Martin Burns noted: >>ETL lets you consolidate it all in a common format, >>with cleansing and de-duping on the way, usually to >>a datamart of some kind. Hi Martin, Spot on! Just wanted to point out that y'all made that sound a bit breezy. In practice, ETL can be one of the 'heavy lifting' areas where all manner of fugly fuzzy-logic matching, multigrain consolidations, day-over-day differencing, and algorithmic derivations get added to your dataset. On the plus side there is that great look you get when you patiently explain that Informatica, while very nice, doesn't actually do *everything* that needs to be done for ET&L. Cheers, RonL. What you is is what you once were when!