All interviews and FDGs were recorded, transcribed verbatim, translated from Shona or Ndebele to English, and transferred to NVivo14.23.3 (QSR International) (61) software version for analysis. In the field, notes were taken, and daily interview summaries were subsequently written to aid interpretation. Guided by the SULF framework, we conducted a hybrid deductive and inductive thematic analysis of data. The deductive aspect of the analysis involved using codes developed a priori from SULF. Among these deductively produced high-level codes, lower-level codes were inductively generated using content analysis. Pattern coding was used to identify patterns across and within the data sources. This allowed the condensation of data into fewer relevant analytical concepts. For validity, pattern coding was conducted by three experienced researchers who reviewed all transcripts and identified the descriptive codes through consensus. This analytical approach helped maintain a focus on the holistic livelihood impact of the COVID-19 pandemic on households.