I think I said somewhere before - it is not for reporting / BI. Where I have done data warehouses before I have left the data as is.

Although I would counter that and say that there will always be a case for checking data and reviewing if there is a cleansing / modifiying requirement. Although generally I would retain the original column in the table and add a new column with the modified data. This retaining the opriginal 'as is' and also the cleansed value.

The data in this case is used for comparisons and number crunching. Therefore on some fields there could be an advantage if the data is prepped first. Was just wondering if going to NULL was the sensible choice.

Cheers

M