{ "id": "9pmjBc7dqrEagOoX", "meta": { "templateCredsSetupCompleted": true }, "name": "AI Customer Churn Predictor", "tags": [], "nodes": [ { "id": "1f854286-b67a-42f9-9711-0056c93f9a0d", "name": "Sticky Note", "type": "n8n-nodes-base.stickyNote", "position": [ -32, -384 ], "parameters": { "width": 900, "height": 1504, "content": "## AI Customer Churn Predictor\n\nAnalyzes user behavior data and predicts churn probability using machine learning algorithms.\n\n## How it works\n\n1. **Trigger** \u2014 Runs daily at 2 AM to analyze active customers\n2. **Collect** \u2014 Fetches customer profiles, activity logs, and transaction history from database\n3. **Engineer** \u2014 Extracts features: login frequency, spend patterns, support tickets, engagement metrics\n4. **Predict** \u2014 Uses ML model to calculate churn probability (0-100%)\n5. **Score** \u2014 Classifies as HIGH RISK / MEDIUM RISK / LOW RISK\n6. **Remediate** \u2014 Creates retention campaigns for at-risk customers\n7. **Store** \u2014 Saves predictions to analytics database\n8. **Notify** \u2014 Sends alerts and dashboards via Email & Slack\n\n## Setup steps\n\n1. **Database** \u2014 Connect your customer database (PostgreSQL/MySQL)\n2. **ML Model** \u2014 Configure ML API endpoint or use Python ML node\n3. **Analytics DB** \u2014 Set up table for churn predictions\n4. **Slack/Email** \u2014 Add webhook URLs and SMTP credentials\n5. **Retention API** \u2014 Configure CRM/marketing platform for campaigns\n6. **Test** \u2014 Run manually first, then schedule daily execution\n\n**AI Customer Churn Predictor**\n\nThis n8n workflow runs daily to analyze active customer behavior, engineers relevant features from usage and transaction data, applies a machine learning or AI-based model to predict churn probability, classifies risk levels, triggers retention actions for at-risk customers, stores predictions for tracking, and notifies relevant teams.\n\n### Key Insights\n- Prediction accuracy heavily depends on feature quality \u2014 ensure login frequency, spend trends, support interactions, and engagement metrics are consistently captured and up-to-date.\n- Start with simple rule-based scoring or AI prompting (e.g., OpenAI/Claude) before integrating full ML models for easier testing and faster value.\n- High-risk thresholds (e.g., >70%) should be tuned based on your actual churn data to avoid alert fatigue or missed opportunities.\n\n### Usage Guide\n- Import the workflow into n8n and configure credentials for your customer database (PostgreSQL/MySQL), ML API (if external), analytics DB, Slack webhook, SMTP/email, and CRM/retention platform.\n- Define feature extraction queries and thresholds carefully in the relevant nodes \u2014 test with a small customer subset first.\n- If using AI/LLM for prediction, refine the prompt to include clear examples of churn signals.\n- Run manually via the **Execute workflow** button with sample data to validate data flow, scoring logic, and notifications.\n- Once confident, activate the daily schedule.\n\n### Prerequisites\n- Customer database with readable tables for users, activity logs, transactions, and support interactions\n- ML integration option: either an external ML API endpoint, Python code node with scikit-learn/simple model, or LLM node (OpenAI, Claude, etc.) for probabilistic scoring\n- Separate analytics database (or same DB) with a table ready for churn predictions (customer_id, date, churn_prob, risk_level, etc.)