Recued
Menu
← Back to recipes

Churn Risk Analyzer

by recued-core v1 52 views

Computes churn risk score from support tickets, NPS, usage trends, and engagement. Generates retention actions when risk is high.

How it works 33 steps

Inspect the data fetches, transforms, gates, and output this recipe runs.

Data fetch
▼ account account-reader-hubspot
Fetch reader from hubspot
▼
Process (32 steps)
⚠ check_account guard
Stop if account is missing
■ ticket_risk_raw math
Calculate account support ticket count 90d times 10
■ is_ticket_high compare
Check if ticket risk raw is greater than 100
■ ticket_risk switch
Map is ticket high to one of: true, false
■ nps_shifted math
Calculate account nps score plus 100
■ nps_normalized math
Calculate nps shifted divided by 2
■ nps_scaled math
Calculate nps normalized times 10
■ nps_risk math
Calculate 100 minus nps scaled
■ usage_risk_raw math
Calculate 100 minus account product usage trend
■ is_usage_negative compare
Check if usage risk raw is less than 0
■ usage_risk switch
Map is usage negative to one of: true, false
■ engagement_risk_raw math
Calculate account days since last positive engagement times 3
■ is_engagement_high compare
Check if engagement risk raw is greater than 100
■ engagement_risk switch
Map is engagement high to one of: true, false
■ part_ticket math
Calculate ticket risk times 0.25
■ part_nps math
Calculate nps risk times 0.25
■ part_usage math
Calculate usage risk times 0.3
■ part_engagement math
Calculate engagement risk times 0.2
■ sum_weights math
Calculate part ticket plus part nps
■ sum_remaining math
Calculate part usage plus part engagement
■ churn_risk math
Calculate sum weights plus sum remaining
■ churn_risk_ge_threshold compare
Check if churn risk is at least setting ai threshold
■ should_ai all
Check if all conditions are true
■ hashed_account hash_replace
Anonymize sensitive fields for safe AI processing
skip: step.should_ai equal false
■ ticket_issue compare
Check if ticket risk is greater than 50
■ nps_issue compare
Check if nps risk is greater than 50
■ usage_issue compare
Check if usage risk is greater than 50
■ engagement_issue compare
Check if engagement risk is greater than 50
■ ai_context pick
Keep only specific fields from
skip: step.should_ai equal false
✦ ai_plan ai-prompt
AI generates a response from a custom prompt
skip: step.should_ai equal false
■ overall_summary to_summary
Format results as a summary card
■ risk_checklist to_checklist
Format results as an actionable checklist
▼
Output
summary overall_summary
checklist risk_checklist
ai_analysis ai_plan
Settings 2 configurable

Configurable at install. Defaults shown — change them anytime in Recued.

enable ai toggle = on
ai threshold number = 50

Uses (2 ingredients)

Ingredient account-reader-hubspot account-reader-hubspot
Legacy AI Custom Prompt ai-prompt

Trust & control

What installing this recipe would let it do. Recued grants these permissions at install — review them there before approving.

What it's allowed to do

Read Legacy AI Custom Prompt read-only

1 more isn't listed here. You'll see it at install.

Data it touches

Warehousedata.nps_scoredata.product_usage_trenddata.support_ticket_count_90ddata.days_since_last_positive_engagement

About

Tags

hubspot accountchurnriskretentionhubspot

Details

33 steps 2 configurable settings recipe_id: analyze-churn-risk-account-hubspot