Microsoft ExcelFormats

Anonymize Excel without losing the ability to analyse it

When you anonymize Excel workbooks, the spreadsheet usually still needs to do its job: count, group, filter or serve as an example. This guide walks through the manual methods in Excel, a column-by-column method table, the places personal data hides – and the automatic route with Diskretio, which replaces each person with the same placeholder across all cells and sheets.

Microsoft ExcelClient_List_Q3.xlsx
StartEinfügenFormelnDatenAnsicht
A2fxJohn Smith
ABCDEF
1ClientDate of birthIBANEmailPhoneFees
2John SmithMarch 4, 1970GB82 WEST 1234 5698 7654 32j.smith@mail.com(212) 555-1234£4,250.00
3Jane DoeMarch 12, 1985GB33 BUKB 2020 1555 5555 55jane.doe@outlook.com(415) 555-0188£1,980.00
4Acme Construction Ltd–GB94 BARC 1020 1530 0934 59info@acme-build.co.uk(646) 555-0142£12,400.00
5Thomas BakerNov 21, 1962GB29 NWBK 6016 1331 9268 19t.baker@yahoo.com(312) 555-0199£760.00
6Emily ClarkeAug 7, 1979GB60 LOYD 3096 3412 3456 78emily.clarke@proton.me(617) 555-0123£2,315.00
7      
8      
ClientsFeesDeadlines

Your formats and the AI tools you work with afterwards

  • Word
  • Excel
  • PowerPoint
  • Outlook (emails)
  • PDF
  • Scans & photos
  • TXT Text & CSV
  • ChatGPT
  • Claude
  • Microsoft Copilot
  • Gemini
  • Mistral
  • DeepL
  • Perplexity

Key takeaways

  • Anonymized means people can no longer be identified. Swapping names for codes while keeping a key is pseudonymization – the data is still personal data.
  • Manually, use Find & Replace (Ctrl + H), a mapping table with XLOOKUP, or RANDBETWEEN values pasted as values.
  • Personal data also hides in hidden sheets, comments, headers, sheet names, pivot caches, charts and file properties.
  • Consistency matters: the same person needs the same placeholder on every sheet, or counts and pivots break.
  • Rare combinations such as department + age + start year can still identify someone – generalise those columns.

Typical Excel files that get anonymized

  • Customer and member lists for analysis or service providers
  • Staff lists, rotas and salary overviews
  • CRM, ERP or ticket system exports for debugging
  • Survey results with free-text answers
  • Test data for software and databases
  • Sample files for training and documentation
  • Spreadsheets you want to analyse with ChatGPT, Claude or Copilot

How it works

  1. 1

    Upload the XLSX

    Cells, sheet names, comments, headers, charts and pivot caches are read – hidden sheets included.

  2. 2

    Choose placeholders or fake data

    [PERSON_1] in every cell for the same person, or realistic substitute names from the Pro plan.

  3. 3

    Keep calculating

    Formulas, references and formatting stay, only the personal values are replaced. Refresh pivot tables once.

Anonymization vs. pseudonymization: what's the difference?

Under the GDPR, pseudonymization (Art. 4(5)) means replacing names with identifiers such as “C-0042” while the key still exists somewhere, for example in a separate lookup sheet. That data remains personal data for whoever holds the key. Truly anonymous data no longer relates to an identifiable person, and Recital 26 puts it outside the GDPR.

In 2025 the Court of Justice of the EU added nuance: pseudonymized data is not necessarily personal data for a recipient who has no realistic means of re-identifying people (CJEU, 4 September 2025, C-413/23 P). For you as the sender, it stays personal as long as you hold the key.

In practice: a mapping table lets you trace results back later, but the table itself must be protected. Replacing without a key means no way back – and gets you closer to genuine anonymity.

Whether a dataset is legally anonymous depends on every column that remains. General information, not legal advice.

How do you anonymize data in Excel manually?