\n- SMTP credentials or email service for alerts\n- Slack webhook URL (optional but recommended for team notifications)\n- CRM or marketing automation API access (e.g., HubSpot, ActiveCampaign, Klaviyo) for creating retention campaigns/tasks\n\n### Customization Options\n- Adjust the daily trigger time or make it hourly for near real-time monitoring of high-value accounts.\n- Change risk classification thresholds or add more tiers in the scoring logic node.\n- Enhance the prediction step: switch from LLM-based to a trained ML model (via Hugging Face, custom endpoint, or Code node).\n- Personalize retention actions: use AI to generate custom email content/offers based on the customer's behavior profile.\n- Add filtering (e.g., only high-value customers > certain MRR) to focus retention efforts.\n- Extend notifications: integrate with Microsoft Teams, Discord, or create tickets in Zendesk/Jira for follow-up.\n- Build feedback loop: after actual churn occurs, update a training dataset or adjust weights/rules in future runs.\n" }, "typeVersion": 1 }, { "id": "2ae5a72b-8d07-4274-8d1d-233863fb9a08", "name": "Sticky Note1", "type": "n8n-nodes-base.stickyNote", "position": [ 928, 112 ], "parameters": { "color": 3, "width": 500, "height": 768, "content": "## 1. Trigger & Collect Data\n\nDaily schedule to fetch customer profiles and behavioral data" }, "typeVersion": 1 }, { "id": "43bb8c82-df36-49f3-89fe-70d908cd5e4a", "name": "Sticky Note2", "type": "n8n-nodes-base.stickyNote", "position": [ 1472, 256 ], "parameters": { "color": 5, "width": 388, "height": 448, "content": "## 2. Feature Engineering\n\nExtract and calculate behavioral features for ML model" }, "typeVersion": 1 }, { "id": "45eec6c4-bb3c-40ae-91fd-f3523ced9e7b", "name": "Sticky Note3", "type": "n8n-nodes-base.stickyNote", "position": [ 1920, 320 ], "parameters": { "color": 4, "width": 596, "height": 408, "content": "## 3. ML Prediction\n\nRun churn probability predictions and score customers" }, "typeVersion": 1 }, { "id": "a0bd3b99-1e56-4d62-9abb-b60829950996", "name": "Sticky Note4", "type": "n8n-nodes-base.stickyNote", "position": [ 2592, 192 ], "parameters": { "color": 5, "width": 550, "height": 612, "content": "## 4. Classify & Remediate\n\nRoute customers by risk level and create retention campaigns" }, "typeVersion": 1 }, { "id": "197eba11-4fc5-477a-9ec7-73b512f5351c", "name": "Sticky Note5", "type": "n8n-nodes-base.stickyNote", "position": [ 3232, 208 ], "parameters": { "color": 3, "width": 550, "height": 580, "content": "## 5. Report & Notify\n\nStore results and deliver churn analytics to stakeholders" }, "typeVersion": 1 }, { "id": "e6ec0958-1d51-45be-a2d3-e9e2edf6b9da", "name": "Daily churn analysis at 2 AM", "type": "n8n-nodes-base.scheduleTrigger", "position": [ 1056, 496 ], "parameters": { "rule": { "interval": [ { "field": "cronExpression", "expression": "0 2 * * *" } ] } }, "typeVersion": 1.2 }, { "id": "4c3ff3c6-e6a4-4a16-8a9d-ce732dbccdc3", "name": "Fetch active customer profiles", "type": "n8n-nodes-base.httpRequest", "position": [ 1280, 304 ], "parameters": { "url": "https://api.example.com/customers/active", "options": {} }, "typeVersion": 4.2 }, { "id": "ca121dcf-1812-463d-8388-10266013e886", "name": "Fetch customer activity logs (30 days)", "type": "n8n-nodes-base.httpRequest", "position": [ 1280, 496 ], "parameters": { "url": "https://api.example.com/logs/activity?days=30", "options": {} }, "typeVersion": 4.2 }, { "id": "ed96d0b9-b6e3-408d-b955-2c3cf2080712", "name": "Fetch transaction history (90 days)", "type": "n8n-nodes-base.httpRequest", "position": [ 1280, 688 ], "parameters": { "url": "https://api.example.com/transactions?days=90", "options": {} }, "typeVersion": 