Crowd geocoding was used to geocode over 107,000 records from USAID's Development Credit Authority in under 16 hours with 145 volunteers. This increased the number of records with administrative level 1 (Admin1) codes from around 67,000 to around 74,000. The crowdsourced data was found to be 85% accurate, which is in line with current geocoding standards. The process demonstrated how crowdsourcing can be an effective no-cost solution for geocoding large datasets and engaging the public to support development work.