- Rollout across all global markets in weeks
- 10% improvement in acceptance with no negative impact on chargebacks rates
- Immediate ROI return in and large reduction in false positives across the globe
FlixBus was founded in 2015 to take advantage of the deregulation of the German rail and bus network. Since then it has expanded to connect over 2000 destinations across 28 countries in Europe and now in the USA. This is over 350,000 connections per day where FlixBus manages the pricing, ticketing and booking for its vast network of buses often supplied by partners. Wherever you go in Europe you are sure to see the green FlixBus livery.
FlixBus’ ticketing goal is to make it as simple as possible for a new or returning visitor to purchase a ticket. To do this the company supports a plethora of payment methods types from cards to wallets, Paypal to direct payments. FlixBus also operates in multiple geographies with a variety of fraud risk profiles.
FlixBus wanted to be able to manage the payment risk in a consistent manner across all of these markets and payment methods. It wanted to do this with no negative impact on the low chargeback rate that it had worked hard to achieve. It also wanted to implement 3DS dynamically, reflecting the tolerance of different markets to the protocol.
After a search FlixBus selected Ravelin as the fraud detection solution.
“To operate at the scale and speed we wanted we knew we would need the automation that a machine learning solution like Ravelin would provide. However what really impressed us was the patience and time that the team at Ravelin took to understand our business.” Dennis Friemerding, responsible for Payments, Fraud & Risk Management, FlixBus
An implementation phase supported by Ravelin’s integration team began in earnest in late 2018 and with a go live at the end of the year. Even working to this rapid pace both companies were able to deliver a rock-solid integration and start to use Ravelin’s recommendations for approving, rejecting and reviewing transactions. Dennis explained:
“We were impressed to say the least with the technical capabilities of the Ravelin team but also their ability to quickly understand our risk profile. For instance we got immediate value from the initial Machine Learning models that they deployed. In the background they were building a model specifically for us based on our data and we were able to get even better performance when we switched after a few weeks to that.”
"Ravelin’s Investigations team was able to provide insight and analysis on what was happening in our customer database. We were immediately able to investigate Ravelin Connect to see suspicious connections in our data. We were honestly and pleasantly surprised to get so much value so quickly”
FlixBus went live on Ravelin initially across its key DACH region, the Nordics and in the US. It is now live in every FlixBus market. The business has seen an immediate and dramatic 10% improvement in acceptance with no increase in chargebacks. This is a significant reduction in the false positive rate the previous rules-based system was causing which resulted in FlixBus and its partners rejecting many good customers.
This hits on a key deliverable to ensure the company and its partners are able to accept more business and is already a significant return on investment.
Concluded Dennis Friemerding: “We’re only at the start of our story with Ravelin but the initial impact has been very encouraging. As we develop a more thorough understanding of the technology and Ravelin of or business we look forward to working on our growth story together.”