Are Apps a game changer for women’s business training? A cautionary tale from Ethiopia

The views expressed in the Let’s Talk Development blog are solely those of the author(s).

Business training programs have long been a cornerstone of development policy, potentially serving as essential tools for fostering economic growth, enhancing entrepreneurial skills, and improving the overall business environment. Yet, the evidence on their effectiveness—especially for women—has been mixed. Women entrepreneurs often face unique challenges, such as time constraints due to unpaid care work and limited mobility, which can hinder their ability to benefit from traditional in-person training programs.
 

The Digital Promise for Women

Digital platforms can scale quickly, reduce costs, and offer flexibility that might better suit women’s schedules. But does this promise hold up in practice?

Our study conducted a Randomized Control Trial comparing  app-based training to  in-person training for 2,000 women entrepreneurs in Ethiopia.  Both training programs covered an identical business curriculum.  The in-person option was delivered through 11 bi-weekly sessions over six weeks, while the app-based version allowed for self-paced learning enhanced with interactive elements. Both programs required participants to have some prior business knowledge, covering fundamentals like marketing, accounting, and business planning.

The study analyzes not only the completion rate for the two modalities but also the cost-effectiveness of the interventions and the impact of the training on women’s businesses. The women involved were specifically chosen to maximize the feasibility of app-based training and provide a “proof of concept” for digital training for women in low-income settings. These participants were growth-oriented entrepreneurs with access to smartphones and the Internet, and they were open to engaging in digital training.
 

Results: A Cautionary Tale About Digital Training
 

• High Initial Participation: Both the app-based (80%) and in-person (76%) training programs saw strong initial take-up. For the app-based training, this was partly due to enumerators who encouraged and guided participants through the first module during in-person information sessions.
• Low Digital Completion: Despite high initial engagement, completion rates (completing at least 8 modules to earn a certificate) were significantly lower for the digital training (22%) compared to in-person training (71%). However, it is important to note that low completion rates are not unusual in the e-learning app industry.
• Ineffective Additional Efforts: Efforts to improve engagement, such as weekly reminders, cash-prize incentives, and offline app access, failed to improve completion rates. A networking feature, designed to foster peer interaction through chat rooms, also had poor results.  This was partially due to a government block on the most widely used chat platform, coupled with limited engagement by participants.
• Underwhelming Effect on Businesses: Despite differences in participation and completion rates, neither training modality had a significant effect on the students’ businesses survival (remaining open) or performance (sales and profits). The study also found no meaningful effects on business inputs, capital, practices, or knowledge gain when compared to a control group not being offered training.

^{Notes: The figure illustrates disparities in training completion between app-based and in-person modalities. For in-person training (right panel), 24 percent of individuals never attend a single session, while 71 percent attend at least eight sessions and earn a certificate. The noticeable increase at eight sessions indicates the binding incentive of the certificate, with most respondents continuing to complete nine to eleven sessions. In contrast, the distribution for app-based training participants (left panel) shows that 19 percent do not complete any app module, and around 50 percent complete just one to three modules. Approximately 25 percent of participants start all app modules, but only 15 percent complete all of them, and 22 percent earn the certificate.}
 

Cost of Digital vs In-Person Training

We also compare the costs of delivering digital vs in-person training. When looking only at the cost per student, the in-person program was approximately 84% more expensive than the digital alternative. However, the average cost per participant who completed the course was much higher for app-based training, considering its lower completion rates.

A key advantage of asynchronous training (such as the app-based program) is that it can be offered to more participants at almost zero direct marginal cost. Nevertheless, an additional experiment conducted within this study highlights that in-person demonstration sessions, where participants were helped to download and log in to the app, were crucial to initiate engagement. The requirement of these in-person sessions raises the overall marginal cost of implementing a digital program.
 

Lessons for Policymaking

Our study offers valuable lessons for the design and implementation of training programs aimed at supporting the growth and performance of women-led businesses:

• The promise of digital technology for delivering training should be approached with caution: Although the initial interest in the app-based training was high, most students dropped off, leading to low completion rates. This lack of follow-through happened despite targeting educated and motivated women entrepreneurs who had smartphones and internet access.
• App commitment is challenging: Keeping participants engaged in app-based training is extremely difficult. Post-experiment surveys suggest that even educated entrepreneurs lacked the digital literacy or confidence to solve basic technical issues, such as trouble logging back in. Stronger gamification features (scoring, competition, prizes) might make an app more engaging and encourage digital learning-by-doing if participants were sufficiently motivated to troubleshoot on their own. Moreover, the higher completion rates of the in-person training indicate that the social dimensions of attending sessions could serve as a commitment device to overcome self-control issues. Similar commitment devices could be designed to increase app take-up.
• Digital training is cheaper to scale but might not be as cost-effective as it first appears:  while digital training offers low-cost scalability, further costs may need to be factored into calculations. In-person onboarding was essential for take-up, raising the cost of delivering digital training.
• Business training alone may not be enough to drive results: The findings show that business training, including app-based programs, is not a game changer for improving the performance of women-led businesses. This suggests that to be impactful, training may need to be combined with other interventions that address the multiple constraints faced by women entrepreneurs.

^{This study contributes to our understanding of the constraints and policy options to facilitate jobs for women and support the growth of their businesses in low- and middle-income settings.  This work is part of the body of research underpinning the World Bank’s Center for Research on Women and Jobs, a research initiative led by the World Bank's Development Research Group in collaboration with the Bank's regional Gender Innovation Labs and other key partners to coordinate policy research efforts on women and jobs.}