Advertisements placed in Web sites and mobile applications using ad networks and exchanges are four times as likely to have malware than those sourced directly from publishers, according to an analysis of advertising malware and safety that also found Australia's rate of ad fraud is lower than that in comparable European countries.
The Integral Ad Science Q1 2015 Media Quality Report evaluated a snapshot of online ads and ranked them according to their TRAQ score, a proprietary measure of advertising quality that incorporates measurements of ad fraud, viewability, brand safety, ad clutter, and professionalism.
Australia's overall TRAQ score was 601 – suggesting ads were generally of lower quality than those in Germany (607), the UK (623), and France (632).
Brand risk – the risk that a brand will be damaged through the posting of inappropriate advertisements – was higher in Australia, at 17.7 percent of advertisements, than in Germany (17.4 percent), the UK (11.1 percent), and France (10.6 percent).
While Australia was also trailing France and the UK in terms of ad viewability, the percentage of fraudulent ads (8.7 percent) was behind that in other countries: 12.9 percent of ads served in the UK were fraudulent, compared to 11.2 percent in Germany and 11.4 percent in France.
Taken as a whole, 14 percent of ads evaluated were fraudulent.
The results are the first out of a dedicated Anti-Fraud Lab that the US-based company opened in Seattle in February, allowing it to conduct deep-dive investigations into advertising fraud and to watch fraud trends over time.
“Fraud has become an increasingly large issue in the advertising industry, which is why we are dedicated to developing the best technology and tools possible to help combat it,” Anti-Fraud Lab director Jason Shaw said upon the lab's launch.
“The capabilities of this lab will enable us to disrupt ad fraud and other deceptive behaviour in all its forms, as well as participate in exciting efforts against cybercrime around the world.” Between-group differences were marked, with ads posted through advertising networks and exchanges earning a TRAQ score of 551 – suggesting they were of less quality overall – and publisher-posted ads averaging a TRAQ score of 673.
Read more: Automation key to defeating new adversaries
Advertising networks were also judged to be riskier overall, with ad fraud at 16.5 percent compared with just 4 percent fraudulent ads served through publishers.
Ad networks had more ads promoting illegal downloads and illegal drugs, making them riskier when it came to protecting brands (13.7 percent brand risk versus 9.7 percent for publishers). Their ads were also significantly less viewable – 41.8 percent versus 52.8 percent for ads published by publishers.
This article is brought to you by Enex TestLab, content directors for CSO Australia.