Powering effective campaigns with better contextuality

In today’s marketing world, it’s all about context. While the internet and modern search tools have been a gamechanger in helping brands target their messages across multiple platforms, it’s also created a sort of ‘Wild West’ scenario when it comes to the sharing of information, which can lead to cases of ads being placed near content they were not meant to be seen with.

At best, such a situation can generate annoyance for the users during their browsing experience. At worst, it may lead to advertisements being placed next to controversial and incendiary content.

It’s therefore no surprise that brands and advertisers are now placing greater emphasis on brand safety then ever before. The days of brute force numbers generated via programmatic advertising are coming to an end, as illustrated regularly in cases where global brands have pulled millions of ad dollars from popular online platforms due to the growing number of incidences of misplaced ads.

Brands and advertisers today are not just expected to understand who their ads are being shown to, but also what sort of content the ads are being placed next to.

Keeping an ear to the ground

To everyday consumers, an ad being placed next to a certain piece of content may mean that the brand is supportive of that content. An energy sports drink brand may run their ads with the intention of reaching as many audiences as possible, but they risk running afoul of some segments if the ads appear on a social media page that advocates content that those audiences don't agree with. In a perfect scenario, that brand would like their energy drink ad to be placed next to content featuring sports stars, for example, but all too often the results have proved otherwise, due to the technological limitations of the tools being used.

Jaguar Land Rover was one of several brands that, in 2017, had ads unintentionally placed alongside highly controversial online content

To be fair, such erroneous placing of ads is not solely the fault of advertisers or agencies. It has more to do with issues with the algorithms, which have failed to protect brands. However, to the average user who is unlikely to know the mechanics behind targeting advertising, the damage will already have been done and they might associate a brand with content that they’re against.

Avoiding this scenario requires brands to take a more active role in finding out where their ads may end up; and many are choosing to do so via 'social listening'. This allows brands to track, assess and respond to conversations about their brand and industry online. It goes a step further than 'social monitoring' – whereby brands comb social media for messages specifically targeted to the brand and respond to them – to understanding audiences by analysing a fuller spectrum of conversation around the industry and topics related to the brand.

How can we listen to audiences better?

What social listening does is help brands to create more effective campaign strategies. It places greater emphasis on brands understanding their audiences deeply in order to place their ads with better context, and hopefully avoid them appearing next to content that may push some audiences' wrong buttons.

How can this be done, exactly? Programmatic tools and algorithms can seem daunting enough but taking a more active role in knowing what audiences really want requires more sophisticated powers of tracking conversations and behaviours.

This is where AI and machine learning come into play, helping us make better sense of the droves of complex data that brands need to sift through to understand audiences and get the right messages out. By leveraging these technologies, we’ve developed a tool that uses a blend of AI and computer vision to identify the specific context of video content, and accordingly serve in-video ads that align with the advertisers’ core messages, whilst enhancing the user experience.

To illustrate such applications, we worked with a popular global fast food chain on the launch of a new product. During the campaign, our tools were used to identify 30 contextual triggers by detecting specified faces, logos, emotions and objects.

For example, in the trailer for the blockbuster Avengers: Endgame, our technology detected the faces of the movie’s cast during heated battle scenes, which then generated an ad copy showcasing the fast food chain’s new product. Simultaneously, our technology detected the logos of content providers like Netflix to generate customised ads during pre-roll.

In another instance, our tools were used to achieve better context with ads amongst specific audiences. For the launch of a new drink by a global drink manufacturer, we sought to generate more brand awareness between adventure-loving audiences. We used our AI technology to detect the faces of the brand’s own ambassadors, popular action heroes and objects aligned with ‘adventurous’ brand imagery to deploy ads during pre-roll.

What both examples show is that, by using more sophisticated tools, we can better contextualise our ads and keeping the rustling of feathers to a minimum. In the first case, our tools allowed the fast food brand to communicate their ads less overtly and ensure minimum disruption to the audience when viewing their content, whilst the second case showed that we can tap into the interests of adventure-seeking audiences to position the drinks brand as a supporter of such activities.

In the current age of marketing, greater pressure has been thrust on brands to ensure that their ads appear during the right situations with the right audiences. To their benefit, there is a fast-growing number of solution providers that have the technology to help brands cut through the online noise, understand their audiences better and help guide them deliver more effective ad campaigns.

With this variety of tools available, the onus now is really on the brands to ensure that their ads and their messages appear in the right context.

Kartik Mehta is chief revenue officer at Silverpush.