Excel 2003

I use Excel for a mailing database. Because there are so many "records"
(approx 6,000) it is likely that there will be different people who share a
name. I don't want to filter out John Hanks at Company XYZ just because the
advanced filter finds a John Hanks at ABC Company.

So question 1 is how does Excel determine which will be filtered out--in
what order does it work. I'd like to understand which John Hanks would be
determined to be the duplicate.

Question 2 is how do I tell Excel that I only want it to look at the company
name column when determining whether the record is a duplicate.

Thanks so much for any help!