What is Omniscore?
Omniscore is a transaction safety rating that can be used in rule creation and during the manual review process to determine the disposition of an order (approve, decline, review). It is the output of Kount's next-generation AI model analyzing hundreds of millions of transactions—their outcomes (including approvals, declines, chargebacks, refunds, etc.) and their real-time linkages and patterns. The AI weighs the risk of fraud against the value of the customer and provides an evaluation (approximating an experienced human fraud analyst) in the form of a score which helps identify good customers, bad customers, and fraudsters.
Omniscore differs from previous scores in that it incorporates the most predictive components of both our supervised machine learning and our unsupervised machine learning, as well as other predictive factors, into one score.
The result is a score that is twice as good at catching fraud and minimizing false positives as any score Kount has ever produced.
How is Omniscore calculated?
Omniscore uses two types of machine learning: unsupervised and supervised. The unsupervised machine learning focuses on short-term linkages and patterns, enabling it to catch emerging fraud attacks and anomalies that supervised machine learning cannot yet learn about due to the recentness of unseen attack types. Our supervised machine learning technology learns from historical data—decisioned orders and their outcomes.
The AI simulates how an experienced fraud analyst would review a transaction. The unsupervised machine learning aspect of Omniscore evaluates the transaction as a human would use instinct. The supervised machine learning aspect evaluates the transaction like the historical experience of seasoned fraud analysts. Together they allow Kount to calculate one highly predictive transaction safety rating that can be relied upon for decisioning orders, so that there is less reliance on manual review and reactive fraud rules. The result is catching more true fraud and allowing more good transactions to generate revenue.
How should Omniscore be interpreted?
Transaction safety is the inverse of transaction risk. Omniscore is an indicator of a transaction’s safety ranging from .1 (unsafe) to 99.9 (safe). A safe transaction will have a relatively high Omniscore and an unsafe transaction will have a relatively low Omniscore.
Designed to make good decisions more intuitive, the Omniscore can be likened to U.S. academic letter grades that range from F to A. Most transactions will rate in the 80s and 90s (Bs and As). Transactions with issues will rate in the 60s to 70s (Ds and Cs). The riskiest transactions rate below 60 (F).
|
Omniscore |
Grade |
Description |
|
90 – 99.9 |
A |
Very safe, multiple indicators of safety found |
|
80 – 89.9 |
B |
Indicators of safety found |
|
70 – 79.9> |
C |
Typically a mix of safe and risky indicators |
|
60 – 69.9 |
D |
Indicators of risk found |
|
0.1 – 59.9 |
F |
Very risky, significant indicators of risk found |
It is important to note that Omniscore is not a decision. It is a prediction of safety that is used by merchants to decision a transaction (either automatically via creating a rule or manually while under review).
How should a fraud rule be created using Omniscore?
Since Omniscore is so accurate in predicting fraud, you can set one rule around it instead of creating large rulesets targeting fraud.
A suggested rule of thumb is to determine the decisioning threshold (at what value the Omniscore is set) based on your desired decline rate:
|
Desired Decline Rate |
Omniscore Fraud Rule |
|
5% |
If Omniscore < 61 Decline |
|
4% |
If Omniscore < 49 Decline |
|
3% |
If Omniscore < 37 Decline |
|
2% |
If Omniscore < 25 Decline |
|
1% |
If Omniscore < 13 Decline |
The decisioning threshold can be adjusted after analyzing decline and chargeback rates, and any other measures of performance important to the customer.
What does Omniscore cost?
There is no additional charge for Omniscore.
Is the AI decisioning orders?
No. The AI evaluates the safety of the transaction and provides the Omniscore value. Omniscore is used for decisioning only when a rule is created to set a decisioning threshold on it or when a manual reviewer incorporates it into their manual decision. In other words, the decision to approve or decline a transaction is, strictly speaking, performed by a rule or a manual reviewer. Whether Omniscore is used in the decision is up to the user to determine.
How will Omniscore change the product?
Omniscore will replace the Boost score and eliminate the need to use the Persona Score in rules. Customers can rely on Omniscore as their ONE score for payments fraud prevention.
Every place in the AWC that Boost displayed, is now displaying Omniscore. The only exception is that rules created previously using Boost will continue to display the “Boost” label.
For example, if a customer currently has a rule to decline if the Boost score is less than 10, the underlying logic will automatically be updated to decline if the Omniscore is less than 10. But the label will continue to display “Boost” even though the score being used is now Omniscore. Should customers want to remove the “Boost” label from rules created prior to Omniscore (because they may find it confusing that the label says “Boost” but the score is Omniscore), they need to recreate that rule by selecting Omniscore while creating the condition and delete the rule with the “Boost” label.
Since Omniscore follows the same score distribution as Boost, users will not experience a noticeable change in the frequency of specific scores. What will be noticed is that the number of transactions declined that were actually fraud will increase and the number of transactions declined that were not fraud will decrease. In other words, Omniscore is Kount’s most effective predictive score for getting more fraudulent transactions and fewer good transactions into the “decline” category.
How is Omniscore different than Boost and Persona Score?
Omniscore combines Boost and Persona technology (along with other predictive values) into ONE score. It eliminates the need for any other score because it is so accurate in predicting fraudulent transactions, simulating seasoned fraud analysts in how they evaluate a transaction (both from a historical experience aspect and an instinct aspect). The combination of Boost and Persona into one score, allows us to deprecate the Boost Safety Rating (with the June 2019 release) and the Persona Score (in a future release).
Why is Persona Score still in the product?
Since Omniscore includes Persona technology, there is no need to use Persona Score in rule creation. Persona Score will remain in the product for a period of time, but will eventually be deprecated just like Boost.
Can a customer test Omniscore without having actual transactions affected?
Yes, a customer can create a "No Change" rule using Omniscore. This will allow the customer to watch how the score performs while not impacting actual transactions. While Kount is confident that Omniscore should be the only score used for fraud rule creation, some customers will want to test it. After three months (enough time for outcome data to be collected), the customer will be able to compare performance with and without Omniscore.