Machine learning (ML) underpins many fraud prevention tools on the market today, but only Decision Manager's machine learning is powered by Visa's sophisticated artificial intelligence (AI) platform and enriched through access to 141 billion VisaNet transactions1. So how could that help your business prevent fraud?
The key to successful fraud prevention
Across the Visa ecosystem, 98.7%2 of Decision Manager customers are confident enough to rely on its automated fraud-decisioning strategies without utilizing manual review.
The accuracy of Decision Manager's automated risk detection allows our customers to maintain industry-leading rates for chargebacks—a critical capability, given that reducing chargebacks is one key way your business can manage fraud3. It also helps businesses optimize revenue by allowing them to:
- Accept more good transactions and reduce false positives
- Streamline the payment experience for genuine customers
This high degree of successful automation is due to Decision Manager's unique combination of best-in-class machine learning and a flexible strategy engine.
Best-in-class machine learning
ML technology has been an essential element of our fraud and risk solutions since our launch more than two decades ago. We've been expanding our ML technologies and expertise ever since.
Unlike any other ML-based fraud prevention solution, Decision Manager's machine learning is powered by Visa's AI platform and enriched through access to 141 billion VisaNet transactions.1 The sheer size of Visa’s dataset and our ecosystem's global reach, advanced capabilities, and 99.9% uptime4 combine to deliver a best-in-class capability that enables average transaction screening times of less than 600ms.1
On top of that, Decision Manager's ML-driven Identity Behavior Analysis leverages Visa data to recognize:
- Legitimate repeat customers
- New customers who have no history with you
to further boost order acceptance rates and reduce false positives.
Flexible strategy engine
Decision Manager integrates ML with a flexible strategy engine that can be configured to meet your business’ unique needs. Building a strategy is as easy as setting thresholds for risk and choosing filters that align with your specific requirements, with no coding required, all in real time. To help you finesse your fraud strategy, Decision Manager incorporates:
- The ability to perform a 'what if?' analysis to quickly test the effectiveness of new fraud strategies against your historical transaction data and assess how they would play out in the real world
- An automated Rules Suggestion Engine, which applies ML models to your historical transaction data to spot patterns and recommend new strategies
- Data consolidation across regions to further enhance risk-scoring accuracy and identify customer spending trends and emerging fraud patterns
Mitigating risk across the ecosystem
As a Visa-owned fraud solution, Decision Manager is fully integrated with other Visa systems, data, and analytics. Its development is supported by 20+ years of insights and our deep relationships with acquirers, partners, and customers.
Businesses worldwide use Decision Manager to take the guesswork out of fraud prevention, which helps mitigate risk throughout the payments ecosystem. See how Decision Manager’s powerful ML foundation could help your business reduce fraud, improve acceptance, and lower operational costs.
1 VisaNet transaction volume based on 2021 fiscal year. Domestically routed transactions may not hit VisaNet.
2 Based on average data collected from all clients on the Decision Manager platform between 10/1/21 – 9/30/22.
3 "Global Fraud and Payments Survey Report 2022," p. 7, Cybersource, Merchant Risk Council (MRC) and Verifi.
4 Data is measured and validated from internal instance of Tableau Server based on payment volume from the Cybersource and Authorize.net Product Fact data source. Provided by payment processing product team. Data is measured and validated from internal instance of Tableau Server based on billable transactions from the Cybersource and Authorize.net Transaction Fact data sources.