Child acute malnutrition remains a persistent challenge in dryland Kenya. This thesis explores why some households are at higher risk than others using a data-driven approach. Paper I highlights the feasibility of smartphone-based surveys for near-real-time, community-led nutrition monitoring. Paper II leverages panel data to identify and analyze variations in household personal network structures and their association with acute malnutrition risk. Paper III underscores the value of dynamic risk assessments for the timely and thorough treatment of acute malnutrition. By addressing evidentiary and analytical gaps, this thesis provides fresh insights into the basic drivers of acute malnutrition, contributing to more resilient local food systems in dryland regions and measurable progress toward nutrition-related Sustainable Development Goals.