The average "innocuous invalid rate" in Anchor Intelligence's network declined to 29.8% in the second quarter of 2010, representing a 36.1% sequential decrease. That means the invalid traffic rate dropped to less than 1% from 7% in the prior quarter, according to the company's Traffic Quality Q2 2010 report released Monday. But the decrease does not necessarily mean good news for advertisers.
Anchor analyzes multiple types of advertising, but predominantly driven by CPC, search ads. While the decrease presents an interesting trend, it does not represent a decline in malicious activity across the Web, but rather a decline in robot traffic such as Web site crawlers and spiders designed to index Web sites as well as a sharp decrease in internal testing traffic like test clicks from ad networks and search engines. The report suggests that since innocuous invalid traffic is generally pre-filtered by ad providers, it rarely impacts an advertiser's ad budget.
While neither the innocuous-invalid rate or the attempted click-fraud rate bode well for advertisers, the malicious intent of the click-fraud invalids generally are worse "because in those instances you're dealing with groups of bad actors who spend lots of time, money, energy, and resources to steal money from the advertising community," says Ken Miller, Anchor's chief executive officer. "Whereas the innocuous is more of an annoying occurrence that many companies can manage, the malicious activity requires systems that are as good as those built by the bad guys, and most companies aren't equipped to do that."
The attempted click-fraud rate remained nearly unchanged, decreasing from 29.2% in Q1 to 28.9% in Q2, but year-on-year the percentage jumped 26%. The Traffic Quality Report aims to provide online advertisers with insights into the magnitude of traffic quality issues observed by Anchor, while providing clear and accurate definitions for the statistics provided.
Another trend that Anchor identified points to targeting smaller, less established ad networks and search engines that may on average become more vulnerable to attacks. It has become apparent that some customers consistently record increases in volume while also experiencing decreases in attempted click-fraud rates. This trend indicates that advertisers are seeing improved performance on ad networks and search engines that defend networks and sites against click fraud.
Anchor also measures click fraud by country. Of the top 30 countries, Vietnam took No. 1 at 37.3% in Q2 when it comes to attempted click fraud, up from 35.4% sequentially. Anchor attributes this to automated, high-velocity traffic from a distinct number of IP addresses as well as traffic from users displaying historically suspicious behavior.
Vietnam's Internet infrastructure is more vulnerable to attack than those of other countries because most computers run on outdated operating systems and Web browsers. Furthermore, the click-fraud attempts originating out of Vietnam in Q2 were believed to be a result of machines that were compromised during politically motivated cyber attacks against blogs criticizing mining projects in Vietnam.
Australia and the United States remain at No. 1 and No. 2, respectively, when it comes to the highest attempted click-fraud rates. The click-fraud rate in Australia increased from 35.2% in Q1 to 36.4 in Q2, and the attempted click-fraud rate in the U.S. decreased from 35.0% to 34.0%.
Click fraud is not the only malicious online activity advertisers need to keep an eye on. Video as a promotional advertising tool will begin to attract more mischief by those who want to exploit the Web and its content.
Anchor has successfully tested its ClearMark technology on video ads, but they don't make up a large percentage of the ads the company monitors, mainly because client demand has not yet reached the medium, Miller says. It likely will eventually.
Hackers took advantage of a cross-site scripting (XSS) vulnerability on Google's YouTube Sunday, hitting sections where users post comments, according to Network World, an IDG News site. The hack focused on clips related to pop star Justin Bieber. It didn't involve malware, but the code did affect YouTube pages so visitors saw tasteless messages or redirected to external sites with adult content.