4.2 }, { "id": "5ac08f4f-5321-40ad-81a1-f027fd635970", "name": "Merge customer and activity data", "type": "n8n-nodes-base.merge", "position": [ 1504, 496 ], "parameters": { "mode": "combine", "options": {} }, "typeVersion": 3 }, { "id": "39030266-1388-4a8f-8b10-4bb70898b93e", "name": "Engineer behavioral features", "type": "n8n-nodes-base.code", "position": [ 1728, 496 ], "parameters": { "mode": "runOnceForEachItem", "jsCode": "// Feature engineering for churn prediction\nconst customer = $input.item.json;\n\n// --- Login Frequency ---\nconst loginDays = customer.activity_logs?.filter(log => log.type === 'login').length || 0;\nconst daysActive = customer.days_since_signup || 1;\nconst loginFrequency = Math.round((loginDays / 30) * 100) / 100; // logins per month\nconst loginTrend = (customer.login_last_7_days || 0) - (customer.login_previous_7_days || 0);\n\n// --- Spend Patterns ---\nconst totalSpend = customer.transactions?.reduce((sum, t) => sum + (t.amount || 0), 0) || 0;\nconst avgTransactionValue = customer.transactions?.length > 0 ? totalSpend / customer.transactions.length : 0;\nconst spendTrend = ((customer.spend_last_30 || 0) - (customer.spend_previous_30 || 0)) / (customer.spend_previous_30 || 1);\nconst daysLastPurchase = customer.days_since_last_purchase || 999;\n\n// --- Engagement Metrics ---\nconst supportTickets = customer.support_tickets || 0;\nconst featureUsage = customer.features_used || 0;\nconst emailEngagement = customer.email_open_rate || 0; // percentage\nconst accountAge = customer.days_since_signup || 1;\n\n// --- Subscription Details ---\nconst planValue = customer.plan_value || 0;\nconst isFreeTier = customer.plan_type === 'free';\nconst daysUntilRenewal = customer.days_until_renewal || 365;\nconst hasCardOnFile = customer.has_payment_method || false;\n\n// --- Feature Adoption ---\nconst adoptionScore = (featureUsage / (customer.available_features || 1)) * 100;\nconst apiCalls = customer.api_usage || 0;\nconst storageUsage = customer.storage_used_percent || 0;\n\n// --- Anomaly Flags ---\nconst suspiciousLogin = loginTrend < -5; // drop in logins\nconst noPurchaseInDays = daysLastPurchase > 60;\nconst lowEngagement = loginFrequency < 2; // less than 2 logins/month\nconst highSupportVolume = supportTickets > 5;\nconst negativeSpendTrend = spendTrend < -0.3; // 30% drop in spend\n\n// --- Compile feature vector ---\nconst features = {\n customer_id: customer.id,\n email: customer.email,\n signup_date: customer.signup_date,\n plan_type: customer.plan_type,\n \n // Core features\n login_frequency: loginFrequency,\n login_trend: loginTrend,\n days_last_purchase: daysLastPurchase,\n total_spend: totalSpend,\n avg_transaction_value: avgTransactionValue,\n spend_trend: spendTrend,\n account_age_days: accountAge,\n \n // Engagement\n support_tickets: supportTickets,\n feature_usage_count: featureUsage,\n adoption_score: adoptionScore,\n email_engagement: emailEngagement,\n api_calls: apiCalls,\n storage_usage_percent: storageUsage,\n \n // Subscription\n plan_value: planValue,\n is_free_tier: isFreeTier,\n days_until_renewal: daysUntilRenewal,\n has_card_on_file: hasCardOnFile,\n \n // Anomalies\n flag_suspicious_login: suspiciousLogin,\n flag_no_purchase_60days: noPurchaseInDays,\n flag_low_engagement: lowEngagement,\n flag_high_support_volume: highSupportVolume,\n flag_negative_spend_trend: negativeSpendTrend,\n \n // Calculated risk indicators\n anomaly_count: [suspiciousLogin, noPurchaseInDays, lowEngagement, highSupportVolume, negativeSpendTrend].filter(Boolean).length\n};\n\nreturn {\n json: features\n};" }, "typeVersion": 2 }, { "id": "a6e2a4e3-05e5-426f-b6e6-8502ea446f3d", "name": "Call ML churn prediction model", "type": "n8n-nodes-base.httpRequest", "position": [ 1952, 496 ], "parameters": { "url": "https://ml-api.example.com/predict/churn", "method": "POST", "options": {}, "sendBody": true, "bodyParameters": { "parameters": [ { "name": "features", "value": "={{ {login_frequency: $json.login_frequency, login_trend: $json.login_trend, days_last_purchase: $json.days_last_purchase, total_spend: $json.total_spend, spend_trend: $json.spend_trend, support_tickets: $json.support_tickets, adoption_score: $json.adoption_score, anomaly_count: $json.anomaly_count} }}" } ] } }, "typeVersion": 4.2 }, { "id": "241db3b1-befb-4b8e-9758-f84ae5d600c8", "name": "Score and classify churn risk", "type": "n8n-nodes-base.code", "position": [ 2176, 496 ], "parameters": { "mode": "runOnceForEachItem", "jsCode": "// Score churn risk and classify into risk tiers\nconst data = $input.item.json;\nconst mlPrediction = data.churn_probability || 0; // 0-1 scale\n\n// Convert to 0-100 scale\nconst churnScore = Math.round(mlPrediction * 100);\n\n// Classification logic\nlet riskLevel = 'LOW_RISK';\nlet actionRequired = false;\nlet campaignType = 'none';\nlet priority = 'low';\n\nif (churnScore >= 70) {\n riskLevel = 'HIGH_RISK';\n actionRequired = true;\n campaignType = 'aggressive_retention';\n priority = 'critical';\n} else if (churnScore >= 50) {\n riskLevel = 'MEDIUM_RISK';\n actionRequired = true;\n campaignType = 'engagement_boost';\n priority = 'high';\n} else if (churnScore >= 30) {\n riskLevel = 'EARLY_RISK';\n actionRequired = false;\n campaignType = 'preventive';\n priority = 'medium';\n}\n\n// Risk factors summary\nconst riskFactors = [];\nif (data.flag_no_purchase_60days) riskFactors.push('No purchase in 60+ days');\nif (data.flag_low_engagement) riskFactors.push('Low login frequency');\nif (data.flag_negative_spend_trend) riskFactors.push('Declining spend');\nif (data.flag_high_support_volume) riskFactors.push('High support tickets');\nif (data.flag_suspicious_login) riskFactors.push('Unusual login activity');\n\n// Retention strategy\nlet recommendedAction = '';\nif (churnScore >= 70) {\n recommendedAction = 'Immediate outreach: Offer discount, upgrade incentive, or VIP support';\n} else if (churnScore >= 50) {\n recommendedAction = 'Engagement campaign: Personalized tutorials, feature highlights';\n} else if (churnScore >= 30) {\n recommendedAction = 'Monitor closely: Nurture with educational content and updates';\n}\n\nconst timestamp = new Date().toISOString();\nconst predictionDate = new Date().toLocaleDateString('en-US', { year: 'numeric', month: 'long', day: 'numeric' });\n\nreturn {\n json: {\n // Customer & Prediction Identity\n customer_id: data.customer_id,\n email: data.email,\n plan_type: data.plan_type,\n \n // Churn Score & Classification\n churn_probability_score: churnScore,\n churn_probability_percentage: churnScore + '%',\n risk_level: riskLevel,\n risk_level_numeric: riskLevel === 'HIGH_RISK' ? 3 : riskLevel === 'MEDIUM_RISK' ? 2 : riskLevel === 'EARLY_RISK' ? 