Skip to content
Go to www.furball.tech
  • There are no suggestions because the search field is empty.

Import Only New Rows with Delta Imports

Compare an updated file with a completed import using all shared columns or selected keys

Overview

Use Import delta when a source export grows between imports, such as a webinar attendee list or a recurring partner report. Upload the updated file from a completed import's history entry, compare it with that import's original file, and prepare only rows that are new under your chosen matching rule.

The comparison is between files, not a check of records currently in HubSpot. A Completed preparation means its files were prepared; it does not prove that someone imported those files into HubSpot.

What this helps you do

  • Avoid preparing the same source rows again.
  • Keep the previous import's preparation and mapping choices as a starting point.
  • Choose whether a changed row should be processed again or only a new identifier should count.
  • Review the delta before downloading files or confirming automated ingestion.

Start a delta import

  1. Choose the correct portal and find the previous import under Previous imports.
  2. Confirm that it is the right baseline and that any intended HubSpot import was completed.
  3. Choose Import delta on its Completed entry and upload the updated CSV or XLSX file.
  4. Check the worksheet, header row and skipped columns. Review the comparison summary and greyed-out rows.
  5. Leave Match rows by at All shared columns, or choose one or more columns.
  6. Continue through Prepare, Map, Details and Review. Check the carried-over settings and the new row counts.
  7. Download the prepared delta files or confirm automated ingestion as usual.

This creates a separate import. It does not overwrite the previous import or modify either original file. Earlier row corrections and manually skipped rows are not copied as decisions for the new file.

Choose how rows are compared

All shared columns

The default compares values across columns that both files share. Surrounding spaces are ignored, but differences in letter case still count.

  • An identical row already present in the previous file is skipped.
  • A new row or a row with a changed value in a shared column is included.
  • Added or removed columns are shown in the comparison. A change only in a newly added column does not make an otherwise matching row new.
  • Repeated identical rows are counted: if the previous file has one copy and the new file has two, the additional copy is new.

Both files need at least one shared column heading. Check the header row if no shared columns are found.

One or more matching columns

Choose Email, an event occurrence ID or another stable column when that value defines what is new. Selecting several columns uses their values together.

  • Keys ignore letter case and surrounding spaces.
  • If a key already appears in the previous file, the row is skipped even if other values changed.
  • Every occurrence of a new key is included. This setting does not deduplicate the new file.
  • If every selected key value in a row is blank, the Studio compares that row using all shared columns instead.
  • Each selected column must exist in both files and be included in the comparison.

For example, matching by Email skips an existing attendee even if their job title changed. Use all shared columns if you want that changed information to be processed.

Keep repeated activity when it matters

For custom events, one contact can have several genuine occurrences. Matching only by Email would skip a new activity for an email already in the previous file. Use an occurrence ID, a combination such as Email and activity timestamp, or all shared columns when those better describe a unique occurrence.

Keep Deduplicate rows by email off when repeated emails represent separate activities. See Prepare and Send Contact Custom Events.

Boundaries and safeguards

  • Import delta is available for completed preparation jobs, not failed or unfinished jobs.
  • The previous original file must still be available.
  • The baseline is the selected import's source file, including rows that may have been skipped during its preparation. It is not a ledger of successful HubSpot writes.
  • Rows removed from a later file do not delete HubSpot records.
  • No new rows means there is nothing to prepare; review the summary or choose a different matching rule.
  • A delta still requires a new upload and review. It is not scheduling, unattended ingestion or two-way synchronisation.

What to do next