In any ROI import, there is the ability to select one or multiple Deduplication Keys. Checking the deduplication key box for a field tells ROI to deduplicate the data based on that chosen field. This means that the system will look for data in that field to append the record if it already exists or to create a new one if it does not. For example, if I have Email as a column on my file that I'm importing into ROI and I select that field as the deduplication key, ROI will look to match the file being imported on Email. If the email address "email@example.com" exists as a contact's email in ROI and it is included in the file being imported, it will append the file fields to that record in ROI rather than create a new contact record. However, if the email address "firstname.lastname@example.org" does not exist as a contact's email in ROI but it is on the file, the system will not be able to find a match for that field and will create a new contact record with the data from the file. This is why it is recommended to set the deduplication key to the most specific identification, such as email or ROI ID.
You do have the ability to select multiple deduplication keys when importing data into ROI. The system reads this as 'Key1 AND Key2 AND...' This means that when multiple deduplication keys are selected for an import in ROI, the system will look for the first dedupe AND the other dedupes selected and if it can't find a match, it will create a new record with these values and update all other fields except for the key fields from the mapping. There are no incremental checks based on keys. The system will check against records in database by ANDing all the keys.