According to the Adobe documentation, file based destinations are destinations that allow exporting data from Adobe Experience Platform as files to an external system or service via SFTP or Amazon S3 protocols. These destinations can export profile attributes, segment memberships, or both, using various file formats and compression methods. Reference: https://experienceleague.adobe.com/docs/experience-platform/destinations/catalog/file-based.html?lang=en

According to the Adobe documentation, timestamp ordered is a merge method that can be used to create a custom merge policy in Adobe Experience Platform for an upcoming use case to ensure profile attributes take precedence from latest datasets and are applied across all datasets automatically. Timestamp ordered merge method prioritizes profile attributes based on their ingestion timestamp, with newer values taking precedence over older values. Timestamp ordered merge method also applies globally across all datasets within a tenant. Reference: https://experienceleague.adobe.com/docs/experience-platform/profile/merge-policies/methods.html?lang=en#timestamp-ordered

According to the Adobe documentation, custom date is a type of segmentation event recency that allows specifying a fixed date or a relative date based on a field in the dataset schema. This type of recency can be used to evaluate new users each day into a segment based on a specific date, such as the anniversary of their subscription date. Reference: https://experienceleague.adobe.com/docs/experience-platform/segmentation/ui/segment-builder.html?lang=en#event-recency

According to the Adobe documentation, I2 is a type of identity label that indicates data that is indirectly identifiable, such as hashed email addresses or phone numbers. These datasets can be shared per company policies, as they do not contain directly identifiable data, such as I1 or 11 labels. Reference: https://experienceleague.adobe.com/docs/experience-platform/data-governance/dule/dule-overview.html?lang=en#identity-labels

The most likely explanation for the dramatic drop in segment size is that the majority of browsers on the website are not authenticated. This means that they do not have a loyalty account or a unique identifier that can be matched with the gender attribute from the loyalty system. Therefore, they are excluded from the segment that ANDs the gender attribute with the browsing behavior. To increase the segment size, the retailer could use other methods to identify or infer the gender of anonymous browsers, such as cookies, device IDs, or third-party data.