Skip to main content
Bounty uses a hierarchy to organize the systems a workspace can connect to, the connections that have been set up, and the curated data objects people work with inside the product. Understanding this hierarchy helps explain the difference between a system Bounty supports, a connection that has been configured, and a data source that is ready to use in analysis, agents, actions, and chat.

​
Connector

A connector is a supported integration type that Bounty knows how to work with. Examples:
  • GA4
  • Meta Ads
  • Tinybird
  • Snowflake
  • HubSpot
The connector is the option you choose when setting up a data system in Bounty. It describes what kinds of systems Bounty can work with. It does not mean the workspace has connected that system yet.

​
Connection

A connection is a connector that has been set up in Bounty. Examples:
  • A GA4 connection
  • A Meta Ads connection
  • A Snowflake connection
  • A Tinybird connection
A connection gives Bounty access to data from that system and determines which tables are available for data sources. A connection can make multiple tables available. For example, a Meta Ads connection can make separate tables available for campaigns, ad sets, ads, and daily performance. Each table represents a different view of the same connected platform data.

​
Table

A table is a structured collection of records available from a connection or warehouse. Tables are usually organized around a specific object, event, or report type. Examples:
  • campaigns
  • ad_sets
  • ads
  • ga4_events
  • A deals table in a warehouse
A table is usually not the final object people work with in Bounty. It is the starting point Bounty can turn into a curated data source.

​
Data Source

A data source is the curated, product-facing view of queryable data that people use inside Bounty. Examples:
  • Meta Ads performance
  • GA4 Events
  • Deal Stage Snapshot
A data source is not the raw table itself. It gives Bounty a named shape for the data: where to query it, which table or prepared query shape backs it, which date and ID fields matter, and which fields should be available to users and AI workflows. Most data sources are backed by one selected table from a connection or warehouse. In some cases, a data source can be backed by a prepared query shape when raw tables need to be cleaned, combined, or narrowed before they are useful in Bounty. Data sources are the objects selected across metrics, analysis, agents, actions, and chat.

​
Data Source Field

A data source field is one curated field exposed by a data source. Each data source field maps to an underlying database column, but it is not just the raw column name. It carries the product-level meaning Bounty should use when showing, filtering, grouping, or explaining that data. Fields can define:
  • The underlying database column
  • The user-facing display name
  • An optional description
  • Value mappings, such as raw IDs or codes mapped to readable labels
  • Value groups, such as several raw statuses grouped under one business label
  • Whether the field is visible and usable in Bounty
Hidden or unselected columns stay in the underlying table, but they do not appear as usable data source fields in Bounty.

​
How They Work Together

Once a connector has been configured, the data hierarchy looks like this: 
Connector
  -> Connection
  -> Table
  -> Data Source
  -> Data Source Fields
When people work with data in Bounty, the data source is the main object they use. Connectors, connections, and tables explain where that data source comes from.

​
Examples

For Meta Ads: 
Meta Ads connector
  -> Meta Ads connection
  -> Campaigns, ad sets, ads, and daily performance tables
  -> Meta Ads data source
  -> Fields such as Campaign, Spend, Impressions, and Clicks
For GA4: 
GA4 connector
  -> GA4 connection
  -> Events table
  -> GA4 Events data source
  -> Fields such as Event Name, Session Source, Users, and Conversions
For warehouse data: 
Snowflake connector
  -> Snowflake connection
  -> Selected warehouse tables
  -> Curated data sources
  -> Curated data source fields
For Tinybird: 
Tinybird connector
  -> Tinybird connection
  -> Available Tinybird tables
  -> Curated data sources
  -> Curated data source fields
The key rule: customers use data sources. Connectors, connections, and tables explain how those data sources become available in Bounty.