The costing is a very valid argument. Feature for feature you could make Oracle match SQL (or vice versa) to the point where the business wouldn't really care as long as they got their reports. The simple reality of it all is that the SQL option will be cheaper - and I have seen clients implement a SQL DW + a dedicated resource for the same price it would have cost them to take the Oracle route. In this period of IT Budget contraints (yada yada) price counts for an awful lot...