How to filter imported customers? – Help Center

Bloomreach does not natively track which import a customer profile originated from. This means you cannot directly filter customers based on the import that created or updated their profiles.

However, there is a workaround. You can use a custom customer property to identify customers imported in a specific batch. By adding an import marker during import, you can later use this property in customer filters to include or exclude imported customers from reports, segmentations, and audiences.

Add an import marker

Add a marker that identifies one import batch. Create a property such as imported_group, and assign the same value to every imported row, such as summer_sale_2026_06_30.

Store the marker as a customer attribute. Then use a customer filter based on that attribute value later.

Set up the import marker

Go to Data & Assets > Imports > + New import > Customers.
Upload your file and preview the data.
Map at least one column to a customer ID in the customer IDs section, so Bloomreach knows which profiles to update.
Add the import marker in one of these ways:
- Map an existing column in your file to a customer attribute.
- Create a new attribute during mapping if the property does not exist yet.
- Click + on the far right of the preview table to add a static value column if your file does not contain the marker column.
Enter one value for the whole batch, such as newsletter_import_june_2026 or crm_sync_batch_2026_06_30.
Set the correct data type for the marker attribute.
Run the import.
Wait for both the import and processing stages to finish before you use the data.

Example import file

The following example shows how you can add an imported_group column to your customer import file:

registered,email,first_name,imported_group
100001,alice@example.com,Alice,summer_sale_2026_06_30
100002,bob@example.com,Bob,summer_sale_2026_06_30
100003,carol@example.com,Carol,summer_sale_2026_06_30

In this example:

registered is the customer ID (hard ID) used to match each row to an existing customer profile.
imported_group is the custom customer property used to identify the import batch.
All customers in this batch receive the same imported_group value.

Filter customers from a specific import

Once the import and processing stages are complete, you can use the import marker in a customer filter.

Create a customer filter based on the imported_group attribute and select the appropriate condition.

For example:

Filter condition	Purpose
`imported_group` equals `summer_sale_2026_06_30`	Include customers from this specific import batch.
`imported_group` does not equal `summer_sale_2026_06_30`	Exclude customers with this import marker.

You can use these filters in reports, segmentations, and audiences to include or exclude the relevant customer group.

Note: Customers without an imported_group value may be handled differently depending on the filter conditions. If you need to include customers whose marker is missing or empty, account for this explicitly in your filter configuration.

Follow best practices

Use a single clear value for each import batch.
Use a naming pattern that stays easy to scan later, such as source_campaign_date.
Create a new value for each import batch you may want to analyze or exclude later.
Keep the marker value consistent across all rows in the same batch.
Set the correct data type so the attribute filters correctly.

Understand the limitations

This method works only if you store your own marker as a customer attribute during the import. You cannot filter one import batch later unless you add that marker first. The marker becomes available only after the import and processing stages finish.

When you reimport customer data:

Import a new value to overwrite the existing value.
Import a blank value to overwrite the existing value.
Leave the attribute out of the import to keep the existing value unchanged.

If you run repeated imports and need last-write-wins behavior, map update_timestamp. Bloomreach compares the incoming timestamp with the existing property timestamp and overwrites the value only if the incoming value is newer. If you do not map update_timestamp, the platform uses the time when the import starts.