1 : 0,\n \n // Key Metrics\n login_frequency: data.login_frequency,\n days_since_last_purchase: data.days_last_purchase,\n total_lifetime_value: data.total_spend,\n spend_trend_percent: Math.round(data.spend_trend * 100),\n support_tickets_90days: data.support_tickets,\n feature_adoption_score: Math.round(data.adoption_score),\n account_age_days: data.account_age_days,\n \n // Risk Indicators\n risk_factors: riskFactors,\n risk_factor_count: riskFactors.length,\n anomaly_flags_count: data.anomaly_count,\n \n // Action & Campaign\n action_required: actionRequired,\n campaign_type: campaignType,\n priority: priority,\n recommended_action: recommendedAction,\n \n // Timestamp\n prediction_date: predictionDate,\n prediction_timestamp: timestamp,\n \n // Original features for reference\n features: data\n }\n};" }, "typeVersion": 2 }, { "id": "e37bc7ba-bd03-41d3-9be4-0744bbb5b2d7", "name": "Route by risk level", "type": "n8n-nodes-base.switch", "position": [ 2400, 480 ], "parameters": { "rules": { "values": [ { "outputKey": "High Risk", "conditions": { "options": { "leftValue": "", "caseSensitive": true, "typeValidation": "strict" }, "combinator": "and", "conditions": [ { "operator": { "type": "string", "operation": "equals" }, "leftValue": "={{ $json.risk_level }}", "rightValue": "HIGH_RISK" } ] }, "renameOutput": true }, { "outputKey": "Medium Risk", "conditions": { "options": { "leftValue": "", "caseSensitive": true, "typeValidation": "strict" }, "combinator": "and", "conditions": [ { "operator": { "type": "string", "operation": "equals" }, "leftValue": "={{ $json.risk_level }}", "rightValue": "MEDIUM_RISK" } ] }, "renameOutput": true }, { "outputKey": "Low Risk", "conditions": { "options": { "leftValue": "", "caseSensitive": true, "typeValidation": "strict" }, "combinator": "and", "conditions": [ { "operator": { "type": "string", "operation": "equals" }, "leftValue": "={{ $json.risk_level }}", "rightValue": "EARLY_RISK" } ] }, "renameOutput": true } ] }, "options": {} }, "typeVersion": 3 }, { "id": "c276d68c-d114-4894-a2c1-7b7a2f620762", "name": "Create retention campaign task", "type": "n8n-nodes-base.httpRequest", "position": [ 2624, 400 ], "parameters": { "url": "https://api.example.com/campaigns/create", "method": "POST", "options": {}, "sendBody": true, "bodyParameters": { "parameters": [ { "name": "customer_id", "value": "={{ $json.customer_id }}" }, { "name": "campaign_type", "value": "={{ $json.campaign_type }}" }, { "name": "churn_score", "value": "={{ $json.churn_probability_score }}" }, { "name": "risk_level", "value": "={{ $json.risk_level }}" }, { "name": "recommended_action", "value": "={{ $json.recommended_action }}" }, { "name": "priority", "value": "={{ $json.priority }}" } ] } }, "typeVersion": 4.2 }, { "id": "ca4e1fe7-be3f-4551-88f2-321fee4b8318", "name": "Store churn predictions in database", "type": "n8n-nodes-base.postgres", "position": [ 2624, 592 ], "parameters": { "table": "churn_predictions", "schema": { "__rl": true, "mode": "list", "value": "public" }, "columns": { "value": {}, "schema": [], "mappingMode": "autoMapInputData", "matchingColumns": [], "attemptToConvertTypes": false, "convertFieldsToString": false }, "options": {} }, "credentials": { "postgres": { "name": "" } }, "typeVersion": 2.5 }, { "id": "77b5a3e0-50ac-4a7a-90fd-5df6a4245a73", "name": "Generate churn analytics report", "type": "n8n-nodes-base.code", "position": [ 2848, 496 ], "parameters": { "mode": "runOnceForEachItem", "jsCode": "// Generate HTML report for churn predictions\nconst data = $input.item.json;\n\nconst statusColor = data.risk_level === 'HIGH_RISK' ? '#ef4444' \n : data.risk_level === 'MEDIUM_RISK' ? '#f59e0b' \n : '#22c55e';\n\nconst statusIcon = data.risk_level === 'HIGH_RISK' ? '\ud83d\udea8' \n : data.risk_level === 'MEDIUM_RISK' ? '\u26a0\ufe0f' \n : '\u2705';\n\nconst reportHtml = `\n\n\n\n \n Churn Prediction Report - ${data.customer_id}\n \n\n\n
\n

${statusIcon} Customer Churn Risk Assessment

