Glossary
Quick reference for terms used across the Pug docs, SDKs, and API. Definitions are kept concise — follow the linked pages for full explanations.
| Term | Definition |
|---|---|
| Organization | Top-level tenant. Owns members, billing, and email configuration. Projects belong to an organization. See Core concepts. |
| Project | Isolated data boundary inside an organization. Owns events, profiles, insights, and API keys. All analytics queries are project-scoped. See Core concepts. |
Public key (pub_…) | Project-scoped write credential for client SDKs. Sent as x-api-key; lets you ingest events and identify users, but cannot read analytics. Safe to ship in a browser bundle or mobile binary. See Authentication. |
Private key (prv_…) | Project-scoped read/write credential for trusted server-side code. Sent as x-api-key; reaches shared.* services (insights, profiles). Never ship in client code. See Authentication. |
| Event | A named occurrence with custom properties and a timestamp — the atomic unit of analytics data. The event name is its kind; names starting with pug. are reserved. See Events API. |
| Property | A custom key/value pair attached to an event (string, integer, double, boolean, or timestamp). Keys must not start with $. Sent in customProperties on each event. |
| Auto-property | A $-prefixed property attached automatically by the SDK or server — for example $url, $browser, $country. $-prefixed keys are reserved and cannot be set as custom properties. See Auto-properties. |
| Well-known event | A standard event name (e.g. page_view, purchase) that carries a typed schema. Custom event names not in the registry are also accepted without schema validation. See Well-known events. |
| Profile | A persistent record representing one user — identified or anonymous — with an accumulated set of traits (properties). Created or updated by identify(). See Profiles API. |
| Anonymous ID | An SDK-generated identifier in the form anon-…, created when a visitor first arrives. Events are attributed to this ID until the user is identified. |
| External ID | Your stable user identifier (e.g. a database user ID), passed as external_id to Identify. Use a value that does not change — not an email address. |
| Distinct ID | The identifier an event is attributed to: the anonymous ID before identify() is called, the external ID after. Stored as distinct_id on every event. |
| Alias | The association of multiple identifiers with one profile — most commonly the anonymous-to-identified merge on the first identify() call (there is no separate alias() method). See Profiles API. |
| Session | A window of activity grouped by the SDK. A new session starts after the idle timeout (default 30 min) or the maximum session duration (default 24 h). The session ID is synced across browser tabs. See Core concepts. |
| Insight | A server-side analytics query specification — event selection, filters, time range, granularity, aggregation, and optional breakdowns. The same spec powers dashboard charts and the InsightsService API. Types: TRENDS, SEGMENTATION, FUNNEL, RETENTION, USER_FLOW. See Insights API. |
| Connect RPC | The API transport protocol used by all Pug services. Supports the Connect, gRPC, and gRPC-Web protocols over HTTP/1.1 and HTTP/2. Request path format: /<package>.<Service>/<Method>. See Authentication and the API overview. |
| Batch | A single BatchCreate call carrying up to 1 000 events. All event ingestion goes through this endpoint — there is no single-event API. See Events API. |
| Sampling rate | An SDK option (0–1) that keeps only a fraction of events. Useful to limit volume on high-traffic pages without disabling tracking entirely. |
Further reading
- Core concepts — detailed explanations of key terms
- Authentication — keys, headers, and error codes
- Profiles API — identity lifecycle, anonymous → identified merge
- Insights API — query types and spec format
- Auto-properties — full
$-prefixed property catalog
Last updated on