Solutions overview
Harness the power of your data to reduce fraud and increase payment acceptance
Tailor-made fraud protection
Detect and stop fraud faster with clear insights
Adaptive solutions for emerging threats
Defend against ATO, promo abuse and seller fraud
Optimize conversion with agnostic authentication
Manage PSD2 and take control of authentication
Online payment fraud
Understand chargebacks, fees & detection
Machine learning for fraud detection
Models, risk scores & thresholds
Link analysis & graph networks
Draw deeper insights from data
Account takeover fraud
Prevention strategies & reputational risk
Policy abuse
Uncover & stop hidden costs
PSD2 & SCA
3D Secure, TRA & exemptions
Global payment regulation map 2022
Track PSD2 & more with a full report
Resource Zone
Deep dives on fraud & payments topics
Blog
The latest fraud & payments updates
API & developer docs
APIs, glossary, guides, libraries and SDKs
About Ravelin
Discover the story about Ravelin
Careers
Join our dynamic team
Customers
Read more about our happy customers
Partners
Join our partner programme
Uncover & stop hidden abuse
Resource zone
Read more about our happy custmomers
Blog / Link Analysis
Marketplaces are uniquely at risk of fraud from two angles: customers and suppliers. Learn how to analyze patterns in your vendor data to see the bigger picture and stop fraud...
Share this article:
Online marketplaces connect customers with suppliers, offering convenient apps or websites like Uber or Deliveroo. A customer places an order, and the marketplace allocates it to their most suitable driver (supplier). Marketplaces are incredibly popular business models, but they are uniquely at risk of fraud from two angles: customers and suppliers.
In recent years, supplier fraud (also known as courier or vendor fraud) has made the headlines. Titles like “1 in 4 delivery drivers admit to eating your food,” and “customers accuse drivers of stealing their groceries” appeared everywhere.
These headlines may seem harmless, but supplier fraud can cause marketplaces eye-watering hidden costs. Plus, it can damage brand reputation which may deter customers and suppliers. Supplier fraud isn’t new, but it is a growing problem.
Let’s take a look at why supplier fraud is a challenge for fraud teams, and how graph networks can give you better insight into the problem…
The Covid-19 ecommerce boom created turbulence for suppliers. Food delivery drivers were inundated with jobs during lockdowns, but ride-sharing drivers struggled to get work. These extremes may have increased the likelihood of opportunistic supplier fraud - drivers might feel overworked/entitled to more money or just in need of some extra cash.
Plus, many marketplaces made incentives more generous in light of the pandemic, either to meet high demand or to convince drivers to come back to work. The more generous the incentive, the more lucrative it is for suppliers to try their luck.
Supplier fraud is difficult to tackle as the last thing a marketplace wants to do is limit their pool of drivers. Marketplaces need expansive networks of suppliers to grow the business - if they blocked every good supplier that abused a policy, it would stop functioning. A ride-sharing app can’t offer rides to the whole of New York with only 20 drivers!
Plus, blocking just one supplier could impact multiple customers and give them a worse experience - longer wait times, higher charges or other disappointments (like a cold meal.)
So how can you stop policy abuse without losing suppliers? Fraud teams have to strike a difficult balance: discourage bad behavior but block as few suppliers as possible. And you don’t want to come down too harshly on suppliers as they could easily just take their services to another app or marketplace.
It’s hard to measure your success at preventing marketplace supplier fraud as there’s no easy metric. Most supplier fraud scenarios won’t result in a chargeback - a supplier using a fake customer account to ramp up their business wouldn’t file a chargeback against themselves! And since marketplaces make refunds and customer services easy, genuine customers are more likely to complain to them directly than file a chargeback.
This means fraud teams often have to rely on anecdotal evidence, customer complaints, and manual review, which becomes less viable as your business grows and you get more suppliers. But if you don’t understand the connections between your suppliers and customers, you can never truly be in control of the situation...
Graph networks are a key tool for managing marketplace fraud, making it easy to evaluate connections between your data. In a graph network, all the information about a customer’s account, email, shipping address, order details and payment information is connected and visible at the same time.
To make life easier, we’ve added suppliers to the graph network. Adding suppliers into the mix makes suspicious connections and patterns clear, helping you assess a supplier’s risk level in the bigger picture. This could help you in a number of ways...
If you can easily see suppliers in the graph network, you can analyze...
Suspicious patterns: Unusual patterns in order times and locations can indicate fraud, like if a driver only takes journeys from the same location at the same time every week.
Supplier order history: Assessing a supplier’s order history can flag suspicious repeat orders from the same customers or locations. A single cab driver and single customer in New York are very unlikely to be allocated to the journey three times in a month!
Supplier-to-customer links: Why would a supplier have the same details as a customer? There’s no genuine reason. If you spot a driver with the same phone number as their most regular customer, it’s a huge red flag.
Supplier-supplier behavior patterns: Suspicious patterns in behavior could indicate that suppliers are communicating with each other to create money-making opportunities. Drivers have been known to tell other local suppliers to turn off their availability so the marketplace will implement a temporary rise in pay.
Supplier-supplier links: Suppliers with the same email addresses or phone numbers could be involved in sophisticated fraud networks.
Once you have all of the data in front of you, you can find targeted ways to discourage opportunistic policy abuse and only block the really prolific fraudsters.
If you’ve spotted a supplier prone to opportunistic incentive fraud, you can contact them with an alert or email to discourage the behavior. Or, you can intervene in supplier allocation to prevent high-risk suppliers from being matched with high-risk customers/orders. This approach means the supplier can still be assigned low-risk orders - you don’t lose a supplier and they can still keep earning.
In extreme cases, you can block a supplier from your network completely - if they are a known prolific fraudster and don’t do any genuine orders. If you choose to ban them, graph networks enable you to make sure they don’t return. If a new supplier with a similar email or phone number ever appears, you can prevent them from rejoining your platform.
For more insight into how to stop fraud using graph networks, watch our webinar: how to use network analysis for fraud detection.
Grace Proctor, Content Writer
Blog / News
You might’ve heard about a new fraud tactic called ‘account pre-hijacking.’ But what actually is it? And how can you prevent it? Let’s break it down.
If you offer a subscription, recurring payments can leave you vulnerable to unique fraud risks. How should you tailor your fraud strategy?
Buy now, pay later is exploding - what risks could this bring your business? We speak with Nelda Biltauere, Fraud Researcher at Ravelin, about BNPL challenges, costs & strategy.