Ecommerce dominated the holiday shopping season this year, with steady year-over-year growth in both desktop and mobile sales. Cross-border eCommerce also increased in this holiday season, as consumers around the world took advantage of discounts and promotions on Single’s Day, Black Friday, Boxing Day, and Cyber Monday. But while there are a multitude of available stats about the volume of online sales, there is very little information about the rate of fraud during the 2015 holiday season.

In this post, we share Riskified’s observations about card not present (CNP) fraud during the 2015 holiday season. We explain how eCommerce merchants can learn from their holiday season challenges and mistakes, and share best practices for improving fraud management performance going forward.

High Volume = Lower Fraud Rates & More False Declines

A common misconception often peddled by the popular media is that fraudsters target eCommerce merchants during the busy sales season, resulting in higher rates of CNP fraud. The truth, however, is that the influx of legitimate shoppers during sales season “dilutes” the amount of fraud. When measured as a percentage of overall order volume, fraud is significantly lower during the holidays. 

The illustrative graph below is based on Riskified data from 2015. While the graph is not drawn to scale, it clearly shows that fraud rates decrease when sales volumes increase. This trend holds true during the holiday season as well. When order volumes spiked in November and December this year, the rate of fraudulent transactions out of the overall volume was lower:

fraud rates decrease when sales volumes increase

Illustrative graph based on Riskified data from 2015 (not drawn to scale)

Not only do fraud rates decrease when sales volumes spike, but eCommerce merchants also end up declining more legitimate orders due to suspected fraud during the holidays. Whether due to the above mentioned misconception that the holidays are a “risky” time, or due to the overwhelming amount of incoming orders and the heightened time-sensitivity, the fact is that on Black Friday and Cyber Monday we saw a 50% increase in false positive declines. 

False Declines = Lost Consumer Trust, Lower Revenue

Consumers respond strongly when their card is wrongly declined. A recent report by Javelin Strategy found that 66% of US cardholders limited their patronage or completely stopped shopping with a merchant following a false decline. The research showed that 1 of every 6 cardholders (equivalent to 15% of the respondents) experienced a false positive decline in 2014.

Two consumer segments that experience more false positive declines than average are young adults – 24% of whom had their card wrongly declined in 2014, and high-income customers – 22% of whom reported having their card wrongly declined. Young adults, aka millennials, also have a stronger response to false positive declines. In fact, 75% of millennials limited or entirely stopped shopping with a merchant following a false decline.

millennials ecommerce false declines impact

This information is relevant to anyone selling online, but is especially important for merchants selling digital goods, such as e-gift cards and digital travel or event tickets. Digital goods merchants pay a higher price for false positive declines than tangible goods retailers, because falsely rejected cardholders can easily and quickly purchase the same product from a competing site.

How Can You Learn from This Holiday Season?

  • Analyze Declined Orders to Reduce False Positives

    Whether you’re using a fully automated fraud prevention system or manual review, your holiday season approval rates should be higher or at least equal to your average “low season” approval rates. If this is not the case, you almost certainly turned away good customers over the holidays.
    To reduce false positive declines and identify areas for improvement, it is best practice to go back and review the orders declined over the holidays by country or by device type (mobile vs desktop). You now have an entire year to research new data sources that could be incorporated into your fraud prevention process, and also to learn about other solutions that will help you reach more accurate decisions next holiday season.

  • Reduce Turnaround Times to Improve Customer Experience

    Quick turnaround time is especially important for last-minute shoppers, and a less-than-optimal shopping experience can harm your brand. You should talk with the customer service team and ask about the most common complaints they received in order to determine if turnaround times were an issue this holiday season. You can also go to social media to see what people were saying about the shopping experience on your site.
    Often, procedural changes can help reduce review turnaround times. For example, incorporating data such as on-site behavior and browsing patterns into your fraud prevention system may help you reach accurate decisions without reaching out to customers for additional information. If backlogs were created overnight, see how shifts could be better scheduled to prevent this issue from recurring next holiday season.

  • Review Chargebacks to Identify Gaps in Your Process

    Hopefully you will not incur any chargebacks for orders placed during the 2015 holiday season. However, in the likely scenario that you do receive chargeback notices over the next few months, try to treat them as a learning opportunity. As a first step, we recommend grouping chargebacks by reason code. Some chargeback codes, such as “item not as described” and “item not received”, are not necessarily related to fraud. Chargebacks of this type usually have more to do with shipping, marketing, and customer service issues. 

    When it comes to fraud-related chargebacks (“unauthorized card usage”), try mapping them by country, by product type, and by channel (desktop vs mobile). Take the time to go back and review the orders for which these fraudulent chargebacks were incurred. Looking at these transactions in retrospect and without the stress of the holiday rush should allow you to identify important data points that were missed the first time around. If you are unable to identify the fraudulent aspects of the transaction upon second review, you are either dealing with friendly fraud or you may need to upgrade your fraud detection capabilities.


Riskified couldn’t be more excited for the year ahead and is ready to take on the new challenges facing our customers and partners. We love talking about eCommerce fraud prevention. If you have any other insights about fraud this holiday season, or if you have any questions for us, feel free to add a comment below or simply shoot us an email.