Deep learning about your DG&S users from just a bit of given information

When you sell a digital product — online subscriptions, game codes, eBooks, gift cards, digital currencies — you can expect to be favorited on the fraudsters’ playlist.

The Nethone Guard fraud prevention solution: 4 Pillar Protection (Profiler, ML, API, & human experts)

Nethone Profiler: verify your users even when given little data from the checkout form

Less friction, more security, more revenue

The basic checkout form is the common way to complete an online transaction. But forms cause friction and are often a reason for cart abandonment. Shortening the form as much as possible is required to keep conversions — we make that possible!

Gathering clues in an instant

Nethone Profiler gathers 5,000+ attributes about every customer to see if they are attempting to anonymize their actions, conceal their true location, or use automation tools to deceive your business. Profiler also observes raw behaviour associated with keystrokes, mouse and accelerometer to inform you whether your customer is real or a fraudster.

Benefits of using machine learning

Merging contexts and quick adaptation

Reviewing 5,000 attributes about every user in real time is impossible for a human, but it’s perfectly suited for a machine learning model. ML releases the full data potential by merging contexts, automates decision making by providing actionable recommendations in plain language to your risk team, and adapts to constant changes in user traffic and fraud methods.

Minimize false positives

If you are still using basic heuristics and manual rules sets, you’ve taken an important first step. Heuristics can be effective in the short term, but they generate false positives over the long term. Since fraud changes all the time, it’s necessary to use decisioning techniques which adapt quickly.

An API that becomes a true fraud prevention data hub

We integrate multiple data sources and teach our ML models how to use them, including:

• What Nethone has learned about online fraud over the years and what Profiler is seeing in real time;
• Fraud attempts that your company has already encountered;
• External data sources from 3rd party partners who specialize in online fraud.

This gives Nethone a 360⁰ view on each online customer to make the best decision possible.

The fourth pillar: human support for the machine learning models

Personalized machine learning

Our clients are assigned a dedicated Data Scientist who understands your KPIs and shares actionable insights about your business.

Our Data Scientist provides personalised service, so that you are never stuck waiting for Customer Support.

The power of partnership

Our Data Scientist cooperates with your risk analysts to find new fraud patterns affecting your business. machine learning models will be tuned to respond to new patterns of activity.