How machine learning innovations put Mobisol at the forefront of PAYG financing

Integrity. Innovation. Ethics. These principles are the foundation on which Mobisol has built its stellar reputation for trusted design and engineering. But the creation of cutting edge solar energy systems is not the only area of business in which Mobisol is leading the pack. Behind the razzle dazzle of exciting new products lies a credit assessment model built on machine learning that is a game-changer in improving the financial health of Africa’s rural residents.

For many new customers joining Mobisol, it’s not only light and energy that they’re getting to enjoy for the first time in their lives – it’s also access to financial services, in a secure PAYG payment plan that allows them to rent to buy their own solar home system. In developing countries without state welfare protection networks, where residents do not have access to banks and instead survive on fluctuating income levels, families are much more exposed to negative financial impacts.

Farming is part of the economic activities of 80% of Mobisol customers, while for 25%, it’s their main economic activity. This leads to irregular cash flows because the economy is anchored around the harvesting season and the price of goods can rise or fall depending on how the season is affected by weather patterns. In this context, gaining access to financial services can save people from going under and form a healthy buffer to the rough tides of cash flow.

Mobisol’s customer finance team has developed a credit assessment model that is at the forefront of financial analysis in developing countries and is designed to benefit customers, the company and the PAYG industry at large. This model protects potential customers from falling into unmanageable debt by ensuring that they have the capacity to complete the payment plan for the Mobisol product they desire. If the product they want doesn’t meet their budget, Mobisol directs them to one that they can afford, which also fosters a credit history for them that will benefit generations to come. This method lowers the risk of non-repayment to Mobisol and helps the overall quality of financial services in developing countries.

     

“We are proud of the fact that we were the first PAYG company to have assessed ourselves in terms of customer protection using the framework set by the World Bank for the microfinance industry.”

Luca Giacopelli,  Risk Account Manager, Mobisol Group

    

It’s vital to Mobisol that the relationship the company builds with customers is positive from the start, and so the credit assessment model requires minimal effort on a potential customer’s part. When a sales agent meets an interested customer, they conduct a five-minute questionnaire using a Mobisol-designed smartphone app. The potential customer is asked to authorize the usage of their data for internal purposes and their information is then sent to the Mobisol database. A member of Mobisol’s credit assessment team checks the relevant credit reference bureau to see if the potential customer has a formal credit history, such as a loan from a bank, and if so, whether they were reliable or not.

If the customer does not have a credit history they are given a chance and sent forward to the next stage. A credit assessment officer will call the potential customer to conduct an interview that takes up to 25 minutes to assess their eligibility. This process uses an algorithm that analyzes the structure of the household, its expenses and income, as well as a credit scoring system based on simplified machine learning techniques. In October 2017 Mobisol introduced machine learning and in less than a year and a half it has improved customer selection and reduced the risk of churn by an outstanding 10%.

Since the whole assessment process is carried out remotely, the potential customer is saved travel expenses because they only need to go to a Mobishop once they have been approved to pick up their product. This system also maximizes productivity for Mobisol because the assessment takes up to 45 minutes, whereas a field agent could only carry out two to three assessments per day. To counteract the risks involved with remote assessment, Mobisol uses its own benchmark of prices and productivity based around 35 economic profiles that range from farmer to hairdresser to double check the calculations provided by the potential customer.

Reporting anonymized data to credit reference bureaus is standard practice in the microfinance and banking sectors to protect customers from over-indebtedness, but this has not been the same for the PAYG industry. However, Mobisol voluntarily contributes anonymized data to credit reference bureaus in the countries in which it operates to minimize overall risk levels within PAYG. These bureaus are essential for improving access to finance and lowering risk levels in developing countries because they act as gatekeepers to prevent customers taking on higher debt levels than they can repay. Collecting this data is even more crucial in countries where people have not had any previous access to credit. By fostering credit history for reliable customers, Mobisol is helping them and their families to secure further access to financial services from other institutions.

Mobisol’s machine learning methods are top notch in terms of credit scoring to reduce risk and the customer finance team is constantly developing new techniques. Data from mobile money operators to predict repayment capacities is increasingly becoming an important part of the process. Mobisol has also demonstrated that it is able to constructively analyze its own methods and change them as needed to benefit the PAYG industry. In 2017 Mobisol was the first PAYG company to assess itself in terms of customer protection.

     

“For the past two years Mobisol has been at the forefront of these credit assessment practices, aggressively pursuing them to benefit the customer, as well as the company and the PAYG industry at large. The assessment highlighted some areas that required improvement and so we made these changes.”

Luca Giacopelli,  Risk Account Manager, Mobisol Group

    

Mobisol’s dedication to working on behalf of customers has brought it to discussions with GOGLA and other organizations about extending the ethical requirements expected of microfinance companies to PAYG. It’s vital that companies self-regulate to treat customers fairly and this is a core element of Mobisol’s ethos to conduct ethical business for the benefit of the planet, and its most vulnerable inhabitants.