Assessing Risk Behind Buy Now, Pay Later concept

Buy now, pay later is fast being accepted as an innovative and alternate means to pay for purchased goods online. Service providers, in order to eliminate or reduce risk are adopting tools, such as Machine Learning, analytics and alternate data to assess the risk associated with credit extension.

We have been hearing about ‘’Buy now, pay later’’ for a while now and have witnessed its popularity over the last few months. Shoppers have shown much interest in the online credit concept and eagerly look to get qualified so they can avail the credit. A good number of online portals have adopted the concept in a bid to reduce transaction failures and provide a pleasant customer shopping experience.

Reports suggest that about 50+ online portals in India have collaborated with fin-tech companies, such as ePaylater to roll out this unique and innovative payment concept that is fast becoming a craze among online shoppers. While the concept finds increasing popularity as more customers wait in line to avail the micro-credit, the question of risk associated with the credit still remains.

How do solution providers assess risk and choose the customers?

In the modern business world where ‘’credit’’ is the catch word, extending micro-credit and allowing shoppers to—Buy now, and pay later, definitely comes with its own set of risks. On the surface of it, the entire process or concept looks quite easy but at the back-end this concept involves a lot of research and analysis; there is a very refined risk assessment process involved.

Assessing eligibility

It is true that not every customer or shopper may be eligible for such a credit and moreover, credit extension should make business sense for both service providers and e-commerce portals. Eligibility is provided only to a certain number of customers and is based on several criteria. As Akshat Saxena of ePaylater succinctly puts it–

‘’For this we have spent six to eight months on building this product. Now we deploy machine learning, AI and other sophisticated analytical tools that combine a lot of information together and assess the expected short risk. Only if the customer is an eligible candidate do we allow the transaction to go through’’


Assessing the risk associated with extending the micro-credit is a big back-end task that involves a team of analytics professionals and programmers. It is not an easy task as the information available about customers is limited and it’s never easy to predict the future payment capacity of a customer. Assessment of a customer or risk assessment is done by sourcing alternative data of a customer. Alternative data is any data that is not related to the customer’s credit footprint but at the same time provides a peek into customer’s purchase and payment patterns.

Some of the major sources of Alternate data are –

  • Utility Bills payment
  • Verified Income
  • Cash Withdrawals
  • Digital Bank Statements (30, 60, 90 day transactions)
  • Bank Balance
  • Multiple Account Types—current, checking, savings, etc.

Fin-tech service providers also resort to social media profiles of customers to better understand and assess them. When it comes to assessing risk associated with a customer, every bit of information helps. The first step in this process is for customer to log in to the account of a fin-tech service provider and provide details such as—Email, Phone number, Contact details, identity proof like an Aadhaar card, etc. The service provider then uses Machine Learning and other analytical tools to assess the customers based on which a customer is profiled as—eligible or not eligible.

Though the aforementioned criteria and analysis are used to carve out a customer profile there is always the element of uncertainty that could completely undermine the analytics and assessment process and lead to a major risk.

Varying levels of credit line

All eligible customers will not fall under the same credit line bracket, some are allowed smaller amounts while others are allowed larger. This again, depends on a lot of factors beginning from credit history, to shopping history to payment capacity. In short, an in depth analysis of each customer is done based on previous shopping and payment patterns that gives rise to his or her profile and the subsequent amount of credit extension.

Based on the profile map, customers are extended credit anywhere from INR 2,500 to 5,000 to even 10,000 in certain rare cases. But again, this depends on both the fin-tech service provider and its associated online portals. Customers who get a lower credit have a chance to increase the limit provided they create a positive credit footprint by regularly paying on time.

Credit system in the market place has been existent since time immemorial, there was risk in those times and there is risk even now. The difference now is that credit system has moved online and at the conjunction of online shopping and online payment the Buy now, pay later concept seems to be a perfect solution for purchases that require lower amounts.

Fin-tech service providers are betting big on online credit extension but they are cautious enough to extend credit for only qualified customers and that too in lower limits, which is a fairly calculated risk. The concept will see greater use as all stakeholders from customers to shopping portals to service providers move in tandem and benefit through each other.

Sahil Arora Author

Leave a Reply