How DATA Apply Blockchain Technology to Solve “Ad Fraud” Problem

4 min readMar 8, 2018


One of the most overwhelming challenges in the digital advertising industry is the abundance of fraudulent data flowing into the entire network, commonly termed as “Ad fraud”.

Currently, more than 40% of digital ad traffic is generated by automated bots which have defrauded advertisers in billions of US dollars annually. Such fraudulent performance marketing deteriorates efficiency of the whole industry, and results in exorbitant costs for advertisers and thereby cuts into the actual revenue for legitimate publishers. Ad fraud is
more of a systematic problem. As the digital marketing and advertising ecosystem becomes more and more complex, the middle layers between advertisers and publishers also swell into multiple stacks. Although the pattern of fraud traffic can be detected, gradually more and more by the evolving machine learning technologies, the ever deeper origin or the source is hardly traceable because of the multi-stacked middle layers that typically excel in extremely complex redistribution, routing and redirecting. Such complexity inevitably has itself invigorated a new set of problems into the digital advertising ecosystem, such as “walled garden” and “data isolation”, etc. The entire AdTech industry is constantly facing the prisoner’s dilemma in that while data sharing globally would maximize the overall performance and effect, yet the optimal strategy for individual user is to keep the behavioral data exclusive locally on the device, be it a PC or mobile. The reason for such dilemma is due to lack of strong incentive to encourage data sharing, and inability to deliver actual value of such incentive if and when it’s available. Autonomous, fair and transparent governance, while simply self-contradictory in a centralized environment, constitutes main thesis in a decentralized and consensus based infrastructure, especially with the blockchain technology.

Since monetization relying on advertising has contributed to large share of revenue stream for most internet companies, the current model developed in the past decades has been built on a centralized economy by gauging end user’s attention to the ads for better targeting. The demand for user attention has been rising as the competition gets tougher and push for volume number of users becomes a tug of war for the gain and for the survival of Internet companies. The stake is so high and the gain is lucrative, thus created loopholes for the birth of fraud data with fake users, installs and fake activation, etc. that continues to flood the entire communication network. Many advanced fraud detection technologies have been developed to tackle the problems, but most of them focused on the after-fact detection and then to analyze patterns of the data to determine the validity of end user if it even exists.

The technology needs to be able to build into preventing the fraud being propagated into the network traffic flow. The data of the user’s attention patterns need be logged and encrypted to be distributed over a decentralized P2P protocol in order to deliver the heightened security with consensus from the community auto-programmatically. Thus it can be validated any time, while stored across the network and shared in real-time.

Applying the blockchain technology and the Artificial Intelligence plowing on deep learning is the foundation technology that can set the standard to ensure the decentralization of the least risky information data flow and storage requirement, where every player in the community or the ecosystem becomes the ultimate winner.

DATA is a blockchain project initiated and governed by Blockchain Data Foundation Limited (“Blockchain Data Foundation”, or BDF from hereafter), a company limited by guarantee to be incorporated in Singapore and governed in a not-for-profit manner. This project is being developed in collaboration with Yomob International Co., Ltd. (“Yomob”), a mobile monetization-as-a-service company.

The DATA project is devised to address the root of the fraud at its core. To ensure that every participant in the ecosystem is motivated to “do good”, the DATA platform applies a reward system to incentivize end users with their attention contribution and to publishers with pruning their sell-side inventory. More granular information is collected and privately encrypted into a built-in peer-to-peer data storage for sharing across the network. Because every user is made aware of varying rewards only appropriate for his/her attention consumed on ads served in games or apps, it is highly catered and more personal. Therefore a user is less inclined to reject, ignore or block advertising. The experience institutes a more intimate relation between the advertisement and its consumer.

On top of that, advertisement consumption, no matter what format, be it banner or interstitial or a video or a playable advertisement, is tracked at the device level. Besides being the base for the hugely beneficial rewarding for the true user, the analysis of the behavioral data allows the subsequent conclusion of whether the user is a real person or a fake bot.

The DATA project’s most innovative edge is to proactively identify the fraudulence based on the pattern or behavior analysis via our proprietary AI algorithm, prior to the fraudulent user data being injected into the entire network. Such preventive measurement is the other side of the coin where reward is granted to a true user for his/her attention consuming the advertisements and to the associated device contributing to in the P2P storage sharing.

The DATA platform and its token, DATA Token (DTA) will serve as the backbone of its blockchain based advertising infrastructure. The system supports micropayment where tokens can be used across applications, such as purchasing virtual items and premium services, or for publishers- redeemed as currency in ad networks, DSPs (Demand Side Platforms), ad exchanges, third-party SaaS (Software as a Service) service and etc.

Follow DATA on












Decentralized AI-Powered Trust Alliance