The Areas Where Brand Safety Measures Fall Short

The problem is that most of the tools in a marketer’s brand safety toolkit have defects. Here’s an overview of today’s most common brand safety tactics and their noteworthy negative shortcomings.

Third-party measurement

Leading social media platforms have finally made brand safety a priority. They’re embracing greater accountability and greater reporting responsibility. Platforms like Twitter and Facebook have opened themselves to third-party tracking solutions, making it easier for agencies and vendors to see where their ads appear. Utilizing third-party measurement is a sensible addition to any brand safety regimen. Unfortunately, it is by no means a standalone defense. After all, third-party measurement did nothing to protect YouTube advertisers affected by its most recent brand safety failure. There’s simply too wide a range of potential brand threats for third-party measurement to handle.

Direct relationships with publishers

In the programmatic ad buying world, brands often have trouble tracking exactly where their ads appear. That opaqueness is dangerous. Buying ad space through exchanges that have direct relationships with a publisher allows marketers to verify the safety of a publisher’s inventory and avoid what Dallas Lawrence, OpenX’s chief brand officer, describes as a “Wild West with fly-by-night technology vendors cutting quality corners and taking brand safety risks in order to drive short-term profits.“ As cut and dry an approach as working only with those listed on a publisher’s ads.txt approved list goes a long way.

So what’s the downside? Premium pricing. But more brands and agencies are embracing this publisher-conscious ad buying approach, so its brand safety value seems equal to its cost––at least for now.

Blacklists and whitelists

While the number of industry pros resorting to these types of brand safety tactics is declining, blacklists and whitelist are still a leading tactic for preventing negative brand exposure. Designed to exclude or selectively include ad inventory from specific sites based on the content they distribute, listing is potent brand safety. Whitelisting, which only allows ads to be placed in approved environments, may in fact be the best brand safety insurance.

But blacklist and whitelists may actually be a bit too potent. Many long-tail websites—like blogs—offer engaged, highly relevant target audiences that marketers employing the blacklist/whitelist approach will simply never reach. By screening sites that don’t have the status and reputation of an elite publisher, brands lose valuable audiences and make accurate targeting more challenging.

Natural language context detection and computer vision

AI that analyzes site context has been a recent boon to brand safety. If an auto brand wants to keep its ads from appearing alongside news stories about car crashes, natural language context detection is ideal. The problem is that the online expanse is increasingly visual. Text can be innocuous while accompanying images and video are brand-threatening.

That’s where computer vision helps. Though video is still a pain point for the technology, image recognition tools integrated into major DSPs have the ability to identify unsafe brand images. Both natural language processing and computer vision brand safety technologies can lessen dependence on crude techniques like blacklisting, allowing advertisers to safely reach audiences anywhere online.

But neither alone is enough. Multi-modal AI-powered context analysis technology, which can sweep an entire site’s images, video and text for brand-threatening content, is on the horizon. Until it arrives, AI context analysis cannot be counted on as a brand safety failsafe.

A safety sachem

Faced with all this brand safety imperfection, what are marketers to do? Recognizing that there’s no brand safety silver bullet is a good first step. The next takes a little more effort. It requires a meticulous and disciplined application of brand safety tactics.

Continue Reading