Kabakoo’s follow-up data suggests that participants show stronger growth mindset, greater continued skill-building, higher labor-market attachment, improved income outcomes, high program satisfaction, and promising spillovers into their wider communities. These findings are based on linked baseline and follow-up data, using control groups and baseline-adjusted models.

Kabakoo’s evidence base spans a registered population of 40,089 learners across three tiers of engagement. Treatment effects in this brief are directly identified for the roughly 5,000 learners who completed structured cohort programs; spillover analyses provide evidence that effects extend to the roughly 35,000 learners reached through active platform use or community exposure.

<aside> 💡

Summary of findings

• Treated participants show a 0.14–0.25 SD baseline-adjusted gain in growth mindset relative to clean controls, reaching conventional significance for the intensive treatment group.
• At the 32-month follow-up, the treatment group shows a clear upward shift in income distribution, with a substantially larger share in the 250k+ FCFA bracket and a smaller share in the lowest earning bracket.
• The Kabakoo Behavioral Index gap (treatment vs clean control) widens from +0.30 at the first follow-up to +0.33 behaviors at 32 months, with the strongest contributions from employee earnings and earnings from services.
• Net Promoter Score rises from 61 (n=358) in early waves to 77 (n=150) in later waves at 32 months.
• Direct ROI ranges from 3.0x to 5.0x across analysis windows and self-report haircuts. Pooled-spillover sensitivity yields 5.1x to 8.4x under current cost accounting.
• Positive spillovers are observed among non-participants exposed through community ties or platform use, with the clearest persistent channels in borrowing networks and short-run financial resilience. </aside>

This report evaluates the impact of Kabakoo Academies’ upskilling program using a longitudinal panel design with baseline measurement and two follow-up waves (12–15 months and up to 32 months post-baseline). Identification relies on a recruit-and-delay control strategy that constructs a clean counterfactual from screened community members with no Kabakoo exposure, with progressive tightening of the comparison group as a robustness check.

The evaluation framework was developed in collaboration with the Chair of Personnel Economics at Paderborn University (Prof. Dr. Martin Schneider), drawing on the randomized control design discussed in “Islam, S., Herdejürgen, E., Kemayou, Y., Kabst, R., & Schneider, M. (2023). From Obstacle to Opportunity How Growth Mindsets Fosters Employability & Entrepreneurship in Mali. In Academy of Management Proceedings (Vol. 2023, No. 1, p. 19425)”. Dr. Jamie Walsh (Senior Behavioral Scientist, The Agency Fund) conducted several analyses of Kabakoo data. The Kabakoo Behavioral Index discussed below was developed in conversation with Dr. Richard Sedlmayr and Dr. Catherine Thomas (Assistant Professor of Psychology and Organizational Studies, University of Michigan).

The current analysis and reporting package were built by Kabakoo staff. A full replication trail is available by contacting [email protected].

Data overview and methods

Treated participants show measurable gains in growth mindset.

Stronger labor market attachment and higher income

Treated participants invest more in their professional capacity.

Kabakoo Behavioral Index: More positive economic and social behaviors

Learner satisfaction rises over time.

Spillover analysis suggests broader community benefits.

ROI remains strong under conservative assumptions.

Limitations