Add gender data to an Excel file

The Excel add-in

Select a column of names, emails or usernames and press the button in the NameGender pane: gender, probability and, if you want them, the sample size, match type and a first, middle and last name split are written into the columns to the right. Repeated names are looked up once, and rows that already have a result can be skipped.

Three formulas work on a whole range at once: =NAMEGENDER.GENDER(A2:A500) returns the gender, =NAMEGENDER.SALUTATION(A2:A500, "de") a ready salutation such as “Sehr geehrte Frau Müller,” in the language you give, and =NAMEGENDER.CHECK(A2:A500) flags junk entries such as “asdf qwerty”. Each takes an optional country, as a code or a name. Results are kept for the session, so recalculating the sheet does not spend credits again.

The API key stays on your computer, not in the workbook, so sharing the file does not share the key.

Install the add-in

The add-in is waiting for its Microsoft AppSource listing. Until then it installs from its manifest file in a minute. In Excel on the web, open Home → Add-ins → More add-ins → Upload My Add-in and choose the file from namegender.com/excel/manifest.xml. In Excel for Mac, copy the file into ~/Library/Containers/com.microsoft.Excel/Data/Documents/wef/ and restart Excel.

Then open the NameGender pane, paste your API key once and test the connection. Every account has free daily credits, so you can try it on a real column before buying any.

Use file upload for one-time enrichment

Upload an XLSX or CSV file in the dashboard, preview the detected columns, select the given-name column and optionally map a country column. The cost is shown before processing begins, so uploading and inspecting a file does not commit credits by surprise.

The result is returned in the same spreadsheet format with derived fields appended. Keep probability, confidence, sample_size and source alongside gender. Those fields let you filter weak rows, audit a result and change your acceptance threshold without buying the same lookup again.

Repeated lookup keys are deduplicated during batch processing and joined back to their original rows. This matters in customer exports where a large row count often contains a much smaller set of unique first names.

Prepare the workbook without cleaning away evidence

Put field names in the first row and keep one person per row. A dedicated given-name column produces more reliable input than asking a spreadsheet formula to split every full name. If you have country data, use a separate column containing ISO two-letter codes.

Do not replace accented characters, Arabic text or Asian scripts with ASCII before upload. The matching pipeline performs its own script-aware normalisation, while manual conversion can merge distinct names or create a spelling that belongs to no source.

Empty cells, company names and ambiguous rows should remain in the file. They are part of the quality measurement. After processing, count unknown separately rather than deleting those rows and reporting accuracy only on the easy subset.

Use Power Query for repeatable refreshes

Power Query can call an HTTP JSON API and expand the response into columns. This is appropriate when the workbook is an internal model that refreshes on a controlled machine. It is less appropriate when the file will be broadly shared, because credentials embedded in a query can travel with the workbook.

Keep the API key in a managed parameter or credential store available only to the refresh owner. Send unique name-country pairs in bounded bulk requests, then merge the result table into the original query. Avoid a custom function that makes one web request for every worksheet row.

For scheduled production enrichment, move the same logic to a server-side job. Excel remains a good review and analysis surface, but it should not become the secret store or retry engine for a business-critical integration.

Questions

Related pages