How graph analytics can prevent buy-now, pay-later fraud

Join online with today’s leading executives at the Data Summit on March 9th. Register here.


A series of integrated smash-and-grab thefts in the San Francisco Bay Area dominated our news feeds early in the 2021 holiday season. Dozens of people stormed San Francisco’s Louis Vuitton store and Nordstrom in nearby Walnut Creek, leaving a handful of luxury items valued at more than $ 100,000. The attacks, according to law enforcement, were carried out on social media and were carried out by people who did not know each other.

Now this is the digital version of organized retail theft – and it’s silent, nameless and faceless – and it uses a new kind of process called BNPL. BNPL (Buy Now, Pay Later) is a type of installment loan that lets you make purchases online and pay it off in weekly, bi-weekly or monthly installments. This method of purchase has become widely popular in the US and Europe. BNPL services are growing at a rate of 39% per year, and PayPal, Amazon and Square are also coming into action and acquiring existing BNPL companies in a multi-billion dollar deal. While buyers can quickly get their hands on Xbox gaming systems, laptops and expensive purses, BNPL has opened the door for potential fraudsters who pay only 25% of the original price for the product and avoid paying the rest.

BNPL Fraud: Account Takeover, Fake Accounts and Digital Flash Mobs

BNPL fraud occurs in various ways. Meanwhile one Account takeover, Fraudsters gain access to existing BNPL customer accounts and make unauthorized purchases. Even cheaters Open a fake BNPL account Using someone’s stolen identity. What helps fraudsters and harms customers is BNPL’s lax identification and verification procedures. Often BNPL providers rely on data, internal algorithms or soft credit checks to determine a person’s creditworthiness. This means that they may miss serious fraud indicators, such as an address or phone number that does not match the applicant. In addition to account takeovers and fake accounts, fraudsters can join together to make purchases. Groups of bad artists spread across different geographical locations and network addresses can attack at the same time. You can see 100 people logged in and buy a ો 600 console for only 20-25% of the product value – and no previous data linking these people. In this digital flash mob the fraudsters go back and sell these items at full market price.

BNPL fraud presents a new challenge for traditional banks that offer their own BNPL; It is basically an instant loan application, at the time of sale, deduct the credit check. This means that banks will absorb the loss of any fraudulent loan. Banks pay customers in advance for customer purchases, meaning they risk losing up to 100% of the loan value through fraud. Also, the fraudster can open a new bank account, get a credit card and start shopping under the same synthetic identity. The fraudster then defaults on the payment, resulting in a total loss. Adventurous fraudsters can also hire helpers to check stolen credit card numbers on a mobile app. If the stolen card number works for small purchases, then it can be used for large purchases.

The graph can detect markers of fraud

The risk of fraud increases as more people use BNPL. During the 2021 holiday season alone, about 40% of people used BNPL loans like Affirm or Klarna to pay for holiday gifts. How can BNPL providers secure their automated digital processes? First, providers can create more stringent identity verification – during account opening and checkout. They may also use machine learning technology to identify unusual shopping activity that may be linked to fraud. Graph analytics is a set of analytical techniques that highlight how organizations like people, places and things are connected to each other. The graph identifies connections, relationships, and patterns. Financial services organizations and credit card providers use graphs to detect potential fraud – during the application process as well as when purchasing. When a person applies for a credit card, for example, Graph can closely examine the features of his application. Are there other apps that share the same email / phone / address / device? What are the short paths from input application to blacklisted application and the number of attachments / hops? Graphs can give different weights to each part of the application to generate fraudulent path scores. Credit card providers, equipped with this score, can predict the risk of a single application – all in real time.

We can similarly apply graphs to BNPL scenarios to actively “catch” fraud during the actual attempt instead of after the fact. Consider this scenario: John fills out and submits the BNPL application. Meanwhile, the BNPL provider pushes its data into the graph, queries are executed and relationships are highlighted. Graph analysis scores out. A low score means high risk, while a high score is likely to be approved. All this can happen in real time if the BNPL provider links the graph database with their other algorithms. Behind the scenes, the graph will analyze various data points, such as John’s name, address, social media accounts, IP address, email address, and date of birth. Is John’s name and date of birth associated with fraudulent petitions? Has the IP address been used for multiple (fraudulent) applications in the recent past? Does John even claim to be that person?

Preventive quality control

Banks can use graph analytics to check Perfection of customer behavior To find potential fraudulent BNPL loan applications at the point of sale. In this way, the credit request is denied before the fraudulent merchant gets his hands on the goods. In addition, Graph Analytics (aka Link Analytics) ensures that there is no correlation between applicants and previous fraud cases or organized fraud rings.

Graph detection can be performed natively if the data is stored in a graph database, but graph algorithms can also be applied to data not stored in a graph format – although in these cases the query may be slow and the results may be incomplete. Some libraries of graph algorithms exist to detect and score relationships between people, places, and events. For a public graph algorithm library, the answer to go is NetworkX. There are also libraries provided by some graph database providers.

If the data is constantly refreshed and updated, real-time analytics allows the organization to find hidden patterns within the data before any transaction or credit application is approved. BNPL providers rely heavily on data to grant or deny an individual loan, these companies need access to the most accurate data results available. Basically, better real-time data gives less successful fraudulent transactions. The implications are huge for BNPL providers, who have historically been the victims of fraud at the cost of doing high-volume business. Fewer frauds, in turn, result in fewer customers facing inconvenience as they wait for their money back after a chargeback.

Todd Blashka is the Chief Operating Officer at TigerGraph,

Venturebeat’s mission Digital Town Square is set to become a place for technical decision makers to gain knowledge about the changing enterprise technology and practices. Learn more

Similar Posts

Leave a Reply

Your email address will not be published.