Online payment fraud is rising fast. In 2021 alone, it increased by 285% as fraudsters exploited the growing volume of online commerce and weaknesses in businesses' fraud defenses. And the threats continue to evolve, with devastating impact: it's estimated online sellers will experience losses exceeding $206 billion between 2021 and 2025.
Payments fraud is not an issue any business operating in the digital economy can ignore. Nor is it an area where businesses can stand still, relying on legacy tools and risk management strategies. The threats are too dynamic, and the cost of not managing them is too great.
We're excited to announce the launch of Fraud Detection Pro to empower businesses to tackle these challenges head-on. The enterprise-grade fraud solution serves the most complex businesses, allowing them to deploy sophisticated fraud management strategies and balance risk and performance globally, maximizing revenue.
We built Fraud Detection Pro with three key customer needs in mind:
Adaptability: Advanced machine learning detects new fraudulent trends and utilizes network-wide intelligence in real-time.
Customizability: Define rules to tailor fraud strategies for specific target segments or criteria, using custom data to build unique segments. Leverage approve and decline lists to block fraudsters and identify legitimate customers.
Optimization: Powerful analytics, reporting, and testing capabilities that inform new fraud strategies and optimize business performance.
Businesses are facing significant economic headwinds. Consumer demand is falling as the cost-of-living crisis hits, while inflationary pressures in supply chains and salaries squeeze margins further. More than ever, merchants need to ensure payments empower them to capture every legitimate transaction and maximize revenue.
But it's easier said than done. At the same time, fraudsters are becoming more sophisticated in their tactics, forcing businesses to deploy stricter risk strategies. One of the biggest trends is increased automated attacks, primarily through bots. These attacks make it easy for fraudsters to scale their efforts. For example, it can cost less than $200 to attempt 100,000 account takeovers, with a success rate between 0.2 to 2%.
Also, that threat is not uniform. Fraudsters go after different businesses in different ways. Ecommerce, airline ticketing, money transfer, and banking services businesses seem more vulnerable to credential stuffing and account takeovers. Online marketplaces are primarily targeted by fake accounts, false advertising, order cancellations and fake buyer/seller closed-loops. The crypto sector gets hit with fake exchanges, wallet takeovers and Man-in-the-Middle Attacks (MITM). Online gaming businesses suffer most from fake third-party top-up services, credential stuffing, account takeovers, and Streaming Potluck schemes.
The lesson for businesses is that a rigid, one-size-fits-all fraud solution is no longer enough. It leaves them to make a zero-sum choice—prioritize fighting fraud but accept that legitimate customers will get blocked (or frustrated enough to abandon their baskets); or prioritize sales and hope that the higher volume counters the higher rates of fraud.
Neither approach is sustainable, leading to lost revenue, chargeback rates increase, a higher cost of transacting, customer churn and damage to brand credibility. We've built Fraud Detection Pro so merchants don't need to make that choice.
"Fraud is an inevitable part of processing payments as it continues to evolve and becomes increasingly creative. We need to be fully prepared to respond quickly and with precision. Checkout.com’s Fraud Detection Pro solution is like using a scalpel, compared to some of the other tools in the market that are more like a sledgehammer."
Philip Quinn, Senior Product Operations Manager, Curve
With Fraud Detection Pro, a business is empowered to tailor its fraud solution to meet its unique requirements and test, learn, and adapt as new threats emerge. Here's how.
Fraud Detection Pro’s Machine Learning feature is trained on billions of hard and soft data points from Checkout.com’s global network of merchants. It learns from patterns of real fraud across multiple sectors and countries, and applies these insights to identify and stop suspicious activity at the point of transaction. Without this, a merchant is more exposed to existing and emerging fraud patterns, as they don't have wider insights and historical data about existing and emerging fraud patterns.
Fraud Detection Pro in action
Company A experiences an instance of fraud. Fraud Detection Pro’s Machine Learning automatically records and learns from this fraudulent activity, storing the payment attributes as potentially abnormal. Weeks later, when scanning a transaction at Company B, Fraud Detection Pro identifies the same attributes as the attack on Company A. This payment transaction is scored as highly likely to be fraudulent and automatically declined.
