Endpoint types

Endpoints can be created from two types of sources: insights or SQL queries. Each has different strengths and use cases.

Insight-based endpoints

Insight-based endpoints are created from existing PostHog insights. They're ideal when you want to expose data from an insight you've already built in the UI.

Supported insight types:

• Trends
• Retention
• Lifecycle

Pros:

• Easy to create from existing insights
• Visual query builder - no SQL required
• Automatic magic variables for breakdowns and date filtering

Cons:

• Less flexible than SQL
• Limited to what the insight UI supports

Example use cases

• Expose a trends chart showing daily active users per customer
• Provide retention data for customer health dashboards

SQL-based endpoints

SQL-based endpoints are created from SQL queries written in the SQL editor. They give you full control over the query and output format.

Pros:

• Flexibility of shaping your data with SQL
• Explicit variables with precise control
• Can query any table in PostHog

Cons:

• Requires SQL knowledge

Example use cases

• Custom aggregations not available in insights
• Joining data across multiple tables
• Complex filtering logic
• Custom output formats for specific integrations

Choosing the right type

Variables

Both endpoint types support variables for runtime customization, but they work differently:

• Insight-based: Use magic variables - the breakdown property and date filters are automatically available
• SQL-based: Define explicit variables in your query using {variables.name} syntax

See the variables documentation for details.