Social media data is rapidly becoming a more useful tool for ecommerce fraud prevention around the globe. The most technologically advanced fraud prevention solutions use machine learning in combination with social media data to determine if a transaction is fraudulent.
For example, an automated solution might see what social media networks a person is connected to and/or how many connections they have to help determine fraud risk. However, most of these machine learning fraud solutions are far from being mature. For most merchants, the best use of social media data is still for manual review by fraud analysts.
North American fraud analysts often approve a transaction by verifying the purchaser’s identity using multiple points of personal information – including social media accounts. Fraud analysts everywhere use people search engine tools like Pipl Pro to confirm an identity by reverse-searching the email address and/or phone number associated with an online order. What differs in the U.S. and Canada is that fraud analysts will also check to see if the person and email address being searched are connected to social media accounts on networks like Facebook, Twitter and Instagram.
Using Pipl’s search engine, fraud analysts can click on the social media account links to check out the customer’s actual social media account and whatever public-facing postings they’ve made. It only takes a few seconds to see if the person is connected to social media accounts and, if so, if the accounts have long been in use and have human generated or bot-like postings.
Beyond general identity verification, there are specific instances when social data will be the most effective type of information for deciding whether to approve or deny a transaction.
One of the first uses of social media accounts for fraud analysts was to reduce customer insults during trips for business and vacations. It used to be that the credit card owners would have to notify their credit card company ahead of any overseas trip or seriously risk having all their transactions in Paris or Rome flagged and blocked for fraud. Those days are largely over thanks in part to people’s habit of publicly announcing their trips on social media. Now fraud analysts can often determine if an overseas or out-of-area charges are due to a client visiting somewhere unusual or whether they are still at home and their card has been stolen.
False fraud claims perpetrated by actual customers or those close to them (AKA “friendly fraud”) can be very hard to catch. However, one way to detect if the goods you shipped actually reached their intended destination is through social media.
Using Pipl Pro, you can go to the social media profile of the person listed on the order and, if they are connected on social media, jump to the social media profiles of their significant other and/or close family members. If the customer, their child or their significant other really were behind the purchase – one they decided was important enough to break some laws – there’s a good chance they will be showing off their new purchase on one of their social media accounts.
Social media is even more useful for loss prevention and fraud investigations than simple yes/no answers on ecommerce transactions. Organized crime groups have targeted brick-and-mortar retail establishments with shoplifting teams for years. Often it will be the job of one gang member to distract shop staff while at least one other gang member starts pocketing goods. The prevalence of omnichannel retail has taken organized crime group tactics to new levels. Now, there are cases where a team member orders goods online for in-store pickup using a fraudulent credit card while another team member is waiting in-store to pick up the goods as the order goes through before there is enough time to review the purchase.
Even when loss prevention personnel and local police identify one of the fraud perpetrators, finding the other group members and taking them off the street can be quite difficult. However, the moment you have identified one crime group member finding the others can be as simple as spending a few hours studying their social media connections, store footage and other details to which you already have access.
Crime groups, like everybody else in the 21st century, are increasingly organized online as well as in the real world. Moreover, despite what you hear in the news about the “dark web” and mysterious criminal hackers, much criminal activity is coordinated via social media networks open to one and all. Finding an entire shoplifting street crew can be a simple as taking a look at a confirmed criminal’s social media profiles and looking for which friends or connections post frequently on their profile, what online groups they belong to and searching with a hawk’s eye for any posts using criminal code or slang. Actual cases have been cracked this way.
Social media networks have rapidly expanded over much of the globe over the past decade. Most social media users no longer reside in the U.S. and the user base is no longer restricted to developed countries. There are plenty of social media networks that are popular in Eastern Europe, Brazil, India and elsewhere and which have accounts tied to a person’s real-world identity via Pipl’s search engine.
North American merchants and fraud prevention solution vendors have been using social media data in some form or another to prevent ecommerce fraud since at least the start of this decade. However, most fraud analysts in European and fast-growing ecommerce markets (such as India) have yet to take advantage of social data for fraud prevention even while they use these social networks in their personal and business lives. Try to be ahead of the curve, not behind it.
Original blog post written by former Pipl Technology Evangelist Ronen Shnidman. Ronen is now Managing Editor @ about-fraud.com