Using Big Data to Avoid Ecommerce Fraud

 recent report estimated that the fraud cost for internet retailers to be $3.5 billion or 0.9 percent of online revenue. To put it in different words, an internet retailer loses, on average, $9,000 to fraud for each $1 million in earnings.

This is a significant quantity of fraud which, ideally, should be a part of the retailer’s profits rather than being added into the cost of doing business. Alas, the fraud rate rises if the merchant supports mobile ships or commerce orders to international customers.

Every merchant strives for fraud losses but that goal has been elusive. With the development of Big Data in the past few years — visit “Recognizing Big Data For Ecommerce,” our previous post — the objective is now within reach. Big Data can protect against fraud by identifying insights which have never been readily accessible to online retailers before. Before introducing how Big Data can help, I will first examine the important kinds of fraud which impacts online retailers.

  • Charge card fraud. This is the most frequent fraud which impacts online retailers. Since the retailer is not able to find the card , thieves can more readily use a stolen or fake credit card. Many retailers use the resources offered by credit card companies such as address verification service, which matches the card number together with the billing address on the card, or the three-or-four digit code on the card, known as the card verification number or card security code, to minimize fraud. But these methods aren’t foolproof.
  • Yield fraud. This takes many diverse forms. The most common are returning merchandize after using it claiming the item wasn’t delivered and selling it via other channels. A few online retailers encourage customers to purchase more goods and return those they do not want. Some customers assume that each and every retailer supports this policy, resulting in higher return fraud. Online retailers that sell high-value items or have a brick-and-mortar shop are more vulnerable to this sort of fraud. Even after taking steps like charging restocking fees for returns, obtaining customer signatures to verify order delivery, and tracking customers that are most likely to perpetrate fraud, return fraud persists.
  • Identity fraud. Identity fraud involves stealing a customer’s personal information: A fraudster logs in, with a customer’s qualifications, stores on the website, and ships the merchandise to another site. It affects retailers heavily as, typically, it contributes to chargebacks and the merchants must pay the chargeback punishment unless they could prove the transaction was a fraud. The fraud may involve having a customer’s personal information to get new credit cards. Identity fraud could lead to larges losses. Having the correct rules in place to assess purchasing patterns of every client, his frequency of ordering, and his purchase shipping addresses is critical. But those measures don’t eliminate it entirely.
READ  Ecommerce Merchants Embrace Brick-and-mortar

Preventing Fraud with Big Data

The usage of Big Data can combat fraud, in three primary ways.

  1. Examine all the data. Previously, retailers used a sample or a subset of the data for fraud investigation. It took too much time and money to use the whole data set. With Big Data, not only does all the information be examined for fraud, but new data sources may also be introduced. Analyzing the entire data set contributes to several advantages: (a) Reviewing all transactions based on the specified fraud rules; (b) Identifying new fraud patterns that get added to the growing list of fraud principles; (c) Minimizing false positives to avoid losing revenue and turning off clients.

    By way of instance, to minimize return fraud, large data can ascertain if the item was actually delivered by assessing data flows from social networks, conducting picture analyses, and cooperating with third parties such as eBay and Craigslist to get their listings. All this information can be aggregated and analyzed using Big Data tools such as Hadoop.

  2. Detect fraud in real time. Live transactions are combined with information from other sources, like existing data warehouses, to detect fraud in real time. This can prevent credit card fraud in which the transaction is screened from a set of pre-defined fraud rules as part of their credit card authorization. Including combining site data with information from client’s social feed, the geo information from client’s smartphone apps, purchase history, and web logs. Since this is done in real time, fraudulent authorizations are diminished. Additionally, Big Data solutions enable assessing historical transactions in the prior weeks, months, or years to identify new fraud patterns, which is automatically added to the set of fraud rules and utilized as part of their real time authorization procedure. Visa recently reported identifying $2 billion in potential annual incremental fraud opportunities by utilizing a strong fraud management system that looks at 500 distinct aspects of a transaction to detect and protect against fraud.

    Another way real-time evaluation prevents fraud is by processing streaming data from sensors attached to high-value things to transmit their place. This reduces return fraud, since the merchant now knows precisely when the product was delivered to the client.

  3. Utilize visual analytics. Other Big Data tools provide the ability to visually analyze data and derive insights, though the data could be coming from various sources. Retailers can use these tools to identify areas, products, and clients with a greater fraud rate based on historical analysis. This explains the areas where money and time should be spent to minimize fraud. Visualization also reduces manual attempts to reviewing every purchase. The accounts can graphically depict the likelihood of fraud for every purchase transaction and connect to email or SMS alerts for escalation, as required.

See also: