# Analytics & Intelligence Canonical URL: https://babadook.tech/sub-modules/analytics-intelligence Entity: sub-modules Content hash: 01058d660dfea621d80b845ec038c4a84712e46502e865f7c17cb458d8231ebf Behavioral, product, and marketing analytics with causal inference and ML-powered insights. Behavioral, product, and marketing analytics with causal inference, propensity modeling, and ML-powered insights that drive data-informed decisions. Analytics & Intelligence processes millions of events to surface patterns, predict outcomes, and recommend actions. ## Analytics That Go Beyond Dashboards subtitle: From event collection to causal inference — understand not just what happened, but why it happened and what will happen next. ### Behavioral Analytics desc: Session replay, heatmaps, funnel analysis, and journey mapping with automatic anomaly detection. ### Causal Inference desc: Go beyond correlation. Identify true cause-and-effect relationships between user actions and business outcomes. ### Propensity Modeling desc: Predict conversion likelihood, churn risk, and lifetime value using ML models trained on your data. ### Marketing Attribution desc: Multi-touch attribution with Shapley value modeling across all channels and touchpoints. ### Processing throughput value: 1M+ unit: events/sec ### Insight latency value: <5min unit: min ### ML model types value: 7 unit: models ## from-data-to-decisions From Data to Decisions DXP Analytics processes every interaction — page views, clicks, scrolls, form submissions, purchases, and custom events — into a unified behavioral stream. Events are enriched with session context, visitor identity, and causal metadata before storage. Causal inference goes beyond correlation to identify which variables actually drive outcomes. Using techniques like difference-in-differences, instrumental variables, and synthetic control, DXP Analytics answers questions like 'Did this feature change actually increase conversions, or was it just seasonal traffic?' Propensity models are trained on your data continuously. Seven model types cover conversion propensity, churn risk, lifetime value prediction, next-best-action, content affinity, segment migration, and anomaly detection. Models retrain on a configurable schedule with automatic performance monitoring. ### data Viz ## Analytics & Intelligence Specifications - **Event Processing:** 1M+ events/sec with sub-second ingestion - **Insight Latency:** <5 minutes from event to actionable insight - **ML Models:** 7 model types with automatic retraining - **Session Replay:** Full DOM replay with privacy masking - **Funnel Analysis:** Multi-step funnels with automatic drop-off detection - **Attribution:** Shapley value multi-touch with custom windows - **Data Retention:** Configurable 90-day to unlimited event retention ## faq section ### Section 1 What is causal inference and why does it matter? Causal inference uses statistical techniques to determine whether a change actually caused an outcome, not just correlated with it. This prevents false conclusions like attributing a revenue increase to a feature change when it was actually driven by a marketing campaign. ### Section 2 How are propensity models trained? Models are trained on your historical data using gradient-boosted decision trees. Training happens on a configurable schedule, and model performance is monitored with automatic alerts when accuracy degrades below threshold. ### Section 3 Can DXP Analytics replace my existing analytics tool? Yes. DXP Analytics covers behavioral analytics (session replay, heatmaps, funnels), product analytics (feature usage, retention, engagement), and marketing analytics (attribution, campaign performance, channel ROI) in a single integrated system. ### Section 4 How does privacy masking work in session replay? Sensitive fields are automatically masked based on configurable rules. Input fields, credit card numbers, and PII are replaced with masked placeholders in replay recordings. Custom masking rules can be applied per page or per component. ## features ml: true causal: true product: true marketing: true behavioral: true ## Relationships ### products - DXP — https://babadook.tech/products/dxp ### industries - Media & Publishing — https://babadook.tech/industries/media-publishing ### capabilities - Advanced Analytics — https://babadook.tech/capabilities/advanced-analytics ### modules - Behavioral Analytics — https://babadook.tech/modules/behavioral-analytics - Causal Inference Engine — https://babadook.tech/modules/causal-inference - Marketing Analytics — https://babadook.tech/modules/marketing-analytics-module - ML & Propensity Models — https://babadook.tech/modules/ml-propensity-models - Product Analytics — https://babadook.tech/modules/product-analytics ### use cases - Behavioral Analytics ### deployment types - Cloud (SaaS) - Self-Hosted