Lack of access to electricity is a major barrier to social and economic development in East Africa. Distributed systems like microgrids are gaining attention as a cost-effective way to extend electricity access to remote rural areas. Load characteristics are important for microgrid power generation system design and financial modeling. This paper characterizes the load profile characteristics of 11 microgrids that PowerGen Renewable Energy, a leading microgrid developer in East Africa, installed between June 2014 and October 2016. Using these empirical load data, we quantify a day-to-day variability factor and a time-step variability factor. These two factors could serve as inputs to HOMER to create synthetic load profiles for use in microgrid simulations. This information is important for practitioners and academics alike as microgrid models like HOMER are widely used to analyze the performance of microgrids, yet there is a lack of empirical data that can be used in these types of models.