- DATE:
- AUTHOR:
- The Fountain team
One Predictable Matching Rule for Data Pipelines
Data Pipelines use one clear rule to decide whether an imported row updates an existing record or creates a new one.
Each import entity matched records its own way, so the same file could update one entity and duplicate another, and a renamed record or changed email could quietly create duplicates or skip rows.
What's new
One rule for every entity — jobs, locations, location groups, workers, openings, prospects, users and applicants all follow it: one match updates, no match creates, and more than one match fails the row with a reason.
Every mapped key counts — each entity matches on a Fountain ID, an HRIS ID and an easy key such as a name or email, with no hidden priority.
Protected HRIS IDs — a pipeline sets an HRIS ID once and never overwrites it, so a rename in your HRIS no longer creates a duplicate.
Owners by email — owner and supervisor fields accept an email as well as a user ID.
Clear failure reasons — each failed row carries a reason in the run log, and the Mapping step explains how rows are matched.
Why it matters
Imports behave the same way every time, duplicates after renames go away, and migrations no longer need workarounds like looking up internal IDs.
What to expect
New and cloned pipelines, one-time imports and default Connector Marketplace syncs use the new rule. Pipelines created earlier keep their current matching; clone a pipeline to move it to the new rule. Names are matched exactly as written, including capitalization.
Availability
Available to all Fountain customers.
