Improving CPA Using Deduplication

About Client

The client is one of the world’s largest and oldest multi-channel bookmaking and gaming companies from the United Kingdom, operating in over 25 countries with annual revenue exceeding €3 billion as of 2020.

Background

  • While affiliate marketers are a boon for the industry, a few indulge in unethical practices that prove to be very costly for businesses in the short and long term. Affiliate frauds have become a rapidly increasing threat due to their attractive ‘low-risk, high return’ nature. In an enormous database of player information, duplicate accounts go undetected.
  • Needless to say, many online casinos suffer huge revenue losses as they incur high costs of acquisition and often lack in resources to identify multiple accounts being created by the same players and pushed by the affiliates to operators.

Challenge

  • The client was incurring high acquisition costs due to affiliate frauds. Affiliate marketing is a massive industry in itself & is a prominent source of revenue across industries.
  • The client offers thousands of wagers over the phone and the Internet on a daily basis and they needed a swift solution; a system that could identify duplicate accounts across their massive portfolio of established brands. They engaged BizAcuity to ensure that necessary steps were taken to identify duplicate accounts.

Our Solution

Bizacuity Has Core-Competencies In Building And Delivering BI Solutions For Clients. A Structured Framework Was Developed By Bizacuity

  • Customer attributes consisting of biographic and demographic data were obtained, and each attribute was given a specific weightage to maximize the correlation between players.
  • We cleaned the data before applying advanced text mining techniques that were used to match names and addresses.
  • After testing and running multiple ML algorithms and fuzzy logic, we were able to improve the accuracy of deduplication for the firm

Outcome

  • Better ROI for the client on Affiliate marketing efforts going forward

The client started to save approximately

80,000 GBP

Per month by suppressing affiliate payments

93.7%

Accuracy in the identification of duplicate accounts