Fraud Analytics

High Level Design

Overview

As the online media streaming industry is significantly growing, every media corporation is struggling to release new content and stable technology, and one should think about the next level. Every company offers the same promotion to new customers - limited time of trial period to try the service and then recurring payment. The common trial time that is given is between 7 to 30 days. About 5% to 10% of the new customers are abusing this offering, by registering unlimited times, using creative methods to use the service for free. The process of re-registration is easy to execute for all the customer types. This is where the fraud analytics dashboard comes into play.

Goals

Monitoring fraud activity volume and value and discover insights about the customers who takes fraudulent actions. Understand how to decrease the number of fraudulent subscriptions.

Objectives

    • Discover the insights behind the fraud memberships number (registered for the first time and started their trial) – age, location, device
    • Identify fraudulent subscriptions volume (% of Active Customers) – registered and found as owners of previous not active subscription (by payment method + name, same phone number, same address + name)
    • Discover the current fraudulent activity monetary value, trends over months, growth
    • Understand the estimated loss of fraudulent activity per year Identify trends and picks of fraudulent activity by special characteristics such as customer’s location, device, gender and payment method

KPIs Architecture

Objectives KPIs Measures Data Source
Identify Current fraudulent activity value volume $ Current Active Fruadlant Customers Loss SUM([CustomerID],sum([Plan])) DIM Customer / FACT Customer Journey
% Growth of Fraud Activity from last month GrowthPastMonth([Past On Trial  Fraud]) DIM Customer / FACT Customer Journey
% Fraud of Active Customers sum([Current On Trial  Fraud])/COUNT([Customer ID]) DIM Customer / FACT Customer Journey
# Fraud Customers sum([Current On Trial  Fraud]) DIM Customer / FACT Customer Journey
Discover the trends behind fraudulent activity % Growth of Fraud Activity GrowthPastMonth([Past On Trial  Fraud]) FACT Past Customer Journey
% Growth of new Memberships GrowthPastMonth([New Memberships]) FACT Past Customer Journey
Insights behind the past fraud memberships # Of Fraud Memberships by AgeRange/Gender [Total PastOnTrial_Fraud] based on Gender and AgeRange FACT Past Customer Journey
# Of Fraud Memberships by Device [Total PastOnTrial_Fraud] based on Device FACT Past Customer Journey
# Of Fraud Memberships by Payment Method [Total PastOnTrial_Fraud] based on PaymentMethod FACT Past Customer Journey
# Of Fraud Memberships by City [Total PastOnTrial_Fraud] based on City FACT Past Customer Journey
# Of Fraud Memberships by State [Total PastOnTrial_Fraud] based on State FACT Past Customer Journey

Entities Relationship Diagram

Plugins & Scripts used

Implementation
Kit

The following resources will enable you to design your dashboard and data model with sample data and then apply it to your own data. Note that you will need to have a previously installed version of Sisense (you can use the free trial version if you’re not a customer).

Sample data and dashboard examples (direct download)

Documentation