I have a SQL script encoded in UTF-8 that contains insert statements with unicode strings that are inserting in a NVARCHAR field. When I run this script, the data inserted in the table is malformed.

The core function of this syntax is to inform the database that the string should be processed using Unicode encoding (such as UTF-16), primarily for matching fields of Unicode types like NVARCHAR.

I've been struggling with utf-8 vs utf-8-sig; it seems like we want to use -sig for reading but not-sig for writing. The BOM should be completely unnecessary in utf-8 but some software adds it anyway, ...