In the market for a fraud prevention tool, but not sure what features you need? We’ve got you covered.
According to statistics, there is a high chance that fraudsters are hurting your business, so you probably already know how frustrating, costly and time-consuming it can be to catch them.
This is true whether you’re trying to stop one of them, or hundreds who attack your company on a regular basis.
Which is exactly why you need a fraud prevention tool.
In this post, we’ll examine reasons why you might be a target of fraudsters, and what kinds of features your fraud prevention tools should use to combat them.
What Makes You a Target of Fraudsters?
The first thing to understand here, is that fraudsters attack all kinds of businesses, and not just the big ones. But of course, some companies are attacked more often than others. Here’s how to know you are in a high-risk vertical:
You Operate As a Digital Wallet
Fraudsters are often looking for the path of least resistance towards cash. So if you are dealing with money in any way, shape or form, this means your business interests them.
Why is that? Well, it’s a lot easier to drain someone’s crypto account than, say, purchase an item on an online store, ship it, and resell it later.
With digital wallets, it’s much more direct. Fraudsters can acquire the login details of your users on the darkweb, impersonate them, and wire the funds to themselves. This is a process known as ATO, for account takeover, and it tends to affect:
- Crypto exchanges
- Loan providers
- Online stores that offer gift cards and vouchers
An important note: there are often two points to secure: the deposit and withdrawal. For the deposit, you want to ensure the funds don’t come from a stolen credit card number.
For the withdrawal, you want to make sure the funds aren’t going to a “drop bank account”, which is an idle account opened specifically to convert digital currencies into cash.
You Are Selling High-Value Items
This one is common sense, but high value items means high risk. Here again, it helps to imagine how fraudsters think to understand this correlation.
Fraudsters tend to automate their actions, as they have a low success rate. So for instance, they acquire a huge list of stolen credit card numbers, most of which are frozen or blocked already, or simply invalid. They will then test all of them with very small purchases, until they think they’ve hit the jackpot.
The next step is to buy something with that card, which they can resell. As you can imagine, it’s a lot easier to sell one Macbook Pro for $3,000+ than 10 entry-level laptops for $300 each, which is why Apple users are often the targets of elaborate phishing scams.
But the list isn’t limited to high-end electronics. Luxury goods like jewelry, collectibles and limited edition artworks are also highly sought-after by fraudsters.
You Are Generous With Your Loyalty Programmes and Sign Up Bonuses
Most online businesses want to reward loyal customers with special offers, coupons and discount codes. Or they want to grow through the power of referral links, which is a common practice in the world of SaaS, fintech and iGaming.
The problem? It’s creating an incentive for multi-accounting. This is the process where fraudsters will create multiple accounts or divert traffic to a site in order to scale their rewards.
At worst, it will be a few individuals creating several accounts. At worst, these bonus hunters, as they are sometimes known, can work through extremely sophisticated methods, including bot attacks and organized criminal rings that hijack traffic from websites or apps so they land on the promotional offer.
Common Types of Fraud to Fight
Based on the examples above, we can highlight a few common fraud types your business may face. These include:
- Transaction / Chargeback fraud: someone buys something on your site with a stolen credit card number. The real user contests the charge, and you have to pay the fee for refunding the money.
This process can be expensive and resource-intensive. Not just in the short term, but credit card companies also have increasingly strict fines for punishing merchants with lax security.
- Bonus abuse / Multi Accounting: one fraudster or a group of criminals create multiple accounts on your site to cheat at casino games, reap the benefits of referral programmes, or abuse discounts.
The biggest challenge here is to spot connections between your customers. This is particularly tough if you want them to onboard without asking too much information.
- Account Takeover (ATO attack): your users lose their login credentials, and a fraudster impersonates them. They drain the funds on the account, spam links, and make the account unusable.
This kind of fraudulent attack is particularly devastating for your business reputation. If your security measures, your users will blame you for letting someone sign in with their account. Worse, your business reputation will suffer, as the user will complain about your service to anyone who listens.
The Must-Have Features of Your Fraud Prevention Tool
Let’s first take a look at one of the biggest challenges faced by businesses: creating a full picture of their users based on a few data points.
With your fraud prevention tool, this is something that can be done through data enrichment.
Ideally, you want to make payment, login or sign up as fast and easy as possible. Asking too many questions increases churn and friction, which makes you less competitive. Luckily, it’s possible to only ask a few questions, and get more answers than meets the eye. Here’s an example of what SEON’s email module can do with data enrichment.
- Your user enters their email address and full name
- You run SEON’s email module for data enrichment
- The system checks open source databases and returns info such as:
- Has that email address been blacklisted before?
- Has it been used to register on social media?
- Was it created with a free or disposable domain provider?
- Does the email address show strange strings of letters or numbers that could point to a throwaway address?
- And much more…
Note that the same process can be run with a phone number, resulting in giving you potential details like a user’s profile picture, last login on Whatsapp, and much more.
As you can imagine, data enrichment can be a treasure trove of info, especially when you combine it with….
Predictive Risk Scoring
In fraud prevention, the more data you have, the more accurate the results are. But what are the results exactly? In most cases, it’s about getting a score of how risky the user action is, whether it’s signing up, logging in, or making a purchase.
The risk will be calculated on a number of rules, which you can edit yourself or copy based on the ones most used in your industry. SEON also makes it very easy to create new rules with custom fields and parameters.
However you set your rules, the point is that they should let you know the risks associated, for instance with:
- A set number of login attempts during a time interval
- A user hiding behind a VPN and emulator to purchase an item
- How far the shipping address is from the card billing address
- And much more…
Whitebox Machine Learning Rules
When your business processes hundreds or thousands of users and transactions, it’s near impossible to create one-size-fits-all rules to generate risk scores.
Enter the power of machine learning. By feeding all the data into a ML engine like SEON’s, the system can train itself to recognize unusual behaviour and suspicious patterns, in order to create very specific rules.
Best of all, SEON’s ML rules are all designed to be delivered in human-readable format, using a decision tree. This is the whitebox part of the engine, which allows fraud prevention teams to understand the rules, accept them, and tweak them manually for increased efficiency.
Real Time Results and Dynamic Friction
Of course, even the most accurate risk scores in the world won’t be useful if you have to wait hours for the results. Which is why it’s always primordial to ensure your fraud prevention tool operates fast, if possible as close as possible to real-time.
SEON’s engine, for instance, delivers results under 1 second, which is so fast your users will never know their data is being assessed for fraud checks.
But what if the data points towards a risky user or transaction? Do you just pause the system and potentially lose conversion? Not necessarily.
With the power of dynamic friction, your fraud prevention tool can trigger the following security questions, only when the risk score meets a certain threshold. This is essentially automating the job of manual reviews, so that you can increase security without sacrificing friction.
Interested in More Fraud Prevention Tool Features?
It might be worth witnessing first-hand how SEON fights fraud with a free test of our platform. You will be able to learn all about the potential integrations, modular features, and powerful, flexible products we have created to help you curb fraud in your business today.
Learn more about our products
Bence is the co-founder and COO of SEON whose vision is to create a safer online environment for merchants in high risk verticals.