Convert CSV, Excel, JSON and TSV without losing the structure

Converting data is not just changing a file extension. Each format can represent different information, so a conversion should begin with one question: what must survive in the output?

CSV and TSV are flat text formats

CSV and TSV represent rows and columns as text. The separator must be interpreted correctly and values containing separators, quotes or line breaks need proper escaping.

Encoding matters as well. UTF-8 is usually the safest choice for multilingual data, but older files may use other encodings.

Excel contains richer information

An Excel workbook can contain multiple worksheets, formulas, formatting and typed cells. A CSV output cannot preserve all of that.

When converting Excel to CSV, decide which sheet and which values matter. If formulas are involved, understand whether you need the formula itself or its calculated value.

Use the Excel, CSV and JSON Converter for common spreadsheet conversions.

CSV to JSON is usually straightforward

A table can be represented as an array of objects: headers become keys and each row becomes an object.

For example:

[
  { "id": "001", "name": "Ada" },
  { "id": "002", "name": "Grace" }
]

The CSV/JSON Converter handles this direct relationship.

JSON to CSV requires a flat schema

JSON becomes more difficult when objects contain nested objects or arrays. A table expects a predictable set of columns, so nested values may need to be flattened, serialized or omitted.

There is no universal lossless conversion from every JSON document to CSV.

Preserve identifiers as text when needed

A spreadsheet application may interpret 00123 as the number 123. That is harmless for a quantity but destructive for a product code or account identifier.

Check columns whose formatting carries meaning before conversion.

Dates and decimals depend on conventions

Dates such as 01/02/2026 and decimal values using commas can be interpreted differently by locale. A conversion pipeline should prefer unambiguous representations whenever possible.

ISO-style dates such as 2026-02-01 are easier to exchange between systems.

Conversion does not clean the data

If a CSV contains duplicates, converting it to JSON simply creates duplicate objects. If headers are inconsistent, they become inconsistent JSON keys.

Clean first when the data itself is the problem; convert when the representation is the problem.

Validate the output

After conversion, check:

• the number of records;
• expected columns or object keys;
• non-ASCII characters;
• identifiers with leading zeros;
• dates and decimal values;
• a few representative rows.

A successful conversion is one where the meaning and structure required by the destination are still intact.