There are 3 distinct primary import modes available to select from to determine the type of import process is to be run as shown below:

#1 - Clear Folder Before Import
This option will result in all contacts being deleted prior to any contacts being imported. This option is only relevant if the import process is designated as "All All".
#2 - <Add All> Only - Allow Dupes
All contacts will be added to the contact folder. There is no check for duplicates when this mode is selected.
#3 - <Add All> Only - Ignore Dupes
Any records found in the input data source that already exist in the target contact folder will be ignored. A duplicate is defined by designating one or more fields as "Key Fields" during the import field mapping process. A <Key Field> can be any available standard or user-defined field that is available for mapping.
#4 - <Add All> Only - Delete Dupes
This option will result in all contacts found in the contact folder that match an input data record to be deleted and will be replaced by the contact information contained in the input data source. This differs from an <Update> in that all existing contacts are deleted first. During an <Update>, no contacts are deleted, only the corresponding data in mapped fields is updated in all matching pre-existing contacts. A duplicate is defined by designating one or more fields as "Key Fields" during the import field mapping process. A <Key Field> can be any available standard or user-defined field that is available for mapping.
#5 - <Delete Matched>
This will result in contacts being deleted from the contact folder that match the <key> fields selected in the mapping phase. It's purpose is to enable a <bulk> deletion process of contacts no longer required.
#6 - <Update>
This provides the ability to update existing contacts from an input data source with several <sub-options> as to how new contacts will be processed and how individual fields will be updated.
<Add New Only>
This option will result in only new contacts being imported based on the key fields specified in the import field map. This option would be useful if new contacts are being added to a database but the primary location for information updates is done in Outlook.
<Add if new>
Determines whether or not contacts that do not already exist in the Outlook folder get added. If the option is not selected, only existing contacts will be updated.
<Update value if blank>
Determines if the value in the Outlook contact field is replaced with a blank value if the corresponding value in the input data file is also blank. Some people want all fields updated all the time regardless of value and others only want the data value changed if there is a value in the input data file. This option applies to all fields in the import field map and cannot be set on a "field by field" basis.
<Update value if Null>
NULL values are special values and are treated differently than <blank> values. To the non-technical person, it is visually impossible to determine whether the contents of the input data file is a blank or null value since both look the same - empty field. If not familiar with NULL values, then this setting should be set the same as <Update value if blank> to ensure no strange import results.
<Append Value to Keyword Fields>
Determines whether or not the contents of a keyword field are replaced with the value found in the input data file or the values are appended to the existing values in the Outlook contact item.