I tried to gauge what percentage of cryptocurrency developers are female

Cryptocurrency likes to talk about decentralization, but its developer base is anything but diverse. In April 2019, journalist Corin Faife set out to measure the gender gap in crypto development by analyzing the commit history of the 100 largest cryptocurrency projects by market capitalization.

How it worked

Faife used Messari's OnChainFX dashboard to identify the top 100 projects, then pulled commit data from every source repository belonging to each project's GitHub organization via the GitHub API. Forked repos were excluded to avoid double-counting. The result was a dataset of 1,026,804 code commits across 100 projects.

Each committer's name was run through Genderize.io to classify it as male, female, or unknown. Faife was upfront about the limitations: gender is not binary, many developers use pseudonyms, and the tool can misclassify non-Western names.

The findings

Of the roughly one million commits, 67.3% came from male-identified names, 4.64% from female-identified names, and 28% were unclassifiable. By individual contributor count, the picture was nearly identical: 4.75% of unique developers were classified as female.

Male-named developers also contributed more per project — averaging 13 more commits than their female-named counterparts.

Some projects appeared to have higher female representation on paper — Bytom, VeChain, Neo, and Theta Token among them — but Faie found these were largely artifacts of Genderize.io misclassifying East Asian names rather than reflecting genuine diversity.

The coverage

The findings were picked up by CoinDesk, Decrypt, and CCN, reaching a broad crypto audience. The 4.64% figure became a widely cited benchmark for the industry's gender problem — concrete enough to cut through the usual vague claims about tech's diversity deficit. It also served as a useful case study in both the power and the limitations of name-based gender inference when applied to a global, pseudonym-heavy developer community.