Solving the mentorship gap in emerging markets

In many emerging economies, entrepreneurship is one of the few viable paths out of poverty and instability. But that promise comes with a paradox: In places where formal markets are not accessible for most people, who teaches entrepreneurs how to build and run a business in the first place? Frank Germann, a marketing professor at the University of Notre Dame, has studied this question throughout his research career. He previously found that mentorship plays a crucial role in the success of entrepreneurs in developing countries. 

Perhaps more surprisingly, he also discovered that marketing-specific mentorship is particularly effective, as it helps entrepreneurs understand how to differentiate their products or services in often amorphous, loosely defined markets. Now, Germann is adding AI to the project with a new study that was recently awarded a 2025 Poverty Research Package funding from the Notre Dame Poverty Initiative, which supports high-impact, poverty-related research projects led by Notre Dame faculty in multiple disciplines. Germann and co-principal investigator Father Arthur Ssembajja (MBA/MGA ’22), a research visitor at Notre Dame and a Ph.D. candidate in economics at the University of Tübingen (Germany), where Germann serves as his co-advisor, are leading a randomized controlled trial in Kampala, Uganda. The trial explores the potential of large language models to provide scalable, AI-driven mentorship for entrepreneurs in emerging markets. The effort is aimed at reducing the costs and difficulties involved with scaling mentorship programs by providing a virtual mentor installed on the entrepreneur’s phone. “We developed a custom LLM [large language model] specifically for emerging market entrepreneurs, to serve as a surrogate mentor,” said Germann. “They'll have this LLM at their fingertips and they can ask any question they want about how to improve their business. We have a particular focus on marketing, but it's also more broadly a business advisor tool.” 

The randomized controlled trial started in November. The entrepreneurs in the trial typically have small-scale enterprises, with a physical structure and several employees. Germann described the types of ventures as typical for emerging markets — “mom-and-pops” for the most part that form the confluence of daily life, including beauty parlors, furniture makers, retailers and electricians. The Notre Dame team trained the entrepreneurs on the tool and made sure it was installed on their phones correctly. They will track the group over the next year to evaluate whether those using the tool are doing better in terms of sales and profits. The researchers are also looking at outcome measures such as anxiety, stress and happiness. “If our tailored AI mentor proves effective, it could democratize access to high-quality business advice for millions of underserved entrepreneurs,” said Germann. “This is an example of how innovative technologies guided by Notre Dame values can expand opportunity, strengthen communities and empower people globally.”