\n

Customer: ${data.email} | Plan: ${data.plan_type} | Generated: ${data.prediction_date}

\n
${data.risk_level.replace(/_/g, ' ')}
\n
\n\n
\n
${data.churn_probability_score}%
Churn Risk Score
\n
\\$${Math.round(data.total_lifetime_value)}
Lifetime Value
\n
${data.login_frequency.toFixed(1)}
Logins/Month
\n
${data.adoption_score}%
Feature Adoption
\n
\n\n
\n

\ud83d\udcca Key Metrics

\n \n \n \n \n \n
Days Since Last Purchase${data.days_since_last_purchase}
Spend Trend (90 days)${data.spend_trend_percent > 0 ? '+' : ''}${data.spend_trend_percent}%
Support Tickets (90 days)${data.support_tickets_90days}
Account Age${data.account_age_days} days
\n
\n\n
\n

\u26a0\ufe0f Risk Factors (${data.risk_factor_count})

\n \n
\n\n
\n

\ud83c\udfaf Recommended Action

\n
\n ${data.recommended_action}\n

\n Campaign Type: ${data.campaign_type.replace(/_/g, ' ').toUpperCase()}\n
\n
\n\n
AI Customer Churn Predictor | ${data.prediction_timestamp}
\n\n`;\n\nconst notificationMessage = `${statusIcon} *CHURN RISK ALERT \u2014 ${data.email}*\\n\\n` +\n `*Risk Level:* ${data.risk_level.replace(/_/g, ' ')}\\n` +\n `*Churn Score:* ${data.churn_probability_score}%\\n` +\n `*Lifetime Value:* \\$${Math.round(data.total_lifetime_value)}\\n\\n` +\n `*Key Metrics:*\\n` +\n `\u2022 Logins/Month: ${data.login_frequency.toFixed(1)}\\n` +\n `\u2022 Days Since Purchase: ${data.days_since_last_purchase}\\n` +\n `\u2022 Spend Trend: ${data.spend_trend_percent > 0 ? '+' : ''}${data.spend_trend_percent}%\\n` +\n `\u2022 Feature Adoption: ${data.adoption_score}%\\n\\n` +\n `*Risk Factors:* ${data.risk_factors.join(', ')}\\n` +\n `*Action:* ${data.recommended_action}`;\n\nreturn {\n json: {\n ...data,\n reportHtml,\n notificationMessage,\n notificationTitle: `${statusIcon} [${data.customer_id}] ${data.risk_level.replace(/_/g, ' ')} - Churn Score: ${data.churn_probability_score}%`,\n shouldNotify: data.action_required,\n shouldNotifyAll: true\n }\n};" }, "typeVersion": 2 }, { "id": "749cee1c-71ed-4c46-acf5-2bffb16fd094", "name": "Filter at-risk customers", "type": "n8n-nodes-base.filter", "position": [ 3072, 496 ], "parameters": { "options": {}, "conditions": { "options": { "leftValue": "", "caseSensitive": false, "typeValidation": "strict" }, "combinator": "and", "conditions": [ { "id": "notify-condition", "operator": { "type": "boolean", "operation": "true" }, "leftValue": "={{ $json.action_required }}" } ] } }, "typeVersion": 2 }, { "id": "859e2c70-dded-45ee-a8a0-118a16762657", "name": "Post churn alert to Slack", "type": "n8n-nodes-base.httpRequest", "position": [ 3296, 400 ], "parameters": { "url": "YOUR_SLACK_WEBHOOK_URL", "method": "POST", "options": {}, "jsonBody": "={\n \"text\": \"{{ $json.notificationTitle }}\",\n \"blocks\": [\n {\n \"type\": \"section\",\n \"text\": {\n \"type\": \"mrkdwn\",\n \"text\": \"{{ $json.notificationMessage }}\"\n }\n }\n ]\n}", "sendBody": true, "specifyBody": "json" }, "typeVersion": 4.2 }, { "id": "c871536b-fb9f-4b73-ba8e-9666959da6ab", "name": "Email report to customer success team", "type": "n8n-nodes-base.emailSend", "position": [ 3296, 592 ], "parameters": { "html": "={{ $json.reportHtml }}", "options": {}, "subject": "={{ $json.notificationTitle }}", "toEmail": "user@example.com", "fromEmail": "noreply@example.com" }, "credentials": { "smtp": { "name": "" } }, "typeVersion": 2.1 }, { "id": "f6b94a2d-c388-4596-9438-f8b593307b22", "name": "Log analysis completion", "type": "n8n-nodes-base.code", "position": [ 3520, 496 ], "parameters": { "mode": "runOnceForEachItem", "jsCode": "// Log