OpenIDEO is an open innovation platform. Join our global community to solve big challenges for social good. Sign Up / Login or Learn more

Profile

Recent contributions

(1)

Contribution list

Recent comments

(3) View all

Our interfaces and technology are certainly readily accessible not only to the farmers we intend to reach, but to other relevant players in the value chains represented. According to latest statistics, from the Communications Authority of Kenya, mobile penetration is at 88.1 % with 37.8 million active subscribers to such services. The internet/data market has over 21.6 million subscriptions. The portion of the Kenyan population accessing internet services reached 74.2 per 100 inhabitants as of January 2016. Mobile money transfer service subscriptions are at 28.7 million users. Entry level smartphone devices entry points are at less than $30, with an estimated 30% of the population estimated to adopt the technology by 2020. This data, as well as our on the ground research which showed that every user polled had the immediate ability via one of our portals to access the system, shows that the use of mobile phones, mobile data, and mobile money services are uniquely and readily available to the Kenyan market. Our solutions are based on the use of such services. In order to facilitate all levels of engagement, or system is available in various formats. First is an SMS gateway which requires only the ability to send and receive SMS messages. The second is our smartphone application, which provides access to all services within the network. We have developed a USSD interface that will reflect the same functions, that is ready but will not go live until we reach a certain threshold of active users due to cost considerations. In addition to this we offer a web portal that users and suppliers within the system can use to access information on their accounts or specific goods.

As stated, the loans being provided by our system are in two different formats – in-kind input loans, as well as cash loans. Loans taken on a purely cash basis are indeed available for use on a consumption basis, while the loans used for agricultural inputs are treated separately, with interest and repayment schedules reflecting these criteria.

Loans not tied to assets such as inputs are indeed riskier than those which are. We seek to mitigate risk in a number of ways. Kenya's loan market is somewhat unique, with unsecured loans based on mobile usage and mobile money history proving successful to date. For example, the leading platform for such loans, which is untargeted toward a specific segment, uses mobile data and MPESA history as a basis for credit scoring. This platform, on latest figueres, shows a non-performing loan portfolio of less than 2%. Our system uses the same data, but allied with 5 other relevant and verifiable data points that take into account the financial and social profile of a farmer. We have a data calculated tiering and grading system, that takes into account the farmer's financial profile, repayment history and production cycle. We integrate with the nationwide Credit Reference Bureau to perform analysis on farmer's ability to service financing. We are in constant contact with users to let them know how and when to make engagement with the system in an easy to understand way. Once we reach a certain threshold of active, engaged users, we will make agreements that will lead to loans provided by our system being insured on a total portfolio value basis. As our system and user base grows, through the use of strategic partnerships, some of which are in place and others which we are constantly pursuing, we aim to shift the profile of loans away from pure cash/consumption loans as much as possible to a model where such loans are facilitated on a similar model to agri-input loans. Users will have the option to request their financing for not only cash or input, but other necessities. Through our partnerships, users will be able to request loans that are to be used directly for various needs, such as device loans, school fee loans, storage facility payments, and transport services. By providing direct facilitation of such payments on a similar model to our input loans through the use of administrated e-value and payments, we aim to reduce risk and increase users' access to vital services.

