Google removed 2 million pages from its ad network each month last year due to content policy violations and 6 million bad ads per day throughout 2018 as part of its efforts to police the web for advertisers, according to a report released today.
The company is at the epicenter of how the web is monetized with advertising offerings for both the buy- and sell-side of the media industry. Those include content moderation and removal tools it showcases each year in the report, which trumpets its brand and user safety efforts.
In total, Google took down 2.3 billion “bad ads” in 2018 for violations of both new and existing content policies. That’s down from the 3.2 billion ads it removed the previous year, according to Scott Spencer, director of sustainable ads at Google.
Spencer described 2018 as a year of “new challenges” as scammers aimed to use online ads to defraud consumers offline.
Such fraudsters preyed on vulnerable communities, for example, by attempting to run ads ostensibly for bail bondsmen and rehab services, according to the report. “We introduced 31 new ads policies in 2018 to address abuses in areas including third-party tech support, ticket resellers, cryptocurrency and local services such as garage door repairmen,” Spencer said.
Those efforts included the use of machine learning to go after bad actors and the launch of 330 “detection classifiers” to better detect “badness” at the page level, according to Spencer.
“Using improved machine learning technology, we were able to identify and terminate almost one million bad advertiser accounts, nearly double the amount we terminated in 2017,” he said, adding, “So, while we terminated nearly 734,000 publishers and app developers from our ad network, and removed ads completely from nearly 1.5 million apps, we were also able to take more granular action by taking ads off of nearly 28 million pages that violated our publisher policies.”