Rich data & powerful ML: Flix and Greyhound’s revenue boost from better fraud prevention

A breakout tech business that has gone truly global, Flix is a German long-distance bus and train service provider that serves 81 million+ passengers a year.

Launched in 2013 and growing organically every year, Flix is active in more than 44 countries in Europe, North and South America, and Asia-Pacific.

First and foremost a travel tech company, Flix doesn't typically own bus fleets; they collaborate with local providers of bus and coach services to serve passengers across an expansive network, enabling them to reach more passengers than before.

Flix also brings everything together digitally for the providers’ convenience, including ticketing, promotion, and payments.

With such impressive growth and reach, Flix has put fraud prevention at the forefront of their strategy:

“With a growing global footprint and millions of digital transactions processed each year, we’ve made proactive fraud detection and prevention a key pillar of our platform security,” explains Krystyna Savotchenko, Fraud Team Lead at Flix.

“Our goal is to stay one step ahead, continuously investing in technology and expertise to protect our passengers and partners.”

In 2021, Flix acquired the iconic American intercity bus operator Greyhound Lines – a key milestone in their expansion. As the largest intercity bus service in North America, Greyhound Lines operates an impressive fleet of 1,700 motorcoaches.

Krystyna, along with her six-person team, governs Flix's comprehensive fraud strategy, leveraging their expert insight and Ravelin’s bespoke solutions to protect the business.

“In our experience, travel introduces an additional level of complexity for fraud prevention. We already have people moving from point A to B. Some people are using VPNs, and some people are providing non-functional phone numbers either intentionally or by mistake.”

Preventing deliberate fraud on top of this can be a complex challenge – one that Flix addresses with the help of Ravelin’s machine learning solutions, enriched with identity data from Mastercard.

The key performance indicators (KPIs) for this strategy are aligned with Flix’s commitment to customer trust and seamless experience: aiming to minimize chargeback rates while maximizing genuine customer approvals in each market.

Ravelin’s core philosophy is to custom-build AI-native solutions for each merchant, recognizing that every fraud profile is unique. This strategy aligns perfectly with Flix’s goals, as their specific risk landscape requires a bespoke defense that generic, one-size-fits-all tools cannot provide.

To achieve this, Ravelin’s Engineers and Data Scientists use a wealth of data, including:

• Merchants’ own data – both real-time and historical

• Consortium network data anonymized from Ravelin’s client base

• Enriched identity signals from Mastercard for fraud detection

This rich dataset is used to build advanced machine learning features that put the numbers into perspective such as analyzing velocity and graph network link analysis for deep context.

AI-native Ravelin then utilizes a multi-model, continuous deployment philosophy, ensuring the system always prioritizes the ML features that bring superior fraud detection results.

Ravelin enhances its fraud risk scoring by integrating with Mastercard’s global identity data. This unique capability provides Flix and Greyhound with the confidence they need when serving customers online.

Mastercard enriches our existing data sets with data that has been vetted, normalized and verified at local, regional and global levels. Mastercard analyzes billions of patterns of how identity data is being used in digital interactions over time to detect patterns and identify risk.

Ravelin utilizes this powerful data stream to enrich customer profiles, analyzing identity signals to help calculate certain ML features.

The result is unparalleled fraud detection accuracy and fewer chargebacks, giving merchants a better understanding of their customers’ true intentions without the heavy lifting of complex data integration.

By leveraging Ravelin’s internal ML features and Mastercard’s identity data, Flix is better positioned to increase uplift and grow securely. The ultimate goal is to enable Flix to accept more payments with confidence and serve customers better.

"The integration of Mastercard's identity enrichment with Ravelin has led to improvements in both acceptance and dispute rates across our markets. This dual benefit has not only bolstered our monthly revenue but also reduced the time and resources spent on managing disputes.” – Krystyna Savotchenko

Flix achieved a major success: by accepting more payments with confidence, they enjoyed substantial unlocked revenue, with incremental revenue uplift reaching impressive new levels in four months. Critically, this success did not incur the typical risk of increased chargebacks. Dispute rates actually decreased for both Flix and Greyhound - meaning fewer instances of fraud despite the higher acceptance volume. This two-fold success solidified their trust in Ravelin’s capabilities:

"This success has reinforced our trust in Ravelin's solutions and their capability to provide consistent performance with their partners." – Krystyna Savotchenko

Greyhound’s partnership with Ravelin and Mastercard has enhanced their fraud prevention strategy, driving increased uplift and supported growth in regional markets.

The notable decrease in chargeback rate achieved in the United States since updating the Greyhound model with Mastercard identity data is a direct result of Ravelin's technical synthesis. Success relies on combining Ravelin's bespoke machine learning, link analysis, and Mastercard's global identity data for industry-leading fraud detection capabilities.

“Integrating Mastercard through Ravelin has further strengthened our fraud management framework. With a substantial reduction in dispute rates, we have gained increased confidence and improved performance in our key markets." – Krystyna Savotchenko

For the forward-thinking travel tech platform Flix and the operator Greyhound, the complexity of payments, trips, and markets is a given. The risk landscape can be dynamic and demands adaptable and resilient technology to support fraud prevention efforts.

Automation, custom ML models, and Mastercard identity data delivered impressive results for both companies, including a 40.7% uplift in acceptance for Flix and a substantial drop in dispute rates for Flix (14.9%) and Greyhound (19.3%)*. Both companies have successfully increased approvals and expanded successful transaction volume while streamlining the time and resources spent on dispute management.

By partnering with Ravelin and Mastercard, you can enjoy the results of a personalized approach to fraud prevention that prioritizes what your company cares about and makes the most of your data – ultimately boosting revenue and powering your secure growth.

*All results in the article sourced from internal Ravelin data, confirmed by Flix.

• Flixbus: Tackling CNP fraud during hypergrowth (video from RavCon 24)

• Mastercard and Ravelin partner to reduce fraud and create frictionless checkout in digital quick commerce

• AI fraud detection with ML & NLP: How Ravelin does it

• How do we build a machine learning model?

• Webinar: How River Island works smarter, not harder, with machine learning