How to anonymize an Excel file before you share it

A practical guide for spreadsheets with personal data

Sending a spreadsheet to a developer, a supplier or an analyst, or pasting it into ChatGPT or another AI assistant, sends every row in it. If those rows describe real people, remove or replace the personal data first. Here is how to do it without uploading the file anywhere.

1. Find the columns that identify people

Start with the obvious ones, which identify a person on their own:

• names, email addresses and phone numbers;
• national ID numbers, such as a Finnish henkilötunnus or a Swedish personnummer;
• IBANs, payment card numbers and customer or account numbers;
• street addresses.

Then look for columns that identify people in combination: a date of birth, a postal code, a rare job title. Each is harmless alone, but together they can point to one person. Free-text columns such as notes or comments often hide names and phone numbers too.

2. Choose what happens to each column

Not every column needs the same treatment. Pick per column:

• Fake it when the data still has to look real, for testing or a demo. A fake IBAN or ID number keeps the format and passes the checksum.
• Mask it when people need to recognise a value without seeing it, like ************1234.
• Redact or clear it when nobody needs it.
• Hash it when you still need to join files on it: the same value always gives the same hash, with your own secret key.
• Generalize it when the rough value is enough: a birth year instead of a date, the start of a postal code instead of all of it.
• Keep the columns that contain no personal data.

3a. Anonymize it in your browser

1. Open the Scrubsheet browser tool and drop in your .xlsx, .xls, .ods or .csv file. It is read by the page itself: nothing is uploaded, and the page cannot send data anywhere.
2. Check the suggested type and method for each column. Detection looks at column names and sample values, and national ID numbers are only recognised when their checksum is valid. For a notes column, you can anonymize just the personal data inside the text.
3. Compare the before and after preview.
4. Download the anonymized copy as .xlsx or .csv.

It is free, with no row limit and no sign-up.

3b. Or anonymize it inside Excel

If the data should not leave the workbook at all, or you need formulas and formatting kept, use the Scrubsheet add-in for Excel. Open it from Home › Anonymize, pick the table or range, check the columns, and it writes an anonymized copy of the sheet next to the original.

4. Check the result before you send it

• Scroll through the anonymized file. Detection helps, but it cannot guarantee it found every piece of personal data, especially in free text.
• Remove columns the recipient does not need. The safest data is data you do not send.
• If you created a mapping file (original value next to its replacement), keep it somewhere secure and never send it with the anonymized file: it undoes the anonymization.

Replacing direct identifiers does not always make data anonymous in the legal sense of the GDPR; that depends on what remains and who receives it. See the FAQ for more.