Fraud Scanner

Fraud Scanner aims to protect your data against different types of fraud, such as click injection, impressions flooding, fake devices, and others that distort results of ad campaigns and destroy the budget.

As advertising costs increase, this industry attracts more and more scammers. Unfair players may be among your users, trusted advertising agencies, and even among reliable partners.

To protect your data and not to pay for fraudulent traffic, track fraud metrics with myTracker.

How it works

Fraud Scanner currently works only for mobile applications.

Every day system monitors the incoming traffic and detects fraud metrics for each installation. For example, how many installs are made from suspicious devices, what percentage of the total number refers to quick installations, and so on.

The system only marks the fraud indicators but does not block installs automatically. So, you can make an informed decision on your own, for example, to fine or disconnect a partner

Stages

Fraud monitoring includes several stages:

  1. Hardware stage. At the first stage, the system recognizes fraud by hardware characteristics of devices. For example when the Android application is installed on the iOS platform, frequent changes of the launch country.
  2. Click stage. At this stage antifraud works when a user interacts with advertising and before he installs the app. The system indicates all doubtful clicks and impressions: quick installs, rolling clicks, low CCR, and etc.
  3. In-app stage. The last stage is fraud monitoring after the app installation. The system distinguish abnormal behavior, such as those who never launched the application, or users with short session.

Based on the results of all stages Fraud Scanner forms a complete picture of fraud that tells us about the quality of traffic in general: strict, confident and soft fraud. You can also see individual fraud metrics such as deferred installs, low CCR, etc.

Benchmarks

To properly assess the threat of fraud, you can use special indicators of fraud — benchmarks. These are reference values, if exceeded, the system signals a fraud by highlighting the indicators in red.

Benchmark is defined separately for each fraud metric based on of machine learning technologies and a large volume of accumulated data for iOS and Android platforms.

For example, benchmark for the Suspicious devices metric for Android is 0.81%. That is, if more than 0.81% of the installs for the selected period are made from Suspicious devices, myTracker will highlight the metric value in red, and you will be able to detect fraud in time. For details, refer to the How to read reports section.

How to use

We recommend regular monitoring fraud:

  1. Use Default templates: Combined fraud metrics, Strict, Confident and Soft fraud, or build reports for Fraud Scanner metrics in the Constructor.
  2. If you detect a fraud, take action to prevent fraudulent attacks (for example, to fine or disconnect a partner).

For details about how to read reports and detect fraud, read the Fraud detection section

The list of available fraud metrics, templates and traffic detail depends on the selected Basic or Premium plan.

Basic plan

Basic Fraud Scanner features allow you to detect fraudulent traffic using Combined metrics, find bots by High Install Rate and recognize click-spamming using the Low CCR metric.

The basic plan is enough to assess the threat of the fraud and see if there is a need to connect the Premium plan.

Premium plan

The Premium plan brings all Fraud Scanner advantages:

  • monitor statistics using Partner, Campaign, SubID, Traffic type, Country dimensions, and find the source of fraudulent traffic;
  • analyze multiple metrics: Suspicious Devices, Quick Installs, Short Session, and so on to recognize signs of fraudulent schemes;
  • use detailed templates, such as Strict, Confident and Soft fraud for easy and effective traffic evaluation.

To go to Premium plan contact your myTracker manager or send e-mail to support@tracker.my.com

In addition

  • Stop cooperating with partners involved in the fraud
  • Insist on maximum details of your partner's traffic
  • Be attentive to the slightest change in fraud metrics
  • Configure event tracking so that you can easily detect conversion funnels. Track not only target events that are easy to falsify, but in-app events, such as passing levels
  • Please keep alert even with a tried partner.

Soon you will see fraud data on the Dashboard and will be able to export data via the API. For details contact our support team