Counterpoint: Ad Verification Doesn’t Have to Be a Cure-All

Recently, Peer39 CEO Andy Ellenthal wrote that ad verification is no cure-all in the wild and untamed online advertising space. And while it’s certainly true that ad verification is too often a reactive system, there is technology that allows for a more proactive strategy. If the goal is to be truly protective of brands, that does indeed call for a more preventative approach. And the Web is not as much of a Wild West as it may seem. It is no longer a new frontier.

Preventative brand protection requires moving beyond basic ad verification. Simplistic binary scale systems, as well as pixel-based, sliding-scale ad-blocking systems, are not all encompassing. Errors do slip through them. But determining whether an ad’s potential location is “safe” for a brand has been possible for a couple of years now. Moving away from simplistic site and binary domain-level categorizations by assigning numeric grades is actually the key to achieving proactive brand safety. Further, domain-level ratings may fail to take into account the specific information needed to determine true safety. Instead, page-level analysis and scoring against brand safety is important to avoid the false positives Ellenthal discusses. With that said, preventative solutions that work well on a page-level do exist. One just needs to know the right questions to ask in evaluating the technology.

It’s hard to tell the difference between what’s safe and what’s not by just looking at the words on a page. Technology has evolved far beyond this simplistic approach, which can overlook a lot of important and relevant data. It’s only possible to create an accurate and comprehensive rating of page content by evaluating all of the content on that page. The best approach involves weighting competing evidence sources that include, but are not limited to, related pages such as links on the page, human scoring, semantic filters, image analysis, site-registration information, HTML-source code evidence, auxiliary constructed variables, and search-engine toxic URL lists.

This model is capable of differentiating that Obama shirtless on the beach (the Ad Exchanger reference Ellenthal points to) isn’t adult content. This method is built to prevent false negatives in addition to false positives, because if too many impressions are blocked, campaigns lose scale and publishers lose revenue — and that’s not the goal either.

Each brand has its own content guidelines. That’s why we also shouldn’t try to retrofit contextual models onto brand safety. By taking into account much more than just domain, malware and semantic data, and including refined and specific page content data — customized to a brand’s choice of safety thresholds — it is possible to build on ad verification and take a truly preventative approach at scale.

As the industry rapidly moves to take advantage of real-time trading, brand safety becomes a more important conversation every day. Brand protection within the auction environment has historically been difficult to assure given the lack of transparency. But there are now solutions that offer pre-bid brand safety as well as more contextual data and services to both the buyer/DSP and seller/SSP sides in order to prevent wasted impressions.

The focus is not on selling out the publishers, which is how ad verification tends to be viewed. Instead, it is on moving beyond ad verification to provide true brand safety and protection in an efficient and comprehensive way that benefits all parties.

Scott Knoll is CEO of AdSafe Media, a provider of brand protection and ad performance services.

https://staging.digiday.com/?p=3319
Digiday Top Stories
Ad rendering preventing in staging

Ad position: web_bfu