Most recently we have been asked by a large online fashion retailer to help them handle a continuous fraud attack, where more than 1000 orders worth $85K were sent to a specific industrial zone in Florida. Most of these were not identified by the Shopify Risk prediction model.

Prior to contacting us, the customer implemented a first aid measure, putting in place a basic rule engine blocking all orders shipped to that specific city. Our first action was to have our risk experts run a simulation recreating the last month’s orders, one by one, allowing FUGU’s unique velocity checks to identify the root fraud pattern in this specific case. (The importance of velocity checks)

Turns out the root cause was the heavy use of a series of debit cards issued with specific BINS by a certain issuer.    We immediately added this insight into our post-payment risk prediction model, automatically blocking all orders shipped to Florida using that specific credit card type.

We also noticed that several other credit card types were used to send merchandise to the same PO Box but they seemed to be 100% valid.

We then suggested a quick and easy fix: to deploy a fulfillment process change so that their superfast fulfilment process would be separated into two routes:

      A.     Fast lane covering 95% of the cases categorized by FUGU initial risk scoring as low risk

      B.     Delayed delivery lane covering 5% of the cases in which FUGU identified as  high risk.

This method allows FUGU’s platform to profile the transaction and distinguish between fraud and valid transactions sent to a PO BOX for valid reasons.

And we found quite a few of those:

Our findings were stunning ,  turns out this specific brand has a large community of fans in Honduras and Columbia that ship to a logistic warehouse In order to save the international shipping costs

We also found several very loyal customers apparently living in the same city that was previously blocked.

All in all we were able to recover 50% of the transactions previously discriminated against based on their address.

The funny thing is that around 60 days after the client began using FUGU the Shopify fraud recommendation started recognizing what FUGU recognized right away, but still with around 50% false declines.

As we Say in FUGU, Every Payments Counts, in this case ruling out shipments to  PO Boxes is easy but leaves disappointed customers unattended. It  simply makes perfect sense to try to verify these transactions rather than simply saying NO. Our automated post-payment transaction verification does exactly that.

FUGU is a new breed of payment fraud solution tracking payments post-checkout, helping merchants safely accept transactions they currently lose to fraud, false declines, and payment churn.

FUGU offers a multi-tier self-learning fraud prevention solution, fighting a wide variety of risk patterns at various points along the transaction life cycle using automatic customer verifications (KYC), pre-shipment consumer behavior scoring and chargeback representations.