In the past 20 years, there has been a substantial shift of commerce to online channels, with most products and services now bought and sold on the internet or through mobile apps. Fraudsters have been quick to take advantage of the online shift, doing their best to stay one step ahead of the businesses and consumers they exploit. It has evolved from simple individuals hacking systems with a stolen credit card, to criminals running larger groups of people and combinations of data, to full on automated scripts hitting systems with scale.
The new era or “Fraud 3.0” of today requires data science and machine learning to fight systematic attacks such as card-testing and bots that occur earlier in the transaction flow. These attacks have a couple of similar characteristics, including the time in the transaction flow at which they occur and the scale of speed and volume at which they happen.
Evolution of product needs
It also means product innovation and adaptation is a must to fight this new level of sophistication so companies can respond to these attacks as quickly as they occur. Historically, businesses mainly used internal data in the early stages of the transaction flow with third-party data sources being used on just a subset of transactions later in the workflow. Internal data is extremely valuable, yet the resistance to adding third-party data wasn’t often because of lack of interest or because they didn’t think they could improve their model performance. It’s more a problem of “operational expense.” Making a third-party call-out was thought to be expensive—expensive in terms of latency. But, staying a step ahead requires leveraging third-party identity verification on a higher volume of transactions, earlier in the workflow, and at extremely fast response times to gain the edge.
Introducing our Transaction Risk API
The Transaction Risk API was built by data scientists for data scientists and designed for easy integration into models. It provides the most predictive identity verification features to fight payment fraud and improve the efficiency of authorizations in under 100 ms by scoring the overall risk of an identity using email, IP, phone, name, and address. Powered by the fastest and most reliable tech stack in market, the Transaction Risk API scales for any low-latency, high volume model requirements and provides low-friction, sub-second identity verification data to verify good customers while segmenting questionable transactions into a higher-friction (and lower-risk) path. For more information about Transaction Risk API, visit the product page or request a copy of our Transaction Risk for machine learning user guide.