service. The engagement applies the Client Success Loop Diagnostic, including the Churn Predictor Governance dimension and its anchors. From there, the playbook registers the predictor, sets the activation mode, codifies the SOP, and test-fires the path.
