>>Hi,
>>
>>I will be discussing (later this morning) with a company who is supposed to export the data from their SQL Server database with the goal that I will import it in my SQL Server database. Initially the agreement was that they will provide me with the CSV format files. My questions is, is this "the best" or "sufficient" that I can then import the data into my database? Or should I ask for another format? (the SQL Server backup is out of the question because their SQL Server database has their other customers' data).
>>TIA
>
>On this case I suggest to uses a dummy sql database where the company store the data.
>From it you can import that data.

Ah yes, how did I fail to see that? The CSV is a lowest common denominator when the source database is something in its own format, like Clarion or Magic or bTrieve or any proprietary format and the old app has an export utility. For anything anywhere close to standard (i.e. that has an ODBC driver), take the original tables. I've seen so many weird things when an app exports to excel... one of date columns is in DDMMYYYY, the next one is in MMDDYY, because one of them was a c(6) and Excel treats them differently.
I go for a backup of the source database whenever possible, no intermediaries. I even have a VM with the latest SQL, for that one case a year when that backup was made by such a version and can't be restored into an older SQL.