By Giovanni Strocchi, CEO, ADmantX.
“We will only invest where we have proof it is safe, efficient, effective, and high quality,” asserted Procter & Gamble’s Marc Pritchard when evaluating the state of digital advertising earlier this year.
His declaration reflects a growing awareness across the global marketing industry of the quality and brand safety risks associated with programmatic advertising when the right controls are not in place. YouTube’s high-profile scandal earlier this year forced advertisers to pay more attention to unsafe placements.
Almost four-fifths (78%) of marketers now realise the potential damage that negative adjacency can cause to brand reputation, and over two-fifths (43%) have experienced issues following a poor placement.
But by narrowing their focus to the multifaceted and highly subjective issue of brand safety, marketers are missing a larger opportunity. To get the best out of their programmatic campaigns, advertisers must take a step back and look at the wider theme of contextual relevance, rather than building their programmatic strategy solely around brand safety considerations.
So why – by honing in on brand safety specifically – are advertisers missing the bigger picture? And how can broadening their view to include contextual relevance increase their confidence in programmatic advertising?
What does brand safety really mean?
There is no single definition of brand safety because a placement that is harmless and sought-after for one brand could prove to be toxic for another.
Even within a single vertical, different definitions of brand safety can apply. For instance, an ad placed alongside a negative news report on the environmental impact of air travel would be disastrous for an airline brand, while transport providers who offer an alternative means of travel – such as coach tour operators – might actively pursue the placement.
Equally an article linking sugar intake with the obesity epidemic would be a no-go for most soft drinks brands, but sought-after content for a producer of low-sugar smoothies.
When adopting a programmatic strategy, marketers need to understand what safety means for their brand specifically – which may not be the same as the other companies with whom they are competing for ad placements.
Basic brand safety tactics are too primitive
Marketers have responded to brand safety concerns by adopting techniques such as blacklists and whitelists or basic keyword filtering at page level. However, these tactics are primitive, allowing unsafe placements to slip through the cracks while also inhibiting reach and scale.
Blacklists and whitelists are unsuccessful at guaranteeing brand safety as they only assess content at domain level. Just because a domain is considered safe doesn’t mean the page-level content will be. Equally, just because a domain is deemed unsafe doesn’t mean the content on every page will be inappropriate.
Relying on keyword targeting and filtering is often unfit for purpose due to the limited precision of those technologies, which focus on individual words and phrases that can have multiple meanings, and can be used in a variety of contexts.
For instance, a kitchenware brand advertising its new range of knives could easily end up advertising alongside a news article about a rise in knife crime – far from ideal. But if the same brand decided the word ‘knife’ was categorically unsafe and avoided placing ads alongside content that included it, they would miss out on a wealth of potentially valuable and relevant cookery-related content that would attract the eyeballs of their target audience.
Damaging content can be split into two types. The first, standard objectionable content, consists of the content the majority of brands will want to steer clear of – relating to alcohol, illegal drugs and adult themes, or negative content that relates to current events such as natural disasters, violence or weapons.
This is problematic for brands because they require a “complete filter” of these themes. The second is industry or brand specific negative content environments posing an even bigger risk for brands – for example, a soft drinks brand advertising on a diabetic website.
No keyword list, blacklist or whitelist can accurately manage correct placements for each individual brand.
Instead, a move towards “intelligent brand safety”, capable of providing a deeper level of understanding, should be adopted. Implementing custom categories that are specific to a brand can avoid unsafe ad placements, relying on data around entities, page sentiment, emotions, and intentions. By tailoring brand safety requirements and avoiding blanket techniques it is possible to achieve a higher level of brand safety.
Relevance brings inherent brand safety
Instead of relying on out-dated brand safety tactics, marketers should take a more organic approach to ad placements, analysing content at page level, in the pre-bid environment. By employing tactics such as semantic analysis and natural language processing, they can read and understand a webpage just as the human brain would, revealing the true meaning and context of the content as well as the emotions and sentiments it evokes.
This type of analysis allows marketers to decide if content is relevant to, and appropriate for, their brand and audience regardless of its domain or whether certain keywords are present or not, and will automatically ensure the placements they do bid on are relevant for their audience and, at the same time, fully brand safe for that specific advertiser.
When brands understand the contextual relevance of on-page content it will not only allow them to avoid negative adjacency issues, but also enable them to target the placements most likely to attract their target audience attention, while they are in the right mind-set to receive and respond to brand messaging.
Bidding programmatically, in conjunction with contextual analysis, and aligning ads with relevant content, marketers can enhance the overall effectiveness of their campaigns while also ensuring brand safety.
Brand safety and contextual relevance have a similar baseline, but very different outcomes. Once brands see the bigger picture of relevance, rather than just focusing on brand safety and employ contextual analysis technologies at page level, they will be able to achieve Marc Pritchard’s progressive vision of placements that are innately “safe, efficient, effective, and high quality.”