Women in Agribusiness Leadership Network: Training Businesswomen to Mentor Businesswomen Across Ethiopia

Ethiopia's USAID-funded Women in Agribusiness Leadership Network (WALN), implemented by ACDI/VOCA, set out to test whether a cascading, two-stage model — train experienced businesswomen as mentors, then have them coach less-experienced women in their own social and business networks — could improve microenterprise performance more cheaply than direct expert-led training alone.

To test whether coaching passed down through informal, unpaid peer mentors can improve microenterprise performance as effectively as, but more cheaply than, direct expert-led training.

The World Bank's Africa Gender Innovation Lab evaluated the programme as a stratified randomized controlled trial across 96 woredas in five regions (Tigray, Amhara, Oromia, SNNPR and Addis Ababa). From a baseline in March-August 2014, 197 businesswomen (99 treatment, 98 control) received formal training in marketing, record-keeping, financial planning and stock control between August and December 2014. From January to July 2015, the treated businesswomen mentored 295 mentees (with 294 more in a spillover group and 539 controls) drawn from their own networks.

Three post-programme survey rounds (through April 2018) show the direct training effect on mentors was substantial: a profits-and-sales index up 0.21 standard deviations, winsorized profits up 80% and winsorized revenue up 62%, alongside a 0.13 SD gain in business practices. Among mentees, business-practice adoption rose 0.27 standard deviations (marketing practices +0.27 SD, record-keeping +0.18 SD), but the effect on mentees' own profits and revenue (+0.065 SD) was not statistically significant.

The study is a useful cautionary case for cascading models: coaching passed down through informal, unpaid peer mentors changed what mentees did in their businesses, but did not reliably translate into higher mentee earnings within the observation window — a gap the authors attribute to the more diluted, less-customized nature of second-hand mentoring relative to the original expert-led training.