The Merge node combines two data streams in your workflow by matching records based on field values — similar to a join in SQL. While the Bond node creates relationships between entities, the Merge node produces a fully flattened dataset. Records that match are combined into a single row with fields from both inputs.
Basic Example
Input 1
Input 2
Result (Match field: ID)
How it works
Drag two data streams into the Merge node. These can come from any upstream nodes in your workflow — Get Data, Filter, another Merge, etc. Choose the match fields
Specify which fields to join on between the two inputs (e.g., Account ID from Input 1 = Account ID from Input 2). You can add multiple field pairs to create a composite key — records must match on all mapped fields.| Setting | Description |
|---|
| Input 1 Field | The field from the first data stream to match on |
| Input 2 Field | The field from the second data stream to match on |
Resolve conflicts
If both inputs carry the same entity, pick whether to keep the data from Input 1 or Input 2.| Setting | Description |
|---|
| Entity Preferences | Per-entity setting that controls which input’s columns to keep when both inputs carry the same entity |
| Virtual Preferences | Per-virtual-column setting that controls which input’s data to keep when both inputs carry the same virtual column. Defaults to Input 1 if not specified |
Choose a join type (optional)
Under the Advanced tab, select a join type to control which records appear in the output. The default is Match & Track.| Join Type | SQL Equivalent | Description |
|---|
| Match & Track | Inner Join | Only records with matches in both inputs (default) |
| Keep All Source | Left Join | All records from Input 1, matched with Input 2 where possible |
| Keep All Destination | Right Join | All records from Input 2, matched with Input 1 where possible |
| Keep Everything | Full Outer Join | All records from both inputs, matched where possible |
| Only Unmatched Source | Left Excluding | Only Input 1 records that have no match in Input 2 |
| Only Unmatched Destination | Right Excluding | Only Input 2 records that have no match in Input 1 |
| Only Unmatched from Both | Full Excluding | Records from either input that have no match in the other |
The Merge node joins two data streams that are already flowing through your workflow. To pull in new data from a connected integration based on a Bond, use the Get Data node instead.
Output
The Merge node produces two output streams:
| Output | Description |
|---|
| Matched | Records that successfully joined across the two inputs |
| Unmatched | Records that did not find a match in the other input |
Example with join types
Input 1 (Subscriptions)
Input 2 (Ages)
Match Field: ID
Match & Track (Inner Join)
Matched:
Unmatched:
Keep All Source (Left Join)
| ID | Name | Age |
|---|
| 1 | Dan | 30 |
| 2 | Yael | 25 |
| 3 | Noa | null |
Keep Everything (Full Outer Join)
| ID | Name | Age |
|---|
| 1 | Dan | 30 |
| 2 | Yael | 25 |
| 3 | Noa | null |
Merge vs Bond
| Feature | Bond | Merge |
|---|
| Purpose | Create relationships | Combine datasets |
| Output | Linked records | Flattened table |
| Behavior | Join-like (logical) | Join + materialize |
| Use case | Enrichment & exploration | Final dataset construction |
Best Practices
- Always verify your match fields — IDs should align exactly
- Use multiple fields if a single key is not unique
- Resolve conflicts intentionally — don’t default blindly
- Check Unmatched output to catch missing data
- Choose merge mode based on your goal:
- Analysis → Keep Everything
- Clean dataset → Match & Track
- Debugging → Unmatched modes
- Bond Node — creates logical relationships between entities rather than producing a flattened table
- Union Node — stacks rows vertically instead of joining side-by-side
