Apollo Agriculture: using data science to scale up affordable farm credit for smallholder farmers in Kenya

Apollo Agriculture uses an innovative approach aimed at scale up provision of affordable farm credit to smallholder farmers so as to increase the number of smallholders with higher yields and incomes. Apollo’s innovation is the use of mobile technology, satellite data and advanced statistical methodologies to reduce the cost of providing the inputs and by predicting farmer credit outcomes and assessing credit risk, increasing the availability and reducing the cost of working capital that is needed to purchase the inputs.

Provision of loans to purchase high quality inputs and insurance and recovery of the loans once produce is harvested and so has become an established means of helping improve the yields and incomes of smallholder farmers in Africa. Organisations like One Acre Fund (OAF) have demonstrated success in both achieving high yields and recycling funds to fund inputs in subsequent years. However there are shortcomings with this approach: (i) ongoing grant funding is needed to sustain operations at a given level because the costs incurred providing agronomic advice and loan default costs cannot be recovered in full from low income smallholder farmers; and (ii) consequently programmes cannot be scaled up to provide comparable benefits to many more smallholders without having to access increasing and permanent grant funding.

Apollo seeks to address the shortcomings in various ways: by procuring inputs in bulk directly from manufacturers; developing and deploying lower cost means of processing loans and providing basic farmer training; and assessing credit risk by evaluating satellite data and remote farmer performance monitoring across the population and over successive seasons. Working capital will be raised by Apollo and on-led to registered small farmers with the expectation that wholesale finance and improved credit risk assessment will increase the availability and reduce the cost of working capital available to smallholder farmers.

In 2017 EfD approved a grant to fund the costs of a data scientist to build credit risk models based on satellite and on-farm data in order to evaluate the effectiveness of the approach in predicting credit risk. This was very much R&D funding at a very early stage in the development of Apollo. Further details of the approach proposed is set out in Apollo’s application document.

Subsequent update discussions with Apollo have taken place since the initial funding. Interesting insights have led Apollo to evolve its approach to take account of performance so far. Summaries of these updates can be found at in the menu on the left of this page.