Insights
Operational dashboards for monitoring the health and performance of your CDP. Insights give data teams and operators real-time visibility into sync execution, audience dynamics, activation coverage, and SLA compliance — without writing queries or checking individual sync runs.
Why Insights?
Running a CDP at scale means managing hundreds of syncs, dozens of audiences, and multiple destinations. Individual sync run logs tell you what happened in one execution. Insights tell you what’s happening across your entire platform:
- Are syncs healthy? — Composite health scores surface failing pipelines before they become incidents.
- How are audiences changing? — Growth and shrink rates reveal whether campaigns are working or audiences are decaying.
- Is data reaching all channels? — Activation coverage matrices show gaps where audiences are defined but not activated.
- Are we meeting SLAs? — Compliance tracking against freshness and latency targets you define.
Where to Find Them
The Insights page has four tabs, and the dashboards below are the cards inside them:
| Tab | What it holds |
|---|---|
| Overview | Sync health score, success rate, row throughput and the activation coverage matrix — the whole workspace at a glance |
| Sync Analytics | The same sync cards narrowed to one sync, plus run duration |
| Audience Analytics | Audience size trend and audience overlap |
| Channels | Channel reach, the audience-by-channel matrix, and a channel field audit |
Available Dashboards
| Dashboard | What It Shows | Key Metrics |
|---|---|---|
| Sync Health | Composite health per sync from success rate, failure streak, duration stability and freshness | Health score, top errors, daily throughput, duration percentiles, schedule adherence |
| Audience Trends | Audience size over time, recorded at each sync | Member count, growth rate, per-run adds and deletes |
| Activation Coverage | Matrix of audiences against the destinations they are synced to | Unactivated audiences, unused destinations, destination utilization |
| Audience Overlap | Shared membership analysis between two or more audiences | Overlap count, overlap percentage, exclusive membership |
Insights API
All insight dashboards are powered by REST API endpoints that you can also query programmatically. This is useful for building custom monitoring dashboards, feeding metrics into your observability stack, or triggering alerts.
Every endpoint is workspace-scoped — {id} is your workspace ID. See Base URL for your instance’s API base URL and Authentication for the required Authorization and X-Workspace-ID headers.
| Endpoint | Description |
|---|---|
GET /api/v1/workspaces/{id}/insights/sync-health | Health scores and metrics for all syncs |
GET /api/v1/workspaces/{id}/insights/sync-throughput | Daily rows added, updated and deleted |
GET /api/v1/workspaces/{id}/insights/sync-duration | Daily run duration: average, P50 and P95 |
GET /api/v1/workspaces/{id}/insights/sync-sla | Whether each sync is running on its schedule, and how overdue it is |
GET /api/v1/workspaces/{id}/insights/audience-trends | Audience size time series |
GET /api/v1/workspaces/{id}/insights/audience-velocity | Per-run rows added and deleted for an audience |
GET /api/v1/workspaces/{id}/insights/audience-overlap | Shared membership between a pair of audiences |
GET /api/v1/workspaces/{id}/insights/activation-coverage | Audience-to-destination coverage matrix |
GET /api/v1/workspaces/{id}/insights/channel-activation | Activation volume by destination channel |
GET /api/v1/workspaces/{id}/insights/audience-channel-matrix | Audiences against the channels they reach |
GET /api/v1/workspaces/{id}/insights/channel-field-audit | Fields sent to each channel, for coverage gaps |
See the API Reference for full endpoint documentation with request parameters and response schemas.
How Insights Are Computed
Almost every insight is computed from run history Zeotap already records — sync runs, their row counts and durations, and the audience size snapshot each audience sync writes when it completes. Reading them costs nothing in your warehouse.
The one exception is audience overlap, which cannot be answered from history: the intersection of two audiences has to be evaluated against your data. That query runs in your warehouse, and is therefore both cached for an hour and rate-limited to ten per workspace per hour.
What follows from this:
- Insight coverage follows sync activity. An audience that is never synced records no snapshots, so it has no trend, however often it is estimated.
- History is bounded by run retention. A period longer than the runs you have kept simply has fewer points in it.
Getting Started
Navigate to Insights in the left sidebar to access the dashboards. Insights become available as soon as you have at least one sync that has completed a run. The more history available, the richer the trend analysis becomes.
For the time-based metrics — trends, schedule adherence, velocity — a week of run history is about the minimum that produces a pattern rather than a pair of points.