This paper evaluates a method of generating a unique dataset that has been underused—a Freedom of Information (FOI) request. The UK’s FOI Act came into force in 2005, allowing the public to make requests of publicly held data. We set out to understand the determinants of the quality responses to FOI requests. We also explain how requests can be made and provide suggestions to construct effective data-driven requests. We applied for data on hate crime from all police forces and on uptake of maternity leave from all universities. We find that observable characteristics of the local area, police force and universities neither determine whether the request was fulfilled, nor the speed of response, suggesting that the data from these FOI requests are representative of the institutions from which they were requested.