The quantitative evaluation was designed as a cRCT, with two stages of randomisation: a village level randomisation where villages are assigned to a treatment group or to a control group, and a
household-level randomisation within treatment villages whereby households are either assigned
to receive the mNutrition content on just the mobile phone of the primary female or on the mobile
phones of both the primary female and the primary male. In villages that were assigned to the
treatment group, sampled households were offered access to the mNutrition content on their
mobile phone, free of charge, through a door-to-door in-person visit. In villages that were assigned
to the control group, no offer of access to the service was made. Though registration for the
mNutrition service was possible for all households regardless of treatment assignment, and there
was some Tanzania-wide advertising of the service through billboards and radio adverts, prebaseline
discussions with the organisation implementing the mNutrition service in Tanzania suggested that take-up of their existing service was low in the study region.
