Segments
Segment Metrics & Trend Interpretation
Selecting a segment shows a right-side panel with live metrics and an event list. This section explains each element and how to leverage it.
1. Top Metric Cards (3)
| Card | Composition | Meaning | Use Case |
|---|---|---|---|
| Total Events Share | segment events / all events + circular progress | Estimated cumulative event share (no date limit) | Identify influential behavior clusters |
| Active Users Share | segment unique users / all unique users | Relative audience scale | Campaign reach sizing |
| 30-day Trend | direction (up/down/flat) + % | Compares rear 7 days vs prior 7 days | Trigger root-cause exploration |
Trend Thresholds
| Direction | Rule (last 7 vs prior 7) |
|---|---|
| Up | ≥ +5% |
| Down | ≤ -5% |
| Flat | Within ±5% |
2. Event Table Columns
| Column | Description | Example |
|---|---|---|
| Timestamp | Ingest time (local/standard per system policy) | 2025-09-09 10:21:19 |
| Event Name | Event identifier | page_view, purchase |
| User ID | Anonymized/hashed internal identifier or '-' | user_xxx |
| Session ID | Session tracking key | session_xxx |
| Device Category | mobile / desktop etc. | mobile |
| Platform | OS / Browser spec | Android, iOS, Edge |
| OS Version | Semantic version string | 17.1, 13 |
| Country / City | GeoIP derived | KR / Seoul |
| Age / Gender | Demographic attributes | 24 / female |
| Language | Locale code | ko, ja-JP |
| App Version / SDK Version | Installed app & SDK versions | 2.7.3 / 3.0.7 |
| Parameters | Raw JSON key/value payload | {"page":"product"...} |
3. Analysis Patterns
| Goal | Method | Example Insight |
|---|---|---|
| Re-validate core behavior | Check if events share ≥30% | Over-broad definition → refine criteria |
| Growth/attrition detection | Focus 30-day directional change | Feature launch or regression candidate |
| Deep dive after segmentation | Manually compare card values across segments | Geo density disparity |
| Funnel improvement input | Cross-reference step drops vs segment profile | Attribute-specific friction |
4. Anomaly Checklist
| Symptom | Check |
|---|---|
| Event share collapse | SDK version distribution / recent release impact |
| User share near zero | Over-constrained filters (excess AND) |
| Spike in +% trend | Potential bot/abuse or new campaign |
| Parameter JSON bloat | Schema drift (unexpected field growth) |