Modeling your Data

Sundial helps you make sense of raw event data by transforming it into structured, queryable building blocks for analysis. These blocks—metrics, dimensions, and entities—form the foundation for scalable product analytics. Modeling your data makes it easier to analyze behavior, segment users, and track performance with precision and consistency.

Sundial’s modeling layer allows organizations to define these objects clearly, consistently, and collaboratively—so that analytics always stay aligned with the business logic.

Metrics

A metric is a time-based aggregation that represents a quantitative measurement—such as signups, sessions, purchases, or active users. Metrics are defined over a specific grain (e.g., daily, weekly) and often include filters or groupings to tailor them to specific analytical needs. Some common examples of metrics include "Total Signups", "Monthly Active Users", or "Revenue per User".

Dimensions

Dimensions add contextual richness to metrics. Each dimension has a set of possible values, which can be used to segment, slice, break-down and filter your metrics, enabling more granular insights. Some common examples of dimensions include "Platform" (e.g., iOS, Android), "Country" (e.g., US, UK), or "Device Type" (e.g., mobile, desktop).

Entities

Entities are a special type of dimension that represent core business objects—like users, accounts, sessions, or products. They are uniquely identifiable and can be used to join multiple metrics and dimensions. Entities act as analytical anchors, allowing you to conduct entity-level analysis (e.g., retention by user, revenue per account). Some common examples of entities include "User ID", "Session ID", or "Product ID".