How do the Data Merge modes work with Imports

The import merge modes work before any of the data is actually imported. The match mode attempts to find existing data in the system first, and based on whether it finds 0 matches or 1 match, the merge mode determines what happens next (more than 1 match results in the row erroring):

• InsertOnly: if the match mode found 0 matches, import the data. Otherwise (1 match), skip.
• UpdateOnly: if the match mode found 1 match, update that matched data. Otherwise (0 matches), the row errors.
• Upsert: if the match mode found 0 matches, import the data. Otherwise (1 match), update the matched data. This works well as a default since it does both inserts and updates (upsert is a combination of the words update and insert).

After that point, we know we are importing the current row of data, so the import service goes column by column and imports each cell on that row from the import file. As it goes column by column, that’s when the field rules get applied.

In the “Edit Field Rules” link, for many text-only fields (like email, phone number, address, etc.) you can set the field rule to UpdateUnlessEmpty, which will import that data unless the text in the import file is empty. If users are looking to only import if the text already in the system is empty, that is not currently an option.