# Split
Divide a dataset into multiple branches based on selected fields or conditions.
The **Split** feature in **Flow Transformation** lets you break a large dataset into smaller, structured pieces that you can process independently in the next steps of your flow.
> This is especially helpful when a connector returns a large table or a complex object, and you only want to work with **specific portions,** such as selected columns, targeted attributes, or filtered sections of the data.
### **Use Case Example**
Imagine you have a Google Sheet storing thousands of rows of product data (Name, Category, Price, Stock, etc.). You only want to extract **specific columns**, for example, **Category and Price**, to send into another system (like an AI step, analytics tool, or external API).
Instead of passing the entire sheet, the **Split** feature lets you carve out only the fields you need.
Using **Includes Only**, you can split the dataset so only the selected fields move forward, simplifying your next steps and making your flow more efficient.
### Step 1: Add a Connector
* Start your flow by adding a connector that contains the data you want to split.
Add a ***Google Sheets*** Connector and select the action ***getDataFromSheet***.
* This step retrieves all the data from your sheet and provides it as the output for the next steps.
### Step 2: Add the Split Transformation
* After your connector step, click the '+' button to insert a new step.
* Search for **Flow Transformation Actions** and select **Split**.
* The Split step divides the data based on the input field and strategy you choose. Configure the Split step as follows:
**Fields To Split Out**
{{steps.getDataFromSheet.output}}
{% hint style="info" %}
*This references the output from the previous connector step.*
{% endhint %}
**Split Out Strategy**
The **Split Out Strategy** lets you decide whether to include only certain fields or exclude certain fields.
As an example, we will choose the *Includes Only* Strategy in this scenario.
{% hint style="info" %}
You can choose Includes Only or Excludes Only, depending on your use case.
{% endhint %}
**Input Fields**
You can select any input field you want to use for splitting, like the Major Dimensions of the data from the sheet, in this case.
Save the step.
### Step 3: Testing the Split Step
You can test the split step to ensure it works as expected:
* Test the step directly from the **Test** inside the Split step.
* Next, test the entire flow from the **Test** button at the top-right corner of the flow editor.
### Using Split Data
Once the split step is configured, the resulting data can be used in further steps of your flow.
For Example:
* Passing specific fields to another connector for processing.
* Filtering the dataset based on certain dimensions before performing calculations.
* Using split data as input for AI agents or custom code steps.
