Thanks to data, businesses today are making informed decisions and taking intelligent steps. While historical data has helped companies keep track of stock, analyze risks and foresee demands, it has also helped companies to offer and recommend new products to their consumers.Moving on the same line, Payu has been using data from the spend patterns of its consumers to offer them new credit products like Buy now Pay later. In conversations with ETCIO, Sachin Garg, Head- Data Science, PayU talks about this process and idea.
“We have two lines of business, one is credit and the other is payments. A person makes money to feed our families, fulfill basic needs or aspirational needs. And sometimes when the needs become more than what he earns, he borrows. Payu has created a journey for this person both for his spending needs as well as borrowing needs,” Garg said.
PayU offers a Buy now Pay later product, ‘Lazy Pay’, which is a short-term convenience loan. PayU has partnered with many merchant partners for this project. And the idea is to give the consumer a one-click payment experience, literally.
“Now, the question is how do we make sure that we allow only the consumer who has the ability to pay us back in the next 15 days. Otherwise, this product will not work for us,” he added.
“It is about making sure that we underwrite consumers who have the ability to pay us back and they are able to get their convenience product. When the consumer asks for a loan, institutions ask for several documents. Here, we do not have such data but we want to make sure that only the ones who have the ability to pay actually get this benefit,” he explained.
For this product, PayU relies on the spending patterns of the consumers. On the payments side, PayU has more than 250 million consumers who have been on its gateway to make the payment and so the company has their spend patterns. That is the only data that PayU uses to analyze if the consumer will be able to pay back or not.
“We have been able to underwrite the set of consumers who have equal in scale to what a bureau would have on their sets. It’s a case where using the large-scale transactional payment patterns that we have of users, we can actually give them convenience products like Buy now Pay later,” Garg mentioned.
“Once we build a model, it is always a continuing process of building new features. We have had this for a year. We are looking to do two things with this. One is, we want to be able to bring this product to more and more consumers and second is we want to make sure that the models are predictive enough so that our losses remain below a certain level,” he added.
If the consumers go to a merchant registered with us and see the option pay by lazy pay. It means that PayU has already assessed their ability to pay back hence the company is offering them this product. So, the consumer just clicks on the option “Pay by Lazy Pay” and the transaction is done as opposed to the other options where we have to remember our card numbers, pay IDs or even CVVs.
“If someone is doing an e-commerce transaction, how do we assess the risk of that? And how do we enable them to use this product if we feel it is a less risky transaction. So it is all about people doing digital transactions and PayU taking the risk for them making it convenient for the user by using ML and AI. But we do have a manual intervention to make sure that the models are fair,” he concluded.
