Preparing for the Holiday Season
October 28, 2015
The holidays are meant to be a relaxing period for employees to take some time off from work and enjoy it with family. Unfortunately in the retail industry, and especially for fraud prevention teams, there is no downtime.
Fraud prevention professionals must buckle up for a busy six weeks. Order volumes increase anywhere from three to 10 times normal volumes. With this spike in order volume comes an increase in the amount of fraud attempts. What can fraud prevention teams do to prepare for the holiday rush? Here are few suggestions:
Create a holiday fraud model or rule set to adjust for the high volume of orders.
If possible, look at last year and see what the spending pattern is for your customers. How many items are being purchased in their shopping cart? What items are being purchased? What is their chosen shipping method? What are the busy days for virtual gift card purchases? Your holiday model needs to be adjusted or you could potentially auto reject good orders or even worse, create a difficult to manage review queue.
Become your marketing team’s best friend.
Try to find out in advance all the marketing campaigns that will drive high order volumes. Campaigns might include product sales, promotions for a free gift card with purchase, or fulfillment guarantees (e.g., buy now and deliver by Christmas).
Use positive lists to identify legitimate repeat customers.
It’s important to look at a combination of data fields in your white lists to prevent account takeover fraud. Look at email addresses and shipping addresses combined together (e.g., Grandma always sends gifts to her grandkids).
Create a virtual gift card fraud prevention strategy.
The highest risk product sales during the holiday season are typically gift cards. Perform a daily analysis of the average gift card purchase amount and then adjust risk based on the average you see each day. Look at as much data as possible: recipient email addresses, gift card messages, velocity of orders in a short period of time. Be careful not to reject a transaction too quickly, though; for example, it’s common to see large corporate orders at this time of year.
While the suggestions above are about preparation, remember to also (and most importantly) take care of your fraud prevention teams. They are your eyes and ears, and provide valuable feedback on the fraud being perpetrated, as well as good orders that might be held up in manual review. This feedback loop can be done quicker than any machine learning model.
About the Author
Mark has been with Accertify since 2008, and is the fearless leader of Accertify's Professional Services, Account Management and Managed Services teams.