Baseline survey for a water and sanitation program
An NGO launching a WASH program across several woredas had a 60-question paper baseline. Previous paper surveys had produced inconsistent household IDs, missing responses, and weeks of manual data entry.
We rebuilt the instrument as an advanced KoboToolbox form: automatic household ID generation, GPS capture at each interview, constraints on ages and quantities, and cascading location selects so enumerators could not misfile a household.
Data entry disappeared as a task. Supervisors reviewed submissions the same evening instead of after fieldwork ended, and the final dataset went to the analyst without a cleaning backlog.