Layer 1 copy

It’s very important to focus on the core of your product and minimize the time spent on stuff that doesn’t add value to the user. Having a reliable and sustainable solution is crucial, so I don’t have to worry about any of that and my team can focus on the bigger picture.

Simon KreuzCo-Founder

Overview

  • Software layer between shippers and e-commerce stores

  • Available via API or web interface

Challenge

  • Billing occurs after multiple labels have already been created

  • Preventing fraud before it occurs

Solution

  • Easy integration with Sift Science

  • Real-time fraudster identification for chargeback prevention

Results

  • 0 Full-time people dedicated to fraud

  • 0 False positives

Overview

Making Shipping Easy

Shippo is a B2B application that empowers merchants to ship orders through a variety of carriers. Founded in 2013, Shippo’s users range from small standalone retailers to chains and consumer-to-consumer shops. Through Shippo, a merchant can easily print shipping labels and track orders. Bigger customers even use the Shippo API to automate their checkout and order processing flow.

The Shippo team is focused on ensuring a positive merchant experience. In order to keep the flow as seamless as possible, Shippo doesn’t charge for individual shipping labels. Rather, every user is only invoiced after they exceed a certain order threshold. For a threshold of $25, for example, a merchant can create multiple labels before having to pay. This way, the merchant’s time spent paying is limited, and they can test out the product for free. To sign up for Shippo, a user just needs to provide a credit card number – that won’t get charged until that first invoice is sent – and email address. Unfortunately, though, this frictionless merchant experience means that Shippo is on the hook for paying for the shipping labels until the invoice is sent and paid.

Challenge

Account Abuse and After-the-Fact Billing = Chargebacks

Early on, technical cofounder Simon Kreuz – who built Shippo’s first API – was able to combat any fraud with an internal fraud prevention tool. The Shippo team found that fraud was generally easy to identify because fraudsters followed a pattern, which allowed them to build their own solution. But as time went on and Shippo’s business grew, the fraudsters hitting the app got more wily.

Shippo found that the majority their fraud fell into one of two categories:

1. Users who sign up with a fake email address and use a stolen credit card number.

2. Users who create labels, hit the threshold, and then create a new account to avoid paying their invoices.

In both cases, Shippo lost money – whether from chargebacks from the accounts with stolen credit card numbers or lost revenue from the unpaid invoices. Shippo needed a solution that could preemptively identify account abuse and prevent users with stolen credentials from purchasing; after a label is generated and used, it’s already too late. It soon became apparent to Simon that his team’s time was better spent growing the business rather than updating their internal tool, and he turned to Sift Science to stop Shippo’s fraud problem.

Solution

Preventing Fraud by Blocking Fraudsters

As he began evaluating Sift Science, Simon was mindful of his developer resources. He was pleasantly surprised to see that Shippo was able to integrate with Sift Science in just a few hours, and the whole process was smooth and relatively effortless. Of particular use to his team was Sift’s ability to spot and troubleshoot integration errors easily.

Shippo began by sending Sift Science as much historical data as possible and relying on the machine learning technology to make sense of this dense information. Using Sift Science, Shippo could quickly separate out suspicious users from legitimate ones, and focus their energies on validating accounts and orders. Right away, Simon’s team saw Sift Science effectively catch fraudsters. Perhaps most valuable was Sift’s knack for identifying bad users at the account-creation stage. Shippo no longer had to wait until they saw a chargeback or a pile-up of invoices before spotting malicious users.

“It couldn't have been better timing: about 3 hours after the rollout, Sift Science in fact did correctly identify a new fraudster who just signed up! I honestly couldn't believe it at first.”

Results

Focusing on the Bigger Picture

With Sift Science in place, the Shippo team has seen a dramatic drop in the number of bad accounts and chargebacks. In fact, Sift Science’s fraud prevention has been so accurate that Shippo has a 0% false positive rate on flagged orders.

“With Sift, we don’t have any false positives. As far as I can tell, all of the fraud that is flagged by Sift is real fraud.”

With an effective fraud solution in place, Simon and the Shippo team have more time to focus on growing the business. Currently, there are zero employees dedicated to fraud prevention at Shippo. Instead, everyone works to add value and create an exceptional experience for their users.