SQL Server 2016 provides built-in support for importing, exporting, parsing and querying JSON documents. However, JSON support is not the same as the native JSON type, and it is implemented using simple text with types like NVARCHAR.

Traditionally, NoSQL systems, which based on JSON documents (such as DocumentDB or MongoDB) are designed to store the data more effectively (using binary format like BJON and not as simple text) and are designed to increase performance (for example using built-in indexes' systems). Parsing text "on the fly" in relational database (like SQL Server), especially with long text types like NVARCHAR(MAX) and large number of rows, can be challenging!

This zip file includs the presentation file (as pdf) and the code files which were used in Ronen Ariely Lecture for the Israeli SQL Server User Group. In this lecture, we gave a glance on built-in JSON support functions in SQL Server 2016, and we focused on tips and tricks to improve performance using JSON on SQL Server 2016 and above.