Basic Attention Token logo Basic Attention Token

Reducing Digital Ad Fraud: A New Deal with BAT

by Basic Attention Token | April 7, 2017 | General , Token , Ad Fraud

Ad fraud is growing dramatically as no adequate solutions have appeared. The Basic Attention Token (BAT) seeks to radically reduce digital ad fraud, which is estimated to be $16B in 2017, according to a report released a few weeks ago by the World Federation of Advertisers. The sheer number of middlemen present in the ad-tech ecosystem inevitably creates entry points for fraud to flourish. Publishers do not accurately know what ads are programmatically served to their ad slots. All of this is part of what recently prompted P&G’s Chief Brand Officer Marc Pritchard to say: “The days of giving digital a pass are over. It’s time to grow up. It’s time for action.”

The industry faces a huge fraud problem with no relief in sight:

The ad fraud problem is not limited to a small group of bad actors and victims. Recent incidents demonstrate that some of the largest players in the space are not immune:

  • In January 2016, CNBC reported that Yahoo was allegedly not serving ads to 30–70% of the locations that they had claimed. Instead of pre-roll video ads being served to the video player placements that demand $20 CPM (cost per thousand views), sources reported that Yahoo was allegedly serving the video ads in smaller $2–4 CPM display ad slots that run alongside video players.
  • In September 2016, the Wall Street Journal reported that Facebook had been overstating video ad metrics by an estimated 60–80% for two years, resulting in a class action lawsuit.
  • In January 2016, Google was exposed by MalwareBytes as having their ad network abused by a fraud operation that targeted European users using layered ad injection that displayed fraudulent cookie disclosure notifications.
  • In March 2017, the New York Times reported that JP Morgan Chase scaled back their ad delivery from 400,000 sites to a whitelist of 5,000 sites, and surprisingly saw little to no change in the cost of ad impressions or visibility.

What does the ad industry propose for addressing the issue?

The Trustworthy Accountability Group (TAG) is a cross-industry accountability program fighting criminal activity across the digital advertising supply chain. Founded by the American Association of Advertising Agencies (4A’s), Association of National Advertisers (ANA), and Interactive Advertising Bureau (IAB), the TAG Certified Against Fraud Program Guidelines are often cited in fraud reports as the ad industry’s response to fraud. The blind spots in these guidelines are obvious:

  • Participants are required to pay annual fees to participate.
  • Self-certification is an acceptable method for participants, with third party certification being strongly encouraged but not mandated.
  • No requirements are made for the resources needed to support Certified Against Fraud compliance aside from the requirements that certified companies always have a designated TAG Compliance Officer.
  • There are no mentions of user privacy, or signs of privacy protection by design, in the guidelines.
  • The TAG Certified Against Fraud Program does not account for fraud occurring outside its purview.

BAT: Radically reducing fraud, empowering the parties that matter.

The system needs major change, not minor rearrangements or additional moving parts. Half-measures will not make a substantial impact. That is why we created the Basic Attention Token (BAT). It creates a platform for a new deal to take place between publishers, advertisers and users. It prices users in from the start and along the way. The BAT underpins more efficient economics relationships, realistically addresses fraud, and protects user privacy.

  • BAT protects against fraud by eliminating third-party traffic sourcing, which the 2015 White Ops & ANA Baseline Bot Study concluded was more than three times as likely to contain bots than unsourced traffic. This ability will substantially reduce the fraud outlined in the aforementioned reports. Also, by reducing middlemen, BAT eliminates the fees, latency and additional round trips in the ad-serving process. This means fewer ads and trackers, thus a much better user experience, with better economics for users, advertisers, and publishers.
  • BAT’s Basic Attention Metrics (BAM) system removes fraud entry points. BAM will measure the user’s direct attention and engagement on the mobile and desktop browser or other BAT-enabled app. This measurement data will be private, because it is not transmitted across the network. By measuring directly on the device, BAM will greatly reduce cross-platform discrepancies that have become common among MRC Accredited 3rd parties. These 3rd parties apply their own methods (some of which rely on Flash, a deprecated and insecure plugin) for measuring “viewability.” By keeping user data on-device, BAM measurements also avoid the privacy problems of ISP and other 3rd party interception associated with transmitting user data over the Internet. Equally important, the BAM will improve reporting of ads viewed to substantially reduce the over-reporting of this data by companies.
  • Fraud hides in the intentional opacity of existing ad-tech. Transparency is a vital component of any legitimate ad ecosystem. To ensure the integrity of the platform for brands and users, BAT code will be auditable as open source. Further, the public and decentralized nature of the blockchain ensures that macro-economic and eventually anonymous transactions will be auditable. will develop no-exception rules against remote tracking and third-party ads that depend on tracking for campaign measurement. This is necessary to avoid arbitrage and Gresham’s law problems.
  • Only the user with the aid of BAT-enabled apps mediates between advertisers and publishers, so third-party invalid traffic and other fraud detection bypassed by bots will be excluded from the start. Private ad impressions and actions will be verifiable using Zero-Knowledge Proof (ZKP) protocols such as Anonize (already used by the Brave browser for its opt-in micro-donation service).
  • Personalized ads will be placed based only on app- and device-local data and analytics that consult a common downloaded inventory catalog. This reduces vulnerabilities to threats such as as fraudulent layered ad injection mentioned in the cookie disclosure example above.
  • Ads and payouts will be rate limited, making bot operations hard to deploy at scale.

This list gives some of our founding operational principles and how they can be implemented to create an open, verifiable direct ad delivery and measurement platform. Ads are guaranteed viewable with BAM, verified via ZKP protocols, and bot fraud restricted by payments being rate limited to known publishers only. Based on feedback from users, advertisers, and publishers, we will develop detailed specifications in consultation among the members of the group.

The goal is to ensure real users view authentic, ads while protecting users’ data performance, privacy, and safety. This is what the ad industry has failed to bring about in almost two decades (studies on fraud in digital advertising began to call attention to this problem as early as 1999). Ad fraud takes more and more from overall ad spend; it has broken trust among all parties in the current indirect ad delivery model. Some ad industry solutions contemplate adding more middlemen to the system, but more layers will not solve the problem.

Digital advertising needs a new deal, where ad fraud is not considered a “cost of doing business.” BAT will provide the token of utility and decentralized platform to uphold the terms of this new deal.