20160525-ANA-Bot_Baseline_FRAUD_IN_DIGITAL_ADVERTISING_44页_7mb

2015 Bot Baseline Report Summary

Core Content

The 2015 Bot Baseline Report, conducted by the Association of National Advertisers (ANA) in collaboration with White Ops, Inc., provides an in-depth analysis of the state of ad fraud in the digital advertising ecosystem. The report highlights the persistent and evolving nature of bot fraud, its financial impact, and the need for improved detection and prevention strategies across all stakeholders.

Main Findings

1. No Advertiser Was Immune to Bots

• Bot rates remained high across all advertisers, with an average decline of only 0.2 percentage points compared to 2014.
• The range of bot percentages increased in 2015 (3–37%) compared to 2014 (2–22%).
• A quarter of the advertisers experienced bot rates of 9% or higher, while the lowest bot level was 3%.
• The majority of advertisers who participated in both 2014 and 2015 saw an increase in bot fraud, showing that the problem persists and is not yet solved.

2. Bot Impacts Vary Widely

• The financial impact ranged from $250,000 to $42 million per advertiser.
• The estimated global losses due to bot fraud in 2016 could reach $7.2 billion, up from $6.3 billion in 2015.
• High CPM media was more vulnerable to bot fraud, with:
  • Display media over $10 CPM showing 39% more bots than lower CPM media.
  • Video media over $15 CPM showing 173% more bots than lower CPM video media.

3. Bot Fraud is Persistent and Evolving

• Bots are not static; they evade detection and adapt to new strategies.
• Sophisticated bots use complex techniques to mimic human behavior, making them harder to detect.
• Freshly infected machines are more likely to be used for high-CPM advertising, while older bots are typically used for lower-value traffic.
• Bot operators shift their focus based on economic incentives, such as moving from display to video advertising.

4. Sourced Traffic Increases Bot Risk

• Sourced traffic (acquired through third-party means) is more than three times more likely to contain bots than unsourced traffic.
• Programmatic buys are especially vulnerable to bot fraud due to the lack of transparency and complex supply chain.

5. Bot Prevention Techniques Fall Short

• List-based bot prevention (such as list-lookup) is ineffective against sophisticated bots.
• General bot detection systems are unable to distinguish between human and bot traffic effectively.
• Blacklisting can remove valid human impressions, leading to loss of legitimate traffic.

6. Bot Traffic is Shifting

• Bots migrate to new platforms and domains when their current methods are blocked or become less profitable.
• The platform of last resort is where low-cost and low-quality traffic is purchased, often from long-tail publishers.
• The profit window of bots remains open as long as new infections are not detected faster than old ones.

Key Information

• 49 advertisers participated in the 2015 study, compared to 36 in 2014.
• The study period was August 1 to September 30, 2015, matching the 2014 timeframe.
• Anti-fraud technologies and proactive policies are essential to reduce bot fraud.
• Combined defenses are necessary to effectively combat bot fraud, as no single solution is sufficient.
• Transparency in the supply chain is crucial for identifying and blocking bot traffic.
• High-value inventory is more susceptible to bot fraud, especially in video advertising.

Recommendations

For All Stakeholders

• Implement sophisticated bot detection and anti-fraud technologies.
• Monitor inventory continuously for bot activity.
• Maintain transparency in all ad transactions.

For Advertisers and Agencies

• Be aware and involved in fraud prevention.
• Understand the programmatic supply chain and request inventory transparency.
• Include non-human traffic language in terms and conditions.
• Use third-party monitoring to ensure compliance with anti-fraud policies.
• Announce anti-fraud policies to all external partners.
• Support the Trustworthy Accountability Group (TAG).

For Publishers, Platforms, and Exchanges

• Relentlessly monitor for fraud.
• Cut off bot sources.
• Maintain transparency and allow buyers to monitor media quality.
• Improve measurement policies to allow full fraud assessment.

Conclusion

The 2015 Bot Baseline Report confirms that bot fraud remains a significant threat to the digital advertising industry. Despite efforts to detect and prevent fraud, bots continue to evolve, evade detection, and target high-value inventory. The report emphasizes the need for unified, proactive, and sophisticated defenses to reduce the financial impact of bot fraud and make it unprofitable for operators.