While the volume of e-commerce and card-not-present (CNP) transactions continue to grow year over year, so does the specter of payment fraud. Protecting digital payments is paramount—as the true cost of payment fraud is far greater than you might think: almost $4 in remediation expenses for every $1 of fraud.
However, it’s not just about saying no. Because inaccurate stock methods can cause revenue leakage of up to 10%, it’s also about accuracy and the ability to approve more transactions. With 25 years of data, analysis, and model refinement, Vesta helps businesses approve more transactions, while still leaving fraud out in the cold.
Improve Accuracy
Inaccurate fraud detection often lets bad transactions through while declining good ones. Vesta’s provides accurate assessment with key insights to help you approve more transactions
Minimize Cost
Gain insights from Vesta’s 25 years of fraud expertise that help you stop payment fraud based chargebacks and stop wasting time and resources on chargeback remediation.
Enable Smart Authentication
Vesta gains insights from transparent data collection such as the user’s known device to make accurate predictions on transactions without adding friction to the the user.
How Payment Protect Works
See exactly how Vesta detects payment fraud, approves more transactions, and guarantees that you stay fraud free.
STEP ONE
Transaction
When a customer checks out, the relevant details about the purchase are sent to Vesta. This is passed via an open API framework that works with a wide range of shopping cart plugins and sends key data sets to the decisioning engine for analysis.
Evaluation Information Includes
- The order information
- Payment information
- Card holder details
- Session key with behavior attributes and device information
STEP TWO
Analysis
Once the transaction data is in the decisioning engine, supervised and unsupervised machine learning models built on two trillion data points evaluate the transaction and determine risk. Each component is evaluated using velocity checks, white and black lists, and stepwise regression.
Evaluation Information Includes
- Phone numbers
- Addresses
- Device Information
- Geolocation
- Zip Codes
- Over 292 million credit card numbers
- And more
STEP THREE
Deep Learning
Over time, Vesta collects more and more indicators of payment fraud—adding to, and refining, a collection of over 25 years of intelligence. Over time, this data increases accuracy, and improves machine learning models.
STEP FOUR
Model Selection
During analysis, the decisioning engine may self-select a different model to use in an effort to gain better risk profiling. Once everything is analyzed, the profiles are then tabulated for a final risk score.
STEP FIVE
Response
A risk score is sent back to the Vesta customer along with 5 key insights on what factors attributed to the score. Customers can then choose to take the action best suited for their business.
Fast
Analysis typically takes less than half a second
Frictionless
Gain insights from Vesta’s 25 years of fraud expertise that help you stop fraud based chargebacks and stop wasting time and resources on chargeback remediation.
Accurate
Allows you to raise approval rates through accurate detection
More Revenue. Less Headaches. What’s not to Like?
Thanks to Vesta’s deep knowledge and mature machine learning models, merchants can enjoy an average of 10% or more revenue through increased approval rates. Stopping fraud is one thing—approving more transactions is another. Vesta does both.