For small sheets, Excel's built-in tools are enough. Always work on a copy, never on the master file. Four methods are common:

  • Find & Replace (Ctrl + H): replace each name with “Person 1”, “Person 2” and so on. Fine for a handful of names, error-prone with spelling variants
  • Mapping table with XLOOKUP: copy the name column to a new sheet, use Data › Remove Duplicates, assign sequential IDs next to it, and pull the ID into the main sheet with =XLOOKUP(A2,Map!A:A,Map!B:B) (VLOOKUP in older versions)
  • Random values: generate substitute numbers with =RANDBETWEEN(10000,99999). The function recalculates on every change, so copy the column and use Paste Special › Values to freeze it
  • Delete or generalise: drop columns nobody needs; turn dates of birth into age bands and full postcodes into the first characters

After replacing, delete the original column rather than hiding it, and convert any formulas that reference names into values. Otherwise the real names are still in the file.

Where does personal data hide in an Excel file?

Visible cells are only part of the picture. Names often survive in places a manual clean-up misses:

  • Hidden sheets, rows and columns
  • Comments and notes, including their author names
  • Headers and footers such as “Prepared by …”
  • Sheet names like “Smith John” and formulas that reference them
  • Pivot caches, which keep the raw data even after the source changes
  • Chart titles, axis labels and cached chart values
  • File properties: author, last modified by, company
  • External links with file paths containing user names, and embedded images

Excel can help you find some of this: File › Info › Check for Issues › Inspect Document lists comments, properties, hidden rows and hidden sheets. It does not check what is inside your cells.

Which method for which column? Data types at a glance

Not every column needs the same treatment – it depends on what the spreadsheet must still do afterwards:

Data typeExampleRecommended methodWith Diskretio
Person namesJohn SmithConsistent placeholder or substitute name[PERSON_1] or fake name (Pro)
CompaniesSmith Builders LtdPlaceholder[ORG_1]
Email, phonej.smith@…ReplaceDetected automatically
IBAN, card numbersGB29 NWBK …Replace or drop the columnDetected, even when stored as a number
Tax IDs, social security / national IDsQQ 12 34 56 CReplaceDetected automatically
Date of birth14/03/1987Generalise to year or age bandDates of birth are replaced
Address, city5 High St, LeedsReplace or generalise to regionAddresses and locations detected
Amounts, quantities1,250.00Usually keepKept
Free text (comments, surveys)“Ms K. was rude”Review text, replace namesNames in running text detected
Rare attributesDepartment + age + start yearGeneralise or removeNot automatic – review yourself

Why consistency is everything in spreadsheets

In a spreadsheet the same person often appears dozens of times, across several sheets and spellings. Deleting or randomly replacing each occurrence would make counts, pivot tables and lookups worthless. Diskretio therefore plans replacements for the whole workbook at once: “Hans Müller” becomes [PERSON_1] everywhere, on sheet 1 just as on sheet 7.

That mapping only applies within one job and is never stored. Two separately anonymized files therefore cannot be linked through their placeholders. If you need references between tables, put them into one workbook as separate sheets before you anonymize.

Anonymize Excel automatically with Diskretio

Diskretio edits the XLSX file in place: cell text, numbers, formulas, sheet names, headers and footers, text boxes, charts and pivot caches are anonymized; comments and external links are emptied; file properties are cleaned and embedded images are replaced with grey areas. If sheet names change, references in formulas, defined names and charts are updated automatically. Formatting, column widths and formulas stay intact.

Choose placeholders like [PERSON_1] (best for AI and analysis), realistic fake data from the Pro plan (for test data and training) or classic black redaction. Terms that should stay, such as product or site names, go into an allow list from the Pro plan; names that must always be replaced go into a deny list.

There are limits: single words without sentence context, such as a column containing only surnames, are harder to classify than running text, so review such columns in the result. Save legacy .xls files as .xlsx first; CSV files are supported directly.

Is the spreadsheet really anonymous afterwards?

Not automatically. Without names, a combination of rare attributes can still identify someone: the only female engineer over 60 at the Leeds site is identifiable even as [PERSON_17]. Before sharing, check whether individual rows stand out by department, age, location or start date, and generalise those columns.

A practical rule of thumb: every combination of the remaining attributes should apply to several people. Where it doesn't, reduce detail – birth decade instead of birth year, region instead of town.

FAQ

Questions. Answered briefly.

Try it with your own document

10 pages per month are free. No subscription, no credit card.