Data Cleaning
Full instructions and common questions
Data Cleaning
Clean CSV, JSON and Excel tables in your browser. Trim fields, normalize numbers and dates, replace text and deduplicate by chosen keys. Review before continuing, export CSV, Excel or JSON, and reuse cleaning rules.
How to use
- Import CSV, TSV, JSON or XLSX and choose a worksheet, or paste and read data. Start with the duplicate-ID sample if needed.
- Add rules and choose deduplication fields; selecting none retains every row. Rules run in order and duplicate keys keep their first record.
- Run cleaning and review row counts and values. Choose “Continue with this result” before subsequent cleaning, summary or splitting uses it.
- Download CSV, Excel or JSON. Save rules for later use and confirm field mappings when restoring them. Reimport source data after refreshing.
Try an example
Sample A starts with four rows and configured trimming and ID deduplication. Cleaning should leave three rows. Confirm the result and group by channel: community has two rows totaling 400, email has one totaling 200.
Limits and privacy
Up to 50 files and 50 MB in total, 100,000 rows, 300 fields and 2 million cells. Numeric normalization changes text representation and is unsuitable for IDs with leading zeros. Dates require a recognizable year-month-day format. Selected keys define duplicates; real people or business identities are not inferred.
Common questions
Why does the next step still show the original data?
Review the result and choose “Continue with this result”. Generated output is not automatically adopted as the next input.
Does saving rules save my table?
Rules store fields, operations and deduplication settings, not source tables. Reimport data and check mappings when reusing them.