Report builder
Build your own report for a question the built-in tabs don't answer: choose a source, metrics, up to two breakdowns and filters. Available on the Reports page in the dashboard or through the API. You can save the report, download it as CSV…
Build your own report for a question the built-in tabs don't answer: choose a source, metrics, up to two breakdowns and filters. Use it on the Reports page in the dashboard or through the API. You can save the report and download it as CSV.
POST /v1/apps/{appId}/analytics/reports/run
Authorization: Bearer <analytics:read-scoped key>
{
"source": "deliveries",
"metrics": ["sent", "open_rate"],
"dimensions": ["day", "channel"],
"filters": { "channel": ["email", "ios_push"] },
"sort": { "by": "sent", "direction": "desc" },
"limit": 100,
"days": 30
}Sources
| Source | Metrics | Breakdowns | Filters |
|---|---|---|---|
deliveries | sent, delivered, failed, opened, clicked, delivery_rate, open_rate, click_rate, failure_rate | day, week, month, channel, provider, campaign, status | channel, provider, campaign, status |
events | count, unique_users, value_sum, value_avg | day, week, month, name, source | name, source |
conversions | conversions, revenue, unique_users | day, week, month, campaign, event_name, kind | campaign, event_name, kind |
Fields not in the list are rejected. Personal data (user ID, email) cannot be a breakdown; the response contains aggregates only.
Calculation
- Rates are the same as in the built-in reports:
open_rate = opened / sent,click_rate = clicked / sent,delivery_rate = delivered / sent,failure_rate = failed / (sent + failed). - Time breakdowns are in UTC. Weeks start on Monday.
- Period:
days(last N days) orfromandto. They cannot be given together; if neither is given, the last 30 days are used. - Default sort: chronological if there is a time breakdown, otherwise the first metric in descending order. Empty rates always stay at the end, in either direction.
Limits
- At most 8 metrics, 2 breakdowns, 50 values per filter, 1,000 rows.
- At most 5,000 groups are read. If that is exceeded,
truncated: trueis returned andtotalscomes back empty, because totals over incomplete data would be misleading. When the row limit is exceeded,truncated: trueis also returned, buttotalscovers all groups. unique_userscannot be summed across rows: it comes back empty intotals.
Saving
| Endpoint | Scope | Description |
|---|---|---|
GET /v1/apps/{appId}/analytics/reports | analytics:read | Saved reports |
POST /v1/apps/{appId}/analytics/reports | admin | { "name": "...", "definition": { ... } } |
GET /v1/apps/{appId}/analytics/reports/{id} | analytics:read | A single report |
PATCH /v1/apps/{appId}/analytics/reports/{id} | admin | Name or definition |
DELETE /v1/apps/{appId}/analytics/reports/{id} | admin | Deletes it; data is not affected |
You can save at most 100 reports per app. Only the definition is stored; a report that uses days reads the current period on every run. In the dashboard,
saving and deleting require the data export permission (analyst and above); viewers can run reports.
User paths
Shows where users go after an event, or how they got to it. A funnel counts the steps you ask about; path analysis also shows the steps you didn't ask about.
Prediction models
Predicts a user's churn risk, or the likelihood that they will perform a specific event, from their past behavior. You use the scores in audiences and journeys.