November 14, 2018
Guest Post by Melayna Gabiou, Marketing Director, Kount
It’s an exciting time of year for consumers and businesses alike as they shop to give and receive the perfect gift. Celebrating the joy of record sales during the holiday season is what businesses dream of. But what if digital fraud disrupts those sales? Disruption can include chargebacks, false positives, declines, manual review of orders, and everything else that goes into financial loss due to fraud. Provided below is a list of 7 best practices for online businesses to accept online orders with confidence while stopping digital fraud this holiday season.
- ‘Tis the season to plan ahead. As you begin planning your holiday fraud prevention strategy, it’s important to determine your goals and end-to-end strategy. The goal is to not just “reduce fraud”. The goal is to measure all KPIs that play a part in increasing sales while minimizing fraud. What is the projected order volume? How much revenue are we forecasting? How many seasonal hires do we need? How can our good customers have an enjoyable buying experience while fraud attempts are stopped? All of these areas impact how you plan and manage fraud prevention.
- Analyze last year’s holiday data. To successfully forecast this year’s holiday trends, it’s important to look at last and current year data. For example, Kount analyzes holiday data as it relates to fraud each year. Kount looks at the volume of transactions vs. fraud attacks across different platforms throughout November and December. The most prevalent area of growth for merchants last year was the growth of mobile apps. Throughout the entire holiday season, mobile app sales grew 269% while fraud attacks on that platform grew 293%. This data showcases the popularity of mobile commerce and its potential fraud.
- Review and report chargebacks. While chargebacks are a financial pain for the business, they can glean insights on preventing fraud. There are two important components with chargeback data:
- Feedback loop. Make sure chargeback data is entered into your fraud prevention solution. Solutions that employ machine learning will use data characteristics from those chargebacks to refine fraud detection scores. Entering this data will also allow a feedback loop for better reporting on chargebacks so you can understand what fraud looks like for your business.
- Training. Find chargeback examples from the current year and previous holiday season to use as a training resource. Perhaps you find that a certain order amount or location or device characteristic was prevalent with chargebacks, you can then train fraud agents how to uncover that information efficiently before approving or declining orders.
- Plan for seasonal hires. If you have a process to review orders manually, planning for the increase in order review is critical. Hire and train fraud agents well in advance of your peak timeframe. It can take weeks or months for personnel to be fully trained. Look at last year’s trends regarding order volume and peak times of the day. Make sure to understand what promotions marketing is sending, then efficiently staff your team accordingly. Have a senior fraud agent on shift at all times.
- Leverage automation. Automated rules help you work smarter, not harder. Setting risk thresholds unique to your business with the ability to automate decisions is a vital part of holiday preparation. This allows you to efficiently manage anomalies rather than reviewing and approving every digital interaction being analyzed. Machine learning technologies and identity verification will help you speed transaction approvals and boost sales conversions.
- Test before going live. It is recommended to test new policy rules, technologies, or processes before your peak season. Your fraud prevention solution should have a few options for testing automated policy rules:
- Run in a test environment
- Run in a live environment with a “no change” outcome
- Run business intelligence reporting to simulate a rule
Give yourself at least a month before implementing something new in your live environment to ensure the outcome is what you anticipated.
- Implement a code freeze. Problems reveal themselves at double or triple the amount of order volume you typically manage. Therefore, unpredicted issues may happen during the holidays. Determine a final date for adjusting code or technologies that are involved in your order processing. Kount employs a code freeze on October 31. In order to meet that code freeze with success, you need to be begin planning well in advance – often in mid-October or earlier.
All of these best practices need to be vetted, tested, and implemented with enough time before peak days. To learn more about how to dial-in your fraud prevention strategy before the influx of orders watch the on-demand webinar, How to Prepare for Digital Fraud This Holiday Season, featuring LUSH Cosmetics and Whitepages Pro.