Risk scoring

Risk scoring assigns a transaction or account a number that summarises how likely it is to be fraudulent, by combining many weak signals into one value. The score is an input to a decision, not a verdict, and is only meaningful relative to where its bands are set.

In more detail

Scoring exists because no single signal is decisive. A new device, an unusual hour and a first-time beneficiary each add a little risk; a score calculates how much they add together. Scores are usually grouped into bands, with low scores passed, middle scores reviewed and high scores escalated.

The misconception is that a high score means fraud. A score is a ranking: it orders transactions by suspicion so the limited attention of an analyst goes to the right ones first. Where the bands sit is a business choice, and moving a band trades missed fraud against analyst workload.

Example

A payment scores high because it combines a new device, a first-time payee and an amount well above the account's norm. A second payment with only the new device scores low and passes untouched.

Related capability

Risk scoring on the platform

Related terms

Discuss your transaction monitoring requirements

Tell us which transaction flows you need to monitor, which systems Akoben would integrate with, and how your fraud team works today.

Request a demo