How to flatten nested JSON to CSV

Many APIs return nested JSON structures — an order object might contain a customer sub-object and a shipping address sub-object. Spreadsheets and SQL tables are flat. This guide shows how to automatically flatten nested JSON using dot-notation column names so every property ends up in its own column.

Step-by-step guide

1. Identify the nesting depth: Look at your JSON structure. If objects contain sub-objects (but not deeply nested arrays), the flatten option handles them automatically. Deeply nested or array-valued fields are stringified into single cells.
2. Enable flatten and convert: Paste the JSON array with flatten enabled. The converter uses dot-notation: {address: {city: 'London'}} becomes a column named 'address.city'. Multiple levels of nesting are handled recursively.
3. Verify the column names: The 'columns' field in the response lists all generated column names. Check that dot-notation names are meaningful for your downstream tool before importing.

Frequently asked questions

What happens to array-valued fields during flattening?

Arrays (e.g. tags: ['a', 'b']) are not recursively expanded — they are JSON-stringified into a single cell ('["a","b"]'). Expanding arrays into multiple rows requires a different transformation that is outside the scope of a flat table.

Can I import dot-notation column names into a database?

Most databases require column names without dots. Rename the columns after import, or set flatten to false and handle the nesting in your application code instead.