How to handle the challenge of brand safety in RTB

More and more brands are adopting programmatic buying to efficiently purchase online media, turning to DSPs and agency trading desks to access audiences on thousands of sites across the web. But as brands buy across multiple sites with no way of physically seeing the actual pages where all of their ads served, they're investing more in brand safety.

As with all other segments of the online advertising ecosystem, the brand safety segment is crowded with different approaches, and a lot of confusion about what "brand safety" actually entails. Right now, there are three dominating schools of thought for brand safety: verification, ad blocking, and preemptive brand safety, each with its pros and cons.

Verification is the oldest mechanism in the digital space and sprang from the bifurcated online media landscape. When advertisers began buying from multiple publishers across hundreds of sites, they needed an independent party to prove that ads actually appeared where the IOs stated they would. Think of the tear sheets used in magazine advertising. With the rise of biddable media through exchanges and then RTB, advertisers needed even greater assurance that their ads were actually served, and it gave verification companies an additional value proposition. These companies now report back on aspects of where ads ran, in what type of sites, and in some cases, pages. Armed with this information, brands could determine whether or not the environments were safe for their ad message and use that information to get goods from their ad network and publishing partners.

Blocking came about as advertisers recognized the need to be able to decide on the fly whether or not an ad impression was appropriate for their brand. Blocking is used in the creative and will basically give the requesting ad server a go or no-go decision on serving an advertiser's ad based on a blacklist, geo-target, keywords in the URL, or some standard objectionable content list. This is a more active approach to brand safety, but still has some drawbacks. Blocking has limited capabilities for identifying content on the fly, and it's costly to the advertiser who has already purchased the ad, or to the publisher, who has to take the impression back after it is rejected. Still, blocking does a better job at addressing the messiness of reconciliation that verification deals with.

Then there is preemptive brand safety, which takes blocking to the next level by allowing advertisers to decide pre-bid if an ad impression is not safe. In order to do this, the safety provider needs to be integrated into the ad server or decision engines that participate in RTB.

As RTB grows in usage, advertisers are making use of all three strategies to ensure that their ads don't show up in objectionable environments. Lately, verification has turned into a tool that is used to produce blacklists of sites. Advertisers use the verification data to add domains to a blacklist, thereby ensuring their ad never appears on that site again. There are companies in the space that produce lists of URLs rather than domains, but this is of little use because DSPs rely on groups of sites, rather than URLs, when building white and black lists.

Blacklisting may seem like a sound strategy, but it's ultimately futile. Barring entire domains is too reactive of a strategy: It simply eliminates entire swaths of sites because the ad server was unable to detect an unsafe environment the first time. Even if advertisers could overcome the difficulty of blacklisting based on URL, we've seen that 90 percent of the pages that show up in RTB are new every five days, so there's little likelihood the brand would ever encounter that URL again anyway.

RTB technology is supposed to automate the buying process, but combing through pages of post-campaign analytics looking for bad URLs adds unnecessary inefficiencies. Because preemptive technology is deeply integrated directly into an ad server or RTB agent, it can utilize page data to avoid serving ads to certain pages in a far more granular approach than blacklisting -- it's essentially a dynamic blacklist of pages, as opposed to the exaggerated blacklist of entire sites. The fresh data means brands can avoid entire content categories, even within the new inventory that makes its way into the exchange every day.

Advertisers are still going to use a variety of offerings to achieve peace of mind, but they should understand that preemptive safety is a comprehensive solution on its own. The value of brand safety is about finding the best content for the advertiser. It's about addressing concerns about ending up on bad pages on good sites, and mining good pages on seemingly misaligned sites. If this is a persistent problem for an RTB buyer, then preemptive safety is the ultimate solution they are looking for.

Alex White is GM of data and trading at MediaMind.

On Twitter? Follow iMedia Connection at @iMediaTweet.

"Road sign, against panoramic blue sky. Take care." image via Shutterstock

 

Comments