Choosing the Right Node

Use this page as a quick reference for which Node Type to choose. Follow each scenario link for in-depth information. Use the Related packages column to find Coalesce Marketplace or built-in packages for each scenario.

Reference

Use the Scenario links for the matching section in Node types by scenario or the primary package doc. Use Related packages for every Marketplace listing or Nodes and Node Types for built-in behavior.

Scenario	Related packages
Modeling Baseline With Built-In Node Types: Stage, Dimension, Fact, Persistent Stage, View	Nodes and Node Types
Base Node Types on Snowflake, Databricks, and BigQuery: Work, Persistent Stage, Dimension, Fact, View, Code; Snowflake adds Factless Fact and SQL Stage	Snowflake Base Node Types, Databricks Base Node Types, BigQuery Base Node Types
Advanced Deploy Merge and Tracking: Dimension, Fact, Persistent Stage, Work, Factless Fact Advanced Deploy variants	Snowflake Base Node Types Advanced Deploy, BigQuery Base Node Types Advanced Deploy
Snowflake Change Feeds and Tasks: Streams and Tasks Node Types	Snowflake Streams and Tasks
When You Land Each Batch in a Staging Layer: Stage (built-in) or Work	Snowflake Base Node Types, Databricks Base Node Types, BigQuery Base Node Types
When You Keep Curated Rows Across Runs on a Business Key: Persistent Stage	Snowflake Base Node Types, Databricks Base Node Types, BigQuery Base Node Types
When You Load Only New or Changed Rows: Incremental Load; Looped Load, Run View, Grouped Incremental load, Test Passed Records, Test Failed Records, Code (Snowflake listing); Incremental Load, Code (Databricks listing)	Incremental Loading package, Incremental Nodes
When You Describe Business Entities and Slowly Changing Attributes: Dimension	Snowflake Base Node Types, Databricks Base Node Types, BigQuery Base Node Types
When You Store Measures in Fact Tables: Fact	Snowflake Base Node Types, Databricks Base Node Types, BigQuery Base Node Types
When You Record Events or Relationships Without Measures: Factless Fact (Snowflake base), Factless Fact Advanced Deploy (Snowflake and BigQuery Advanced Deploy), or Fact or Persistent Stage	Snowflake Base Node Types, Snowflake Base Node Types Advanced Deploy, Databricks Base Node Types, BigQuery Base Node Types, BigQuery Base Node Types Advanced Deploy
When You Want a Simple View Object in the Graph: built-in View or package View Node Type (Snowflake, Databricks, BigQuery); Create As View on other types per listing	Nodes and Node Types, Snowflake Base Node Types, Databricks Base Node Types, BigQuery Base Node Types
When You Own the Create and Run Templates: Code	Snowflake Base Node Types, Snowflake Base Node Types Advanced Deploy, Databricks Base Node Types, BigQuery Base Node Types, BigQuery Base Node Types Advanced Deploy, Incremental Loading package, Incremental Nodes
When You Transform Entirely in Hand-Authored SQL (Snowflake): SQL Stage	Snowflake Base Node Types
When Several Sources Feed One Conformed Node: Multi Source or chained Work	Multi Source Nodes
Advanced Deploy Options by Node Family: Snowflake and BigQuery matrices on listings	Snowflake Base Node Types Advanced Deploy, BigQuery Base Node Types Advanced Deploy
Files, stages, pipes, and APIs on Snowflake: External Data Node Types	Snowflake External Data Package
Batch files into Delta on Databricks: External data package	Databricks External data package
Dynamic tables on Snowflake: Dynamic Table Node Types	Snowflake Dynamic Tables package
Iceberg tables on Snowflake: Iceberg Node Types	Snowflake Iceberg Tables package
Materialized views on Snowflake: Materialized View Node Types	Snowflake Materialized View package
Interactive tables on Snowflake: Interactive Table Node Types	Snowflake Interactive Table package
Semantic layer on Snowflake: Semantic View and Semantic Query	Snowflake Semantic View package
Lakeflow pipelines on Databricks: Streaming and pipeline Node Types	Databricks Lakeflow Declarative Pipelines package
Materialized views on Databricks: Materialized View Node Types	Databricks Materialized View package
Near real-time stream with deferred MERGE on Snowflake: Deferred Merge streams	Snowflake Deferred Merge package
Calendar, PIVOT, UNPIVOT, and advanced SQL on Snowflake: Functional Node Types	Snowflake Functional Node Types package
Calendar, PIVOT, and UNPIVOT on Databricks: Functional node types	Databricks Functional node types package
Create or alter instead of drop-recreate on Snowflake: Create or Alter Node Types	Snowflake Create or Alter package
Metrics and profiling on Snowflake: Data quality Node Types	Snowflake Data quality package
Great Expectations-style tests: Test Utility Node Types	Snowflake Test Utility package, Databricks Test Utility package
Masking and row access policies on Snowflake: Data security views	Snowflake Data security package
Data Vault 2.0 methodology: Hub, link, satellite, PIT, and related Node Types	Data Vault by Scalefree for Snowflake, Data Vault by Scalefree for Databricks
Cortex and related ML and AI on Snowflake: Cortex Node Types	Snowflake Cortex package

Decisions before you pick a Node label

• What each row stands for: One row in the target should match one business instance at the level you need (for example one customer, one order, or one shipment line).
• History: Overwrite in place (Type 1 style) or keep prior versions (Type 2 style or event-level facts).
• Keys: Business keys anchor merges on Dimensions and many Facts; Facts without a business key often use insert-only loads. Defaults are described in Nodes and Node Types.
• Persistence: Staging-style types often replace each batch; persistent types keep prior rows unless the template replaces them.