Fraud Detection Pro allows merchants to set rules that match their unique risk tolerance. And just as merchants have different risk appetites, so an individual business will want to apply different criteria to its sales.
For example, a merchant will probably deem a returning customer using the same IP address, buying something expensive as ‘safer’ than a new customer from a flagged IP address buying something of lower value. Fraud Detection Pro gives merchants the flexibility to treat these two transactions (and other segments) with different risk classifications and to build appropriate fraud strategies for each.
Merchants using Fraud Detection Pro can use this segmentation capability in varied ways. For example, they can separate high-risk versus low-risk geographies, segment by different payment methods, or even detach specific product codes that experience more fraud attacks than others. As well as setting individual fraud thresholds for different segments, merchants can also define what action is triggered when a threshold isn’t met (e.g., accept or decline, send to 3DS, or undertake a manual review).
These actions provide merchants with additional levers to block more fraud or reduce strictness to allow more transactions through—and adapt these measures as their risk appetite develops.
Fraud Detection Pro in action
Company B identifies that its fraud rate is low. They see an opportunity to take on more risk by accepting more transactions to maximize acceptance rates. Using Fraud Detection Pro, the merchant adjusts its Machine Learning settings to be less strict (i.e., increasing the score at which transactions are declined). The merchant continually tests and adjusts these settings until it finds the optimal balance between fraud costs and acceptance rates.
Fraud Detection Pro includes a range of other capabilities that enable merchants to build on their Machine Learning setup and fine-tune their risk strategy by building specific rules that address the unique challenges of their business.
Enhanced protection with flexible rules: Merchants have unlimited scope for creating rules that suit their business, including advanced velocity and custom rules. Multiple rule criteria can be grouped to produce a weighted behavioral score for more accurate outcomes. And merchants can safely experiment with new rules by testing the potential impact of a rule before pushing it live.
Block fraudsters and allow legitimate customers: Merchants can upload and maintain approve and decline lists to identify legitimate customers and ensure they deliver a frictionless payment experience while keeping known fraudsters out.
Higher authentication: Merchants can validate risky transactions by applying additional authentication using Checkout.com’s integrated Authentication solution. Businesses can set different levels of authentication to increase or minimize friction as needed.
Streamlined manual reviews: Merchants can access a centralized view of all transactions requiring manual review, alongside rich contextual data on why each transaction has been flagged, saving time and increasing the accuracy of decisions.
Risk strategy builder: Merchants can review their various risk strategies via a decision-tree visualization and adjust strategies by adding, amending or subtracting components.
Fraud Detection Pro in action
Company C, a gaming merchant, uses insights from Fraud Detection Pro to understand that new players using prepaid and credit cards carry a high potential risk. With Fraud Detection Pro, the gaming merchant includes the account age in the payment request and builds a rule that blocks transactions from a prepaid card or debit card for all accounts less than one week old.
Fighting fraud should never be about solely prioritizing security. It must be a holistic approach that equally considers the customer experience and protects legitimate transactions. Research from Checkout.com found that 34% of people have been permanently put off a shopping site by a declined payment; but a much higher 52% will abandon their baskets if the payment process is too complex.
Unlike many fraud tools today that mandate strict risk strategies that create friction and frustrate customers, Fraud Detection Pro has been built to work with the customer experience.
Fraud Detection Pro enables merchants to balance fighting fraud and protecting the customer experience with a three-step process:
Analyze: A highly detailed and visual analytics dashboard and library of reports help merchants monitor the impact of their risk strategies on key metrics such as conversion, acceptance rates and performance against scheme chargeback limits.
Investigate: Merchants can investigate and diagnose the root causes of false positives and false negatives and use these insights to eliminate overly burdensome anti-fraud measures.
Test: Shadow mode testing provides a safe way to examine the potential impact on the customer experience of proposed new rules, Machine Learning settings, or entire risk strategies without affecting live traffic.
Risk management needn't be purely a defensive approach. Instead, it can maximize revenue by increasing acceptance rates and reducing friction for legitimate customers, leading to more conversions.