Since starting up its new anti-fraud tool that intercepts fraud, DoubleVerify says it has seen 62% fewer brand safety “violations” when it comes to video impressions across devices.

Starting four months ago, the digital media and analytics company says advertising categories using its Video Filtering software include travel, technology, consumer packaged group, finance and food services.

Looking specifically at each area, ad fraud and violations were 85% lower video in instances where bots and crawlers pretend to be legitimate users, and there were 88% fewer geographic violations. Specific platform fraud violation rates were down 66% for connected TV, 69% for desktop, 69% for mobile apps, and 71% for mobile web.

Before the launch of its video filtering software, DoubleVerify says fewer than 40% of video impressions were eligible for “post-bid blocking” on desktop and mobile web -- and on CTV and mobile apps this kind of blocking is virtually impossible.

More effective blocking -- by its video filtering -- allows marketers to re-expressed video impressions. DoubleVerify collects data from an ad request and runs it through the company’s detection models for advanced fraud, brand safety and suitability, and geographic issues.

The company says its video filtering tool can be used for campaigns directly through a publisher or through a DSP.