Andrew Morsy, UK MD at Sizmek, says that in the dangerous world of digital advertising, artificial intelligence can help brands safeguard their reputation.
It was only recently that global brand and confectionary giant Mars made the decision to pull its ads from YouTube. Mars made the decision after it found that ads were being placed alongside content linked to violent crime in London. With Mars’ YouTube spend speculated to be around £5m, the hit is a big one to the video platform and is likely to have other brands looking at exactly how their own ads are being positioned. It’s certainly not the first brand to make this kind of decision; M&S, HSBC, McDonald’s and Audi all saw fit to pull ad spend on Google in 2017, with Cisco adding to the list this year.
While the swift move by Mars will have mitigated the long-term damage to the brand’s reputation, the incident is a further example of just how difficult the digital advertising landscape has become for companies to navigate. In fact, our own recent survey of more than 500 advertising decision makers found that 62 per cent of marketers feel the media landscape has become too complex. If a global giant like Mars – with its sizeable budget and extensive resources – can’t keep its brand safe, how can other businesses hope to keep their ads from being displayed alongside offensive or extremist content online?
With 85 per cent of marketers saying that improving brand safety is a high or critical priority, the issue is evidently creating headaches for companies of all shapes and sizes. The key to helping those companies is creating a culture of trust; trust between businesses and trust between the business and consumers. And that means taking responsibility, teaming up with the right technology partners and using AI to keep brands safe.
As digital advertising continues its ascent and takes an even greater share of the whole ad market, this trust will become even more important – otherwise brands risk losing the reputation and consumer loyalty that they have spent years building.
The battle between brand safety and campaign effectiveness
While our global survey showed that 61 per cent of marketers have a third-party brand safety solution in place, many are failing to keep their ads from being seen alongside inappropriate content. Despite the difficulties in telling if an ad has been delivered to a harmful web page, four in 10 marketers said that they believe their company has had its ads placed in a brand-unsafe environment.
Essentially, one of the biggest barriers to brand safety is the concessions that marketers often have to make regarding campaign effectiveness. The astronomical growth in digital media spend has resulted in a diverse, rich, digital marketing landscape, but the ability to achieve results from ad campaigns and convert audiences is often hamstrung by brand safety solutions.
In fact, 64 per cent of marketers agreed that the pursuit of brand-safe digital environments negatively affects the performance of their ad campaigns. So how can marketers ensure that they don’t have to settle for a trade-off between ad effectiveness and brand safety?
Putting AI at the heart of brand safety
Clearly, marketers are struggling to understand exactly where their ads are being delivered, along with the nature of the content on that particular site. This is where AI comes into play and will be a core part of a truly effective brand safety solution. As semantic analysis and natural language processing will be used to develop a human-like understanding of the meaning of content – helping brands identify fake news or damaging content and avoid brand safety issues – AI-based technologies will increasingly be used to detect and filter out fraudulent activity.
As we continue to see rising investment in digital advertising and programmatic, the role of AI will become even more critical; it will move from a ‘black box’ decision-maker to a trusted advisor that delivers insight and analysis about the sites it is targeting and why. Using the latest AI technology in the right way, therefore, doesn’t just make a brand’s ad “safer” – it is also capable of helping marketers uncover patterns of information easily.
Moreover, with machine learning able to do the heavy lifting, all this analysis of vast quantities of data will happen in real-time. This means that brands can both benefit from smarter campaign insights and feel reassured that they are operating in a safer environment.
Where does the responsibility lie?
In an industry as complex and sprawling as digital advertising, it’s no wonder brands are now treading more carefully over where, when and how their ads are placed. Mars may not have placed the ads on YouTube directly, but they certainly bore the brunt of the headlines; consumers will be quicker to recall the brand in question, rather than where or how the ad was placed. One recalls Warren Buffett’s famous quote about taking 20 years to build a reputation and five minutes to ruin it.
Another key piece in the brand safety puzzle, then, is ownership – and marketers are clear that the primary responsibility lies with the brand itself (38 per cent) more than anyone else. This is likely why 80 per cent of brand marketers have developed brand safety guidelines for their agencies and partners to follow. But this doesn’t mean that responsibility lies with the brand alone; media-buying agencies, publishers, DSPs/ad tech vendors and ad exchanges/networks all have an obligation to ensure they are doing their best to keep brands safe.
Unfortunately for many brands, one small mistake can have a devastating impact. From advertisements of ferry cruises placed next to images of the sunken Costa Concordia, to brands such as Mercedez-Benz and Waitrose inadvertently appearing next to extremist content, it’s easy to see how brands can get sucked into something they have little control over.
With 84 per cent of marketers hoping to find better value from their brand safety solution in the coming year, they should undoubtedly be looking to AI to deliver this value and keep their company from becoming the next news headline.