In today’s e-commerce landscape, fraud prevention and detection solutions must include tools such as graph analysis to track data across numerous transactions.
Over the last few years, the e-commerce sector has witnessed unprecedented growth. As volumes of digital transactions continue to skyrocket, merchants and payment providers find themselves amid a fraud epidemic. Last year, fraud and cybercrime cost the United
Kingdom a staggering
£4 billion — a significant rise from the
£2.4 billion reported in 2021.
As the tactics employed by fraudsters become ever more sophisticated, traditional anti-fraud controls are in danger of becoming overwhelmed and, dare I say, obsolete. Today, I’ll be explaining why I believe it’s imperative that digital businesses continually
adapt to this new age of heightened fraudulent activity, as without the right technologies, many will struggle to survive — particularly if they are unable to identify chains of activity or defend themselves against attacks from fraud rings.
What is a fraud ring?
Fraud rings are organised groups of professional criminals and are typically much more experienced than small-time thieves or opportunists. Members of a fraud ring usually bring specialist skills to the group, such as social engineering,
carding or identity fraud.
Although fraud rings are highly organised and often sophisticated, they do create recognisable patterns in merchant traffic. Numerous fraudster-controlled accounts may reveal a relationship with each other by exhibiting similar features or behaviours. For
example, if a stolen credit card is used to make a purchase, the payment information will be passed to a payment processor to complete the transaction. In this scenario, the chain of fraud involves multiple pathways:
In the past, fraud detection solutions were not powerful enough to monitor these complex activity chains. Instead, they were limited to flagging large or unusual transactions that didn’t align with a cardholder’s regular spending habits.
Identifying fraudulent transaction chains
In 2023, outdated fraud solutions unable to identify chains of suspicious activity are simply redundant, especially in a landscape where cybercriminals are becoming increasingly more sophisticated in their efforts. Modern scammers are adept at covering their
tracks, with an average fraudulent scenario involving bank accounts, fund transfers and purchases. In addition, multiple pathways also exist, including two or more banks, the criminal and the entities receiving the funds.
Without the ability to identify chains, outdated fraud systems often throw up false positives or negatives. False positives flag legitimate activities as fraudulent, while false negatives are instances where fraudulent activities are missed. Both outcomes
are detrimental to businesses and cause unnecessary disruptions in everyday operations, significant financial losses and often poor customer conversion.
As e-commerce and digital payments become increasingly popular, it becomes imperative that anti-fraud systems can identify transaction chains and link them to different entities. To truly understand and fight fraud in 2023, businesses must map out entire
chains of events using data analysis and pattern recognition while identifying links between seemingly unrelated activities.
Enhancing tech solutions
Although powerful, modern risk control management systems (RCMS) are generally not enough when it comes to detecting and preventing chain attacks. Instead, it is now crucial that payment providers and e-commerce retailers deploy the most innovative technologies
to support their fraud prevention and control strategies.
Graph analysis is the perfect example of one of these technologies and is known to augment the work of an RCMS. A graph analysis model allows multiple activities in a chain to be analysed and blocked, discovering the accounts involved in fraudulent attacks
as well as so-called “hidden fraud” — scammer accounts that look genuine and go undetected by traditional solutions. If fraudulent activities are found, graph analysis can block chains of activity, and when fraudsters attempt malicious actions again, the threat
is neutralised. The entire process is repeated until the criminals have exhausted their efforts and ended their fraudulent activities.
Graph analysis is an impressive tool that adds another layer of protection to already sophisticated anti-fraud solutions. With all that being said, an effective and comprehensive strategy should always combine automated monitoring with manual analysis by
experts. For Ecommpay, our approach ensures a 97%+ fraud-detection and prevention rate without affecting customer interactions.
An effective anti-fraud tool
I have personally seen the effectiveness of graph analysis while working with merchant clients. In data collected by Ecommpay, the technology was found to decrease the number of fraudulent transactions suffered each month significantly.
In one example, a merchant reduced fraudulent transactions from a peak of $73,961 to $6,680, representing a reduction of over 90%.
Similarly, graph analysis reduced another merchant’s fraudulent transactions from spikes of $27,793.59 and $22,323.23 in consecutive months to just $4,612 — a decrease of 83% and 79%, respectively.
It’s worth noting the ability of graph analysis to detect fraud improves when presented with greater volumes of data. That means the system doesn’t require recent information and can be used to identify new fraudulent attacks based on historic data from
months or even years ago.
Ultimately, graph analysis offers numerous advantages over traditional anti-fraud systems. Fraud chains are fluid and dynamic in nature, with new players appearing over time and continually evolving their tactics. Therefore, modern solutions must employ
equal dynamism to terminate the activity of fraud rings and their chains of activity, quickly depleting the scammers’ resources and thwarting their attacks.
There are no one-size-fits-all solutions
Alongside revolutionary technology, payment providers and e-commerce merchants should carefully consider how their fraud detection and prevention measures are structured.
Many companies have outsourced their fraud-protection capabilities to cybersecurity and other preventative firms. Unfortunately, this trend is taking place as an influx of new digital payment adopters emerges, primarily triggered by the pandemic. The latest
wave of e-commerce customers is amongst the most vulnerable, as many have not yet been exposed to or warned of the latest methods used by scammers to trick them out of their personal data and money.
Although merchants are rightly concerned with growing their customer base and improving checkout conversion, those improvements cannot come at the expense of stolen consumer data and funds. If a business employs risky practices to boost its performance and
sales, it could suffer major damage to its reputation.
Payment providers have a responsibility to educate and inform merchants about the consequences of fraud and how to improve their risk-control systems. In return, merchants must ensure that anti-fraud controls are tailored to their industry and needs — whether
that’s retail, finance or travel. By utilising proprietary systems, merchants can adjust their anti-fraud filters accordingly, maintaining high conversion levels while achieving maximum revenue and low fraud rates.
The increasing sophistication of scammers and the complexities of identifying fraud rings and chains make it essential for payment providers and their merchants to have robust anti-fraud solutions in place.
Today, organisations need to invest in the latest fraud prevention and detection solutions — including graph analysis — to track data across numerous transactions. In addition, merchants and payment providers must strengthen their relationships and decide
if they can each tailor their product offerings to suit their business needs better, while utilising innovative technologies to ensure they effectively protect against, block and mitigate fraud rings and chains in 2023 and beyond.