Developer DocsData Ingestion & Tracking
How Intempt Works: A One-Page Mental Model
Understand Intempt's behavioral tracking framework — how automatic signals and manual API calls combine into a unified user behavior dataset.
How Intempt Works: A One-Page Mental Model
Intempt operates as a behavioral model, not a simple event collector. It combines two complementary signal sources to build a complete picture of each user.
Two Signal Sources
| Source | Description |
|---|---|
| Automatic signals | Emitted by the platform — page views, sessions, DOM interactions |
| Manual API calls | Initiated by your application code — purchases, identifications, consents |
Together, these create a unified user behavior dataset that powers analytics, journeys, personalization, recommendations, and experiments — without requiring duplicate tracking.
Core principle: Automatic events give context. Manual APIs give meaning.
The Six Foundational APIs
| API | Role |
|---|---|
consent() | The gatekeeper — determines whether user events are accepted or rejected |
identify() | Assigns a known identity to a user, linking events across sessions and devices |
alias() | Merges two identifiers into a single user profile for identity reconciliation |
track() | Captures general user actions, system events, and product catalog data |
record() | Manages richer, structured events with user/account attributes for significant business moments |
group() | Associates users and events with accounts or organizations for multi-tenant contexts |
Implementation Sequence
Proper instrumentation follows a logical progression:
- Session initialization — Intempt starts tracking the session automatically
- Consent verification — Check and record user consent before tracking sensitive data
- User identification — Link anonymous events to a known user with
identify() - Identity reconciliation — Merge identifiers across devices/channels with
alias() - Behavioral tracking — Record actions and events with
track()orrecord() - Organizational association — Associate the user with an account using
group() - Product synchronization — Keep the product catalog in sync for recommendations
What Each Layer Enables
| Layer | Enables |
|---|---|
| Automatic events | Analytics dashboards, session replays, heatmaps |
identify() + alias() | Cross-device journeys, unified profiles |
track() + record() | Segments, triggers, revenue attribution |
group() | Account-level analytics, B2B journeys |
consent() | GDPR/CCPA compliance, suppression logic |
| Product sync | Recommendations, catalog-based personalization |
