The impact of the COVID-19 pandemic on the higher education sector has been overwhelming, with emergency responses that have affected decision-making processes. Yet, our understanding of higher education instructors' perspectives regarding the process of data-driven decisions, especially in times of emergency, is still limited. We aimed at characterizing the types of data-driven decisions that higher education instructors have made in their courses. This was done while asking the instructors to reflect upon a face-to-face (F2F) course that was suddenly shifted to emergency remote teaching (ERT), due to the COVID-19 pandemic outbreak. Taking a qualitative approach, data were collected via an open-ended online questionnaire distributed among 109 higher education instructors from different countries. The findings suggest that the instructors mentioned a wider range of data sources, and a wider range of data-driven decisions while referring to the ERT mode, compared with their F2F instruction. In F2F teaching, the instructors mostly provided students with real-time educational assistance. In ERT, the instructors mostly adjusted the course requirements, promoted collaboration among students, and offered them social and emotional support.