Globally about 800 million people live without electricity at home, over two thirds of which are in sub-Saharan Africa. Ending energy poverty is a key development priority because energy plays an enabling role for human wellbeing and economic activities. Planning electricity access infrastructure and allocating resources efficiently requires a careful assessment of the diverse energy needs across space, time, and sectors. However, because of data scarcity, most country or regional-scale electrification planning studies have been based on top-down electricity demand targets. Yet, poorly representing the heterogeneity in the electricity demand can lead to inappropriate energy planning, inaccurate energy system sizing, and misleading cost assessments. Here we introduce M-LED, Multi-sectoral Latent Electricity Demand, a geospatial data processing platform to estimate electricity demand in communities that live in energy poverty. The key novelties of the platform are the multi-sectoral, bottom-up, time-explicit demand evaluation and the assessment of water-energy-agriculture-development interlinkages. We apply the methodology to the country-study of Kenya. Our findings suggest that a bottom-up approach to evaluating energy needs across space, time, and sectors is likely to improve the reliability and accuracy of supply-side electrification modelling and therefore of electrification planning and policy.