The JSON to CSV Converter expects a non-empty array of objects. CSV has one cell at each row-and-column intersection; JSON can contain arrays and objects inside values. Decide how to represent those structures before exporting.
Reproduce a quoting test
Try an object whose name field contains A, B, whose note includes quotation marks, and whose identifier is the string 0012. A comma-containing field needs quotes in comma-delimited CSV. Quotes inside a quoted field are doubled. The identifier must remain text when imported into a spreadsheet.
The converter uses the union of object keys as headers. A missing key or null produces an empty cell. Nested objects and arrays are serialized as JSON text inside a cell, not expanded into rows or columns. Flatten deliberately if your analysis needs separate columns.
Import instead of double-clicking
Use the spreadsheet’s text import controls to choose delimiter and encoding. Import identifiers as text so leading zeroes survive and long numbers are not rounded. Cells beginning with formula characters are prefixed with an apostrophe for spreadsheet safety; this intentionally changes the raw cell value.
Verify the output
Compare row counts, headers, a comma-containing value, a newline and a missing field. Keep the JSON alongside the CSV. CSV cannot preserve every JSON type without conventions, so a successful download does not prove semantic equivalence.
Quoting conventions are described in RFC 4180. Spreadsheet behavior can still vary by application.