Mappings use exact CSV header names
With name=product_name, only a header exactly equal to name changes. The CSV header itself is not automatically trimmed or case-normalized, so Name and name are different names.
Free web tool
Rename CSV headers in bulk using a mapping or a replacement header list.
Rename several CSV headers at once, such as name → product_name and price → unit_price. The data cells are left untouched; only the header row is mapped.
old=new.With name=product_name, only a header exactly equal to name changes. The CSV header itself is not automatically trimmed or case-normalized, so Name and name are different names.
“Mappings” is the number of valid mapping lines read from the box. “Renamed columns” is how many headers actually changed in the CSV. If those numbers differ, a typo or a mapping for a column that is not present is a likely reason.
Only the first header row is rewritten. Cell contents, row order, and column order stay the same. This is useful when a receiving system requires different field names but the data itself is already correct.
This tool does not validate a schema. If you map both first_name and last_name to name, the result can contain two columns named name. Check the header before sending the file into a system that requires unique names.
Lines without =, or with nothing on the left side, are not used as mappings. Leading and trailing spaces around the old/new names in the mapping box are trimmed. For a large rename set, the renamed-column count is the fastest first check that your map hit what you expected.