That is an odd requirement. Can I assume the table schema is changing? If it isn't then it would likely be better to truncate the table instead. If you delete the table then you'll need to recreate the table before you import. The import process imports into an existing table, it won't create a new one first.
To delete and recreate the table you'll need to use Execute SQL
to delete the existing table and create the new table before your Data Flow task runs to actually import the data. That should be pretty straightforward but note that if the table has any relationships, triggers or indices then you'll need to recreate them as well. Like I said a truncation to get rid of the data may be easier.
For a general walkthrough of the process of recreating the table refer to this similar question that was already asked. The answerer walks through the process of creating the table. You just need to add the table deletion before the creation.