Help requested with data/table modeling

I suggest you start building a seperate Employee Dimension table, or set of tables. Table1: Employee ID and Employee Name, where ID is the unique key. Table2: Employee Name variations with [Proper Name] as one column and [Name as someone typed into the system because there is obviously no validation going on] as the second.

Table 1:

ID       Name

==     ====

1234    John Smith

Table 2:

Name                other names

====                =========

John Smith        John Q Smith

John Smith        Smith, John

John Smith       Smith, John Q

Then after you get frustrated with having to manage this list as you continually find more variations, go to the people that actually enter the data and tell them to adopt some standards when adding people's names.