• by Laurie Sullivan  @lauriesullivan, May 7, 2024

Worldwide fraud rates for optimized ad campaign meant to protect brands fell to all-time lows in the second half of 2023, but those unprotected rose 14 times higher, according to data released Tuesday.

Rates for optimized campaigns fell to 0.6% from 0.9% the two prior years. The year-over-year rate in the Americas dropped to 0.1 point.

Global media measurement and optimization platform Integral Ad Science on Tuesday released the nineteenth edition of its Media Quality Report (MQR). The annual report analyzes more than 280 billion digital interactions daily from across the globe to provide advertisers and publishers with benchmarks to measure the quality of their digital advertising campaigns and inventory, from ad fraud rates and bots, to viewability and challenges related to artificial intelligence, deepfakes, and misinformation.

“In a year marked by global elections, the Olympics, and the UEFA European Championship, marketers are increasingly focused on protecting brand reputation and allocating budgets toward content that minimizes risk, especially in complex channels like CTV, social media, and the open web,” said Jeremy Kanterman, VP, Research & Insights at IAS. “AI-generated deepfakes pose a unique threat, while, in parallel, consumers’ perceptions of brands can shift when ads are seen near risky content.” 

The data shows that much of the reduction in global ad-fraud rates in the second half of last year was driven by a 0.4 point year-over-year (YoY) drop on desktop, where display and video impressions averaged 1.0% and 0.9%, respectively. Despite the improvement, desktop ad fraud rates remained three times higher in the second half of 2023 compared with mobile.

Non-optimized campaigns on desktop display increased two points to 13.0% between the first and second half of the year. Desktop video is the only format where fraud rates for non-optimized campaigns declined annually, landing at 5.7% in the second half of 2023, a format consumers watch only for an average of 22 minutes per day compared with 79 minutes for mobile video, according to an eMarketer February 2024 report.

The reduction in the Americas broke a three-year increase that pushed ad-fraud rates to an all-time high in the first half of 2023.

With increasing amounts of fraud, marketers are most focused on navigating threats from AI-generated deepfakes and misinformation, as consumers continue to hold advertisers accountable for ad placements around risky content.

Brand risk averages worldwide remained essentially unchanged annually, declining 0.1 point from 2022 to reach 1.7% in 2023, but brand risk continues to change rapidly. AI-generated deep fakes pose unique challenges around content authenticity and misinformation.

Increased election coverage and political ad spending in the U.S. and Argentina in 2023 likely contributed to an annual hike of 0.2 points that pushed the Americas to a 2.1% brand risk average. The U.S. registered a 0.3 point average brand risk increase from the first half of the year to the second.

Brand risk on desktop remained unchanged worldwide across both display and video, coming in at 1.6% and 1.4%, respectively -- lower than on mobile.

For mobile web ads, brand risk fell, but it was highest around mobile web display for the second year in a row -- reaching 2.0% in 2023 but down from 2.3% the year prior. Mobile web video came in at 1.7%, following a 0.4 point decline year-over-year.

IAS also looked into brand risk vulnerabilities around major sporting events, specifically the Super Bowl and the FIFA Women’s World Cup.

Sporting events are considered generally safe content for brands, but unforeseen risks may arise during live broadcasts and streams with continued fallout well after the event. For example, the trophy ceremony of the Women’s World Cup featured unexpected high-risk negative content.

Super Bowl brand risk in the U.S. averaged 3.5% in the week leading up to the game in February 2023, which was 59% higher than the 2.2% benchmark for the U.S.