"Unique key" or "Primary Key" are concepts that are coming from relational database. You can find a lot of explanations/documentation on internet about these.
You can consider the data file that you want to load in Anaplan as a "table".
However, in order to identify a "unique key" of your file you need to answer this question:
What is the criteria to identify uniquely a row in the file ?
Is it already present a column that could identify uniquely a row in the file? (something like Invoice_ID ?)
If not, is there a combination of columns that could give as result a unique key of the file? ( example: Customer_ID + Product_ID + Date ?)
If you still do not have a positive answer after you concatenate all the columns of your file, then you need to return to the provider of the file and ask him/her to group (aggregate) the data based on the columns of the file in order to have a possible unique key of the file.
If you will have a "unique key" of the file then you will be able to use "Choose properties below to build the unique identifier for the imported items" option when import data in the numbered list without receiving any error.