Risk engendered by the trust of third party reporting is always a concern, and we have several structures in place aimed to increase verifiability and accuracy of such reports. The majority of our data points sourced for the assessment of credit risk are based on automated, digital systems where the data is requested via API and comes directly, in unadulterated format from the relevant system. Biographical data is sourced from the nationwide Integrated Population Registration System (IPRS). Credit reference checks are performed by check to the national Credit Reference Bureaus, which provides detailed history of financial engagements and obligations on any citizen. Mobile money history and mobile data and credit usage is sourced directly via Safaricom's MPESA API and through user approved download of statements and history. As these data points come directly from their relevant systems, they are accurate, immediately available, and incorruptible. Data reflecting purchase and usage of agricultural inputs is sourced from our closed-loop verification and electronic value system, with cash and opportunities for false reporting eliminated by the need for real-time verification within our system. In order to reduce the risk of of false reporting in terms of farm outputs and sales, we have are forming partnerships with the relevant endpoints at which the sale of farm goods and reporting thereof is performed. Our initial focus, as the system grows, is on particular value chains which we have partnerships with, which have in built methods of reporting and verification. For example, one of our initial value chain partnerships represents a pool of over 15,000 farmers. This organization has a daily, digital reporting system of weights and amounts of goods produced which are input into the system automatically on a daily basis. We are able to source this data directly. In addition to this digital reporting of outputs, we are making relationships with major off-takers and aggregators of sales, by which we can source the actual financial data that the sale of these goods reflect. As a further verification step, we have liasion officers who work in conjunction with the relevant farmer groups and off-takers in order to provide on site verification of such transactions.

Input purchases, while necessarily seasonal in nature, are an important indicator of the performance of a farmer and therefore a very useful metric by which to facilitate credit scoring. Verified purchases of the legitimate, relevant agricultural inputs such as fertilizer, seed and pesticides help us to amend credit profiles as they demonstrate a) a tangible measure of the financial capacity necessary to service such a purchase, and b) that the farmer is using the correct and legitimate modes of input that will serve to enhance their yields, and therefore future income and financial potential. This information, while important, is not the only data that feeds into a farmer's credit score. We ally this information with several other data points, including biographical data, Credit Reference Bureau information, market prices, farming output yields and sales, mobile money history and mobile credit and data usage. By combining this array of relevant information in conjunction with input purchase history, we are able to more accurately build credit profiles and histories.

As far as agricultural inputs, we have implemented control structures to verify and monitor their purchase. In order to monitor the mode of purchase, i.e. outright or via informal store credit, we seek to firstly eliminate direct cash purchases from these transactions, and have them be facilitated completely through electronic value. Input loans in our system are disbursed through the use of Chipua electronic value, which can be exchanged directly for agricultural inputs at participating agrovets. In the case of a customer who has received such a loan, the monitoring is simple. Upon purchase of an input, the customer first performs the verification check on the product they are buying, and then receives the opportunity to confirm the purchase via the app or SMS. In confirming the purchase, our system logs the transaction, and we are able to settle the payment directly with the participating agrovet, who logs the transaction on their mPOS device as an e-value payment, which we are able to verify through our system and the user's input. In cases where the user has not yet received a loan, but wishes to purchase outright an agricultural input, we offer the user the ability to directly purchase e-value through our SMS gateway or application via mobile money, which they can then exchange for the input they desire. We seek to make this attractive via supplier level discounts, which make such a transaction desirable over a traditional cash purchase. In these cases, such purchases require confirmation within the system, and we have a record allowing us to know how it was purchased. It is still possible for customers to buy inputs from agrovets on an informal store credit basis, but in the case of such transactions, our credit scoring algorithm takes this into account in understanding that while the customer has received a relevant input, they have not yet made full payment for it, and amends the user's credit score accordingly. It is not possible to claim an outright purchase while receiving credit, as the agrovet can not enter the purchase into the system as an outright purchase without the relevant systemic confirmations.

We recognize that the provision of in-kind, input loans rather than cash loans would serve to reduce risk. However, we also recognize that the intended customer base for this product have several, pressing day to day needs that they need to meet - such as food, school fees, transport, wages, storage costs and many others - that they are not able to service currently. In many cases the farmer needs to subsist first, before they are able to concentrate their expenditure on agricultural inputs that will improve their farming capacity. For this reason, we aim to provide users with access to both types - cash loans that would meet such costs for non-agricultural needs, as well as the in-kind loans for farming inputs that our electronic value system facilitates. The inherent risk profiles of these differing types of loans are reflected by our scoring algorithm in our repayment and interest structures, with the time periods and repayment schedules being structured to accordingly service the level of risk represented.