If you use Ali's approach to create the unique ID's then you could then use INDEX/MATCH to pull the data based on those ID's?