Digital analytics, behavior tracking, and audience insights.
Key Functions
| Function | Description | AI Opportunity |
|---|
| Event Tracking | User actions, custom events | Auto-capture |
| Session Recording | Replay user sessions | Insight extraction |
| Heatmaps | Click, scroll, attention maps | Pattern detection |
| Funnel Analysis | Conversion paths, drop-offs | Optimization suggestions |
| Cohort Analysis | User segments over time | Predictive cohorts |
| Attribution | Multi-touch, channel impact | ML attribution |
| Real-time | Live dashboards, alerts | Anomaly detection |
| User Identification | Cross-device, anonymous to known | Identity resolution |
| A/B Testing | Experiment analysis | Auto-optimization |
| Reporting | Dashboards, scheduled reports | Natural language queries |
Core Entities
| Entity | Fields | Volume | Sensitivity |
|---|
| Events | name, properties, timestamp, user | Very High | Medium |
| Users | ID, properties, first seen, last seen | High | High |
| Sessions | duration, pages, device, location | Very High | Medium |
| Pages | URL, title, engagement metrics | High | Low |
| Recordings | video, events, duration | Very High | High |
| Experiments | variants, metrics, results | Medium | Low |
| Funnels | steps, conversion rates | Medium | Low |
| Segments | rules, membership | Low | Low |
Integration Points
| System | Data Flow | Direction |
|---|
| Website/App | Events, user data | Inbound |
| CRM | User enrichment | Bi-directional |
| Marketing | Campaign attribution | Bi-directional |
| Advertising | Conversion tracking | Outbound |
| Data Warehouse | Full export | Outbound |
| Product | Feature usage | Inbound |
Data Retention
| Data Type | Typical Retention | Compliance Driver |
|---|
| Event data | 12-24 months | Storage costs |
| Session recordings | 30-90 days | Privacy, storage |
| Aggregated data | Indefinite | Trend analysis |
| User profiles | Until deletion request | GDPR/CCPA |
Evaluation Criteria
| Criteria | Weight | Notes |
|---|
| Data accuracy | High | Sampling vs complete |
| Privacy compliance | High | GDPR, cookie consent |
| Query flexibility | Medium | Custom analysis |
| Real-time | Medium | Time-sensitive decisions |
| Integration depth | Medium | Your stack |
| Pricing model | Medium | Event-based can scale poorly |
Market Leaders
| Product | Strength | Best For |
|---|
| Google Analytics | Free, ubiquitous | Basic needs |
| Mixpanel | Product analytics, funnels | Product teams |
| Amplitude | Behavioral analytics | Product-led growth |
| FullStory | Session replay, DX insights | UX teams |
| Heap | Auto-capture, retroactive | Less instrumentation |
| PostHog | Open source, all-in-one | Engineering teams |
AI Disruption Potential
| Function | Current State | 2027 Projection |
|---|
| Insight discovery | Manual exploration | Auto-surfaced |
| Anomaly detection | Basic alerts | Predictive |
| Attribution | Model-based | True incrementality |
| Segmentation | Rule-based | Behavioral clustering |
| Reporting | Query-based | Natural language |
| Session analysis | Manual review | Auto-summarized |
Build vs Buy: Buy. Analytics infrastructure is commodity. Focus on using data, not collecting it. Self-host (PostHog) if privacy is paramount.