MoneyMan implemented the systems of statistical analysis of data of SAS Enterprise Miner and SAS Text Analytics

April 22, 2015. The MoneyMan company announced implementation of the systems of statistical analysis of data of SAS Enterprise Miner and SAS Text Analytics. Their use allowed to increase efficiency as processes of credit scoring and target marketing, and processes of collecting of arrears.

Goals and Objectives

The business model of online microfinance of MoneyMan requires permanent development of scoring model. As assessment of solvency of the borrower is performed completely remotely, i.e. there is no stage of an internal meeting, an important role is played by the intellectual automated data analysis about the potential client.

Project Progress

MoneyMan has a unique technology stack of own development. One of advantages of the company is its innovation model of decision making about issue of a microloan "Scoring 5.0" which algorithm includes more than 500 units among which there are rating models. Feature of a system of scoring is use of the centralized decision making in the automatic mode. The technology platform allows to manage flexibly decision making process, cutting off level and also "trees" of solutions. The model makes the decision on issue of a loan on the basis of a set of data, including technologies of multiple search, internal credit history, an antifraud service "national Hunter", data of several bureaus of credit histories and other external sources, for example, information on payments of potential borrowers into accounts of mobile communication and data from accounts of borrowers on social networks. Assessment of solvency of the borrower is also influenced by such exotic parameter as behavior of the user on the website MoneyMan.

On the one hand to increase efficiency of credit scoring and detection of cases of fraud, and with another – to estimate and predict needs of the potential and existing clients, the MoneyMan company implemented the solutions SAS Enterprise Miner and SAS Text Analytics. The methods of deep data analysis (data mining) and the deep analysis of the text (text mining) entering them allow to detect the hidden patterns, interrelations and trends and other uncommon and useful knowledge in the accumulated and collected information. The company performed implementing solutions and setup of analytical models generally by own efforts.

Development Plans

As for opportunities of analytical CRM, using tools of SAS the MoneyMan company is going to develop the new marketing campaigns including aimed at increase in efficiency of programs of loyalty of borrowers and also to personify both product offers, and techniques of collecting arrears. In further MoneyMan is going to develop the mechanism of selection of the unique offer for each client on the amount and term of a microloan with the prediction algorithm of a response of the borrower to such personal offer.

Result

SAS Enterprise Miner containing effective methods of statistical analysis allowed MoneyMan to aggregate and investigate data arrays, using advanced methods of forecast and descriptive modeling. The user-friendly interface gives the chance to quickly create models using the convenient interactive environment, to compare models and also to manage them according to business challenges, representatives of the company integrator reported TAdviser.

Use of tools of text analytics of SAS Text Analytics, in turn, allowed MoneyMan to work with the unstructured customer information arriving from different sources, including comments of call center operators, data from social networks from forums, information on addresses and complaints, etc. Intellectual algorithms of SAS detect and take important knowledge, patterns from text arrays and, then, allow to enrich the available data structure.

"At the end of 2014 the NPL 90 level + on the loan portfolio of MoneyMan in Russia did not exceed 14%. At the same time the level of approval of requests for microloans (approval rate) was 15% that is a high rate in an online microfinance business model. We estimate that application of the systems of predictive analytics of SAS in combination with unique developments of the scoring MoneyMan model until the end of the current year will increase the level of approval of requests for more than 30% when preserving the current credit risk. It will allow to increase the qualitative loan portfolio, to show бóльшую profitability that will affect investment attractiveness of the company positively", – the director of MoneyMan risk management Ekaterina Kazak noted.