Why Do Syncs Process More Rows Than Expected?

Question

Why do syncs process significantly more rows than the number of new records generated in the source?

Environment

Connector: Rakuten Advertising

Answer

Fivetran uses a 100-day replay window for Rakuten Advertising incremental syncs. During each sync, we process data from the previous 100 days to capture updates and corrections to historical records.

Due to source limitations, we can't reliably update existing records in place. Instead, we insert previously synced records again when the source returns them with updates or corrections. For more information, see our Rakuten Advertising limitations.

As a result, syncs may process more rows than the source created since the previous sync. This is expected and doesn't indicate a full re-sync. You can't configure your connection to process only newly created records.