Skip to main content
CSV First Aid

CSV date normalizer — mixed dates to ISO 8601

One column with 01/03/2024, 2024-03-01, and 'March 1, 2024' mixed together is a column that can't be sorted. We rewrite everything to YYYY-MM-DD — unambiguous, sortable, safe for every database.

Dates normalized

Before

15/03/2024
03-15-2024

After

2024-03-15
2024-03-15

Drop your CSV file here

or click to browse

The "date format" fix will be auto-detected.

What is this and why does it matter?

Mixed date formats are one of the most common data quality issues. They happen when data comes from multiple sources, manual entry mixes with automated exports, or locale settings change between users. Is 01/02/2024 January 2nd or February 1st? Without context, it's ambiguous.

CSV First Aid analyzes each column for date-like values. It detects common formats: DD/MM/YYYY, MM/DD/YYYY, YYYY-MM-DD, DD-Mon-YYYY, and more. When a column has mixed formats, the tool normalizes all values to ISO 8601 (YYYY-MM-DD), the unambiguous international standard.

Ambiguous dates (like 01/02/2024) are resolved using the column's dominant format pattern. If 90% of dates in the column use DD/MM/YYYY, the ambiguous ones are assumed to follow the same convention.

How it works

  1. 1Drop your CSV. The detector scans each column for date-like patterns.
  2. 2Mixed-format columns are flagged. The diagnosis shows how many dates need normalization.
  3. 3Apply → all dates become YYYY-MM-DD. Download.

FAQ

What if a date is genuinely ambiguous (like 01/02/2024)?

The tool uses the majority format in the column to disambiguate. If most dates are clearly DD/MM, ambiguous ones are parsed as DD/MM too. Edge cases are flagged as 'partial' in the fix report.

Can I choose a different output format?

Currently the output is ISO 8601 (YYYY-MM-DD). This is the safest format for data interchange since it's unambiguous and sorts correctly. Custom output formats are planned.

What about timestamps (dates with times)?

Date-time values like '2024-01-15 14:30:00' are preserved as-is. Only the date portion is normalized when the pattern is purely a date.

Related tools