Brand safety on a global scale and in real-time

It’s become all-too-familiar for digital ads to appear next to unsavoury content on the web. Just last week, The Straits Times ran an article highlighting ads for Singapore brands seen on incendiary websites. The article mentions that online display and video ads of some of Singapore’s biggest government agencies, banks, telecommunications companies, and airlines had been featured on sites that promoted hate speech and/or were deemed unsafe.

When contacted, most of the brands were unaware that their ads had shown up on such unsafe environments. Additionally, the brands say they regularly updated their online exclusion lists by adding in new sites, keywords, and categories. These two worrying points reflect an issue that often gets taken for granted: brand safety.

It's hardly a new concept, but there are several entrenched misconceptions that lead advertisers to protect their brands with simplistic or ineffective methods. This is especially true in trying to reach a global audience, since not all brand safety approaches work across markets to offer brands and CMOs the peace of mind they need.

Domain blacklist versus keyword approach

Some may believe that brand safety can be ensured by blacklisting domains wholesale. While blacklists are an effective short-term solution, they can fail to account for the mix of high and low quality pages found on most domains, which can then limit advertisers’ scale if the entire domain is banned. Bad actors also stay ahead of poorly maintained blacklists by making small changes to their domains.

Another way to ensure brand safety is to prohibit ads from running on sites with designated keywords; however, keywords can fail to account for critical contingencies. For example, ISIS is often abbreviated to "IS," which is a common word that appears on most websites. Moreover, words require context to discern their true meaning. A keyword list featuring "kill" can't tell if it's meant to be taken metaphorically (no issue) or literally (a big issue).

Dynamic solutions for a global marketplace

The most effective brand safety models go beyond keyword lists or blacklists by employing sophisticated techniques that leverage real data to categorize sites by a range of concerns to advertisers, such as: adult content; illegal downloads; alcohol; hate speech; violence; illegal drugs; and offensive language.

The sites are assigned scores that reflect the likelihood of inappropriate content, and are dynamically updated and verified, so brands can make informed risk assessments and reliably stop ads from appearing on sites that potentially threaten their reputation.

Additionally, these advanced, data-science-driven models must constantly adapt to meet the challenges of a globalized market by covering multiple languages.

New frontiers in brand safety

Video brand safety: As the amount of digital video content grows, there will need to be more oversight of its brand safety. Online video content can be just as risky as textual content, if not more. You don’t want your ad running as a pre-roll ad unit before a graphic beheading, for instance. TV buyers who oversee digital video budgets may have a false sense of confidence here, because they’re used to linear broadcast content, which is well-regulated.

Mobile in-app brand safety: Advertisers are only beginning to shift media dollars into the fast-growing app ecosystem, where there’s an array of various types of content: from social media, to dating apps, gaming apps, and news. And because this space acts differently than traditional web, the content doesn’t fall as easily into the same categories. In order to effectively analyse brand safety, brands should take a closer look at a user’s in-app behaviour and experience, but this also brings up privacy concerns that must be addressed.

Location-based brand safety: Location-based targeting is becoming more popular for advertisers as they look to connect their message with consumers in a specific time and place. The augmented reality game Pokémon Go became an overnight sensation, growing a massive audience in a matter of hours and led to amazing location-based options for ad insertion. What many may not realize is that, due to the nature of the game, consumers were sometimes led to inappropriate locations. As location-based advertising continues to grow in popularity, having the ability to quickly ensure ads show up in the right place will become even more critical.

With digital ad spending forecasts for Southeast Asia projecting double-digit growth for 2017 alone, and an ever-evolving and highly complex digital advertising ecosystem, it is even more important for brands to remain vigilant and protect their reputation. This is especially true in trying to reach a global audience, since not all brand safety approaches work across markets to offer brands and CMOs the peace of mind they need. 

A global marketplace therefore needs smarter, more dynamic solutions to protect brands. This will allow marketers to adapt to the needs, make the right decisions for their brands and keep up with the news cycle in real-time.

Niall Hogan is MD of Integral Ad Science Southeast Asia