ANA-2018-2019年网络广告流量欺诈报告（英文）-2019.5-38页_3mb

2018-2019 Bot Baseline Summary

Core Content Overview

This report from White Ops and the ANA provides an in-depth analysis of ad fraud in the digital advertising ecosystem, highlighting both progress and ongoing challenges. It outlines the current state of fraud, the measures taken to combat it, and future recommendations for improving transparency and security.

Major Findings

• Fraud Reduction: Fraud losses are projected to be $5.8 billion in 2019, down from $6.5 billion in 2017, despite a 25.4% increase in digital ad spending. This shows the industry is making structural gains against ad fraud.
• Fraud Stymied: For the first time, more fraud attempts are being stopped before they are paid for, thanks to improved measures like DSPs and SSPs filtering bid requests, and clawbacks.
• High-Standard Validatability: Only a small percentage of impressions are fully validatable at the highest standard, with desktop display at 51.8% and desktop video at 36.8% being the best performers. Mobile video is the worst at 28.8%.

The Good News

• Desktop Fraud Decline: Desktop ad fraud has reached its lowest level in the history of the Bot Baseline study, with display fraud at 8% and video fraud at 14%.
• Structural Improvements: Advertisers are increasingly using ad tech platforms with built-in fraud prevention, and initiatives like Ads.txt have reduced desktop spoofing.
• Industry Collaboration: The formation of coalitions like TAG has led to the dismantling of major botnets, including 3ve, and increased accountability for cybercriminals.

The Bad News

• Transparency Gaps: Less than half of all impressions are fully transparently validatable, with many still relying on outdated formats like VAST 2.
• Inconsistent Monitoring: Real-time independent monitoring is still inconsistent, with many platforms using low-fidelity validation methods.
• New Fraud Frontiers: Fraud is moving into newer areas like mobile, OTT, and SSAI, where it's more difficult to detect and measure.

Key Challenges

• Outdated Technology: Many publishers still use VAST 2, which does not support JavaScript and limits fraud detection capabilities.
• Walled Gardens: Some platforms have limited third-party validatability by design, making it harder to monitor fraud.
• Human Error and Evasion: Cybercriminals are using more advanced evasion techniques, such as tag evasion and selective tag execution, to bypass detection.

Recommendations

• Adopt VAST 4: Marketers should push for VAST 4 support, which enables vendor verification and viewability, and is the gold standard in transparency.
• Increase Validatability: Advertisers should demand higher validatability standards across all platforms and formats.
• Enhance Third-Party Measures: Use dynamic fraud detection methods like JavaScript-based verification, SDKs, or other real-time integrations to ensure accurate monitoring.

Glossary and Partners

• The report includes an ad fraud glossary for clarity.
• A list of ANA member companies that participated in the study, including major brands like Unilever, State Farm, and Coca-Cola.

Conclusion

While the industry has made significant strides in reducing ad fraud, the fight is far from over. New formats and evasion techniques are emerging, and transparency remains a challenge. However, the momentum is there, and with continued collaboration and investment in better technology, the industry can further reduce fraud and protect advertising budgets.