churn analysis completion\nconst timestamp = new Date().toISOString();\nconst data = $input.item.json;\n\nconsole.log(`\u2705 CHURN ANALYSIS: Customer ${data.customer_id} | Risk: ${data.risk_level} | Score: ${data.churn_probability_score}% | Action: ${data.campaign_type} | Time: ${timestamp}`);\n\n// Aggregate statistics\nconst stats = $getWorkflowStaticData('global').stats || {\n totalAnalyzed: 0,\n highRiskCount: 0,\n mediumRiskCount: 0,\n lowRiskCount: 0,\n avgChurnScore: 0,\n lastUpdated: null\n};\n\nif (data.risk_level === 'HIGH_RISK') stats.highRiskCount++;\nelse if (data.risk_level === 'MEDIUM_RISK') stats.mediumRiskCount++;\nelse stats.lowRiskCount++;\n\nstats.totalAnalyzed++;\nstats.lastUpdated = timestamp;\n\n$getWorkflowStaticData('global').stats = stats;\n\nreturn {\n json: {\n success: true,\n customer_id: data.customer_id,\n risk_level: data.risk_level,\n churn_score: data.churn_probability_score,\n timestamp,\n cumulativeStats: stats\n }\n};" }, "typeVersion": 2 } ], "active": false, "settings": { "executionOrder": "v1" }, "versionId": "d4a74901-f1bc-46d0-ae0a-88864d51c3df", "connections": { "Route by risk level": { "main": [ [ { "node": "Create retention campaign task", "type": "main", "index": 0 }, { "node": "Store churn predictions in database", "type": "main", "index": 0 } ], [ { "node": "Create retention campaign task", "type": "main", "index": 0 }, { "node": "Store churn predictions in database", "type": "main", "index": 0 } ], [ { "node": "Store churn predictions in database", "type": "main", "index": 0 } ] ] }, "Filter at-risk customers": { "main": [ [ { "node": "Post churn alert to Slack", "type": "main", "index": 0 }, { "node": "Email report to customer success team", "type": "main", "index": 0 } ] ] }, "Post churn alert to Slack": { "main": [ [ { "node": "Log analysis completion", "type": "main", "index": 0 } ] ] }, "Daily churn analysis at 2 AM": { "main": [ [ { "node": "Fetch active customer profiles", "type": "main", "index": 0 }, { "node": "Fetch customer activity logs (30 days)", "type": "main", "index": 0 }, { "node": "Fetch transaction history (90 days)", "type": "main", "index": 0 } ] ] }, "Engineer behavioral features": { "main": [ [ { "node": "Call ML churn prediction model", "type": "main", "index": 0 } ] ] }, "Score and classify churn risk": { "main": [ [ { "node": "Route by risk level", "type": "main", "index": 0 } ] ] }, "Call ML churn prediction model": { "main": [ [ { "node": "Score and classify churn risk", "type": "main", "index": 0 } ] ] }, "Create retention campaign task": { "main": [ [ { "node": "Generate churn analytics report", "type": "main", "index": 0 } ] ] }, "Fetch active customer profiles": { "main": [ [ { "node": "Merge customer and activity data", "type": "main", "index": 0 } ] ] }, "Generate churn analytics report": { "main": [ [ { "node": "Filter at-risk customers", "type": "main", "index": 0 } ] ] }, "Merge customer and activity data": { "main": [ [ { "node": "Engineer behavioral features", "type": "main", "index": 0 } ] ] }, "Fetch transaction history (90 days)": { "main": [ [ { "node": "Merge customer and activity data", "type": "main", "index": 1 } ] ] }, "Store churn predictions in database": { "main": [ [ { "node": "Generate churn analytics report", "type": "main", "index": 0 } ] ] }, "Email report to customer success team": { "main": [ [ { "node": "Log analysis completion", "type": "main", "index": 0 } ] ] }, "Fetch customer activity logs (30 days)": { "main": [ [ { "node": "Merge customer and activity data", "type": "main", "index": 1 } ] ] } } }