Google has 160 separate filters to block fake traffic in China

Nearly a quarter of digital ad content report fraudulent traffic—ranging from 3 to 37 percent—as reported in The White Ops paper released by the Association of National Advertisers.

“Fraud affects everyone in the ad value chain,” said Richards Gilbert, Head of Performance Solutions, N. APAC at Google.

At the Google Data Summit in Shanghai this September, he told an audience of China-based data marketers that while invalid traffic was especially prevalent in programmatic, display ad fraud occurs across various types of content.

Additionally, bots need to build up “profiles”, like visiting premium sites and putting items in shopping carts, in order to appear “human” and avoid detection, which means invalid traffic can involve all types of transaction models.

However, Gilbert stressed that “the more valuable the content you have, the more attractive it is to fraudsters.”

Gilbert noted there were many different kinds of fraud, such as fake users, ineffective clicks, fake ads, faked data centre traffic and equipment hijacking. But broadened the definition to include any kind of traffic that does not represent the intentions of consumers.

“CTR, CPC, and even CPA can be manipulated,” he said.

He told the audience that as the owner of the biggest display ad network, the biggest ad exchange and the largest app network, as well as major properties such as YouTube and Gmail, Google was committed to combating fraud.

Google’s response to fraud covers technology, personnel and policies. The company has developed 160 separate filters to monitor and block fake traffic. 

“We use automated and manual detection systems combined with machine learning and deep research to identify anomalies in traffic patterns. We block any activity that’s found to be invalid or fraudulent,” Gilbert said.

Google has the industry’s largest anti-fraud team, with more than 100 researchers and engineers focused on identifying and eliminating ad fraud.

Gilbert said the company’s strong fraud policies also protected advertisers. “Advertisers do not need to pay for fraudulent traffic."

“Google’s anti-fraud teams and systems regularly detect fraud. If any fraud is discovered the advertiser will be refunded and the source of fraud will be penalised.”

In 2015, the company stopped more than 125 million ads for drug and healthcare products that had breached FDA regulations.

Gilbert said Google is a big supporter of the IAB Transparent Supply Chain Initiative, which requires that media buyers receive identifiers that identify every upstream party involved in the supply of media.

Aside from anti-fraud measures, Google is also pitching its holistic data solutions.

“Google’s trusted data solutions can connect fragmented and siloed data into a holistic perspective, turning quality data into actionable insights for brands,” said Guo Zhiming, general manager, Google Data Insights and Solutions, Greater China & Korea.

Google’s GA360 technology offers a single view of the customer, Guo said.

“It delivers end-to-end tracking which enables advertisers to obtain deeper consumer insights to optimize marketing decisions.

“It allows you listen to the details of each customer brand interaction and, if needed, alter the course of how you engage with your customer to achieve better business results.”

Additionally, Google’s solutions can deliver the important metric of ad viewability, “thereby helping advertisers achieve true brand exposure.”