Is it possible to create a table containing a column of company names and determine, for each company name, whether there are two or three duplicates of that name in the rows just above the current entry?
