TAG Fraud Benchmark Study

benchmark_2018-690737-edited.png

The 614 Group today released an updated analysis by The 614 Group showing a reduction in fraud of more than 84 percent in TAG Certified distribution channels vs. the broader industry average.

The analysis measured more than 75 billion ad impressions in 2018 from five of the largest ad agency holding companies, and it found that invalid traffic (IVT) rates were held to less than two percent in TAG Certified channels for the second year in a row. The 614 Group expanded its analysis from 2017 to include Sophisticated Invalid Traffic (SIVT) and a broader range of inventory types, including billions of video, mobile web, and in-app ads.

“There’s a saying in the technology industry that once can be lucky, but twice means you’re good,” said Mike Zaneis, CEO of TAG. “By reducing fraud rates in TAG Certified channels to less than two percent for the second year in a row, TAG’s Certified Against Fraud Program has firmly established the success of its model to reduce invalid traffic and stop the flow of money to the criminals who profit from it. This study reinforces why TAG’s anti-fraud certification has become a baseline requirement for doing business in the digital ad supply chain, and we look forward to building on its success as we expand our programs globally.”

Inventory Type

Desktop display, desktop video, mobile web display, mobile web video, in-app display, in-app video

Types of Fraud Examined

General Invalid Traffic (GIVT), Sophisticated Invalid Traffic (SIVT)

Volume of Impressions Examined

75 billion (including traffic from the world’s largest media companies)

SIVT + GIVT Rate (TAG Channels Only)

1.68 percent

Overall Fraud Rate (Blended Benchmark)

10.43 percent

Reduction in Fraud

84 percent

Study Duration

January - August 2018

Data Examined

Campaigns executed by major advertising agency holding companies: Annalect (Omnicom Media Group), GroupM, Horizon Media, IPG Mediabrands, and Publicis Media

 

TAG Fraud Benchmark Study

Press Release