Skip to main content

When to use Splitter

Choose Splitter when one dataset needs to become multiple outputs.

Splitter is most useful when Records should be routed to different Files, groups, review queues, or downstream processes based on repeatable Rules.

Strong signs that Splitter is the right Tool​

Splitter is a strong fit when:

  • one input File contains Records for several audiences or destinations
  • Records can be assigned to groups using clear Rules
  • multiple outputs are needed from the same source
  • output groups should be created consistently across recurring runs
  • unmatched or uncategorized Records need to be isolated for review

Typical business situations​

Splitter is commonly used in situations like these:

Separate by ownership or destination​

Examples:

  • one File per region
  • one File per department
  • one File per account owner
  • separate outputs for different vendors or partners

Separate by business condition​

Examples:

  • approved, rejected, and needs-review records
  • current, past-due, and future-dated transactions
  • standard records and exception records
  • high-priority and normal-priority queues

Prepare multiple downstream files​

Examples:

  • separate upload files for different systems
  • files tailored to different review teams
  • grouped exports for reporting packages
  • batches for staged processing

Good fit vs poor fit​

Good fit​

Use Splitter when the goal can be described as:

  • "Create separate outputs for these groups."
  • "Route each Record based on these conditions."
  • "Isolate Records that do not fit the expected groups."

Poor fit​

Splitter is probably not the first Tool to use when:

  • you only need one filtered output
  • values must be cleaned before grouping is reliable
  • duplicate records need to be resolved
  • the main task is joining related files
  • you need aggregate metrics rather than split files

Ask this before you start​

Finish this sentence:

"This file needs to be divided into ..."

If you can name the output groups clearly, Splitter is often a good fit.

If the groups are vague or changing constantly, clarify the routing rule before building the configuration.

Start with output design​

Before building rules, define the outputs.

Ask:

  • What should each output represent?
  • Can a Record belong to more than one output?
  • What should happen if a Record matches no output?
  • Which output counts would look suspicious?
  • Who will use each output after the run?

Rule-of-thumb decision guide​

Use Splitter when the central task is routing one dataset into multiple outputs.

Use another Tool first when the central task is cleaning, deduplicating, joining, filtering one subset, or summarizing values.