Alexi Mostrous — author of the Times’ exposés covering YouTube’s brand safety issues last year — has been hailed as the hero digital advertising needs. And to an extent, this is true. Without Mostrous bringing ad misplacements into public view, the industry might not have been forced to prioritise the online safety of brands and the correct placement of ads.
But while the current active focus is positive, confusion about whether ads are safe or not remains. For example, programmatic is often linked to brand safety concerns due to the perception it limits control over where ads are displayed. Yet automated technology is becoming fundamental to fast, large-scale digital ad trading: amassing £3 billion in spend last year alone.
So how can CMOs utilise efficient tools to expand reach in 2018, while keeping their brand secure?
The answer to this question begins with an investigation of brand safety; including what it now means, and why it matters.
Defining brand safety
In short, brand safety describes the process of gaining full transparency into online media so that ads are not placed within or beside inappropriate or potentially harmful content. The longer explanation is more complex. Each brand has its unique offering, vision, and values, so there is no prescriptive list of unsafe media. But generally speaking, brand safety can be defined as steering clear of content that is: illegal (in terms of copyright or subject), salacious, offensive or directly opposed to specific brand principles and image.
Why is it important?
The key reason brand safety matters so much is simple: reputations are fragile. A good name can take many years and significant investment to build — for brands and publishers — but seconds to crumble. With just one misplaced ad, or low-quality ad served on a premium site, negative associations can be created that damage audience relationships; reducing advertising returns for brands, and revenue for publishers.
Most CMOs are aware of this risk, with studies showing 78% believe inadvertent links with unsuitable content could diminish brand reputation.
Most CMOs are aware of this risk, with studies showing 78% believe inadvertent links with unsuitable content could diminish brand reputation. Yet there is a gap between this knowledge and implementation of the requisite defences; 22% of CMOs in the same study have seen their ads placed alongside compromising content.
It seems that, perhaps due to ease or familiarity, many brands are still relying on legacy tools that don’t provide sufficient protection for digital ads – such as keywords searches. And this dependence on outdated technology may also be what’s driving the view of programmatic as a hazardous mechanism.
Old tech for a developing issue
Keyword searches have long been a common means of deciding where ads should be placed. And their popularity has soared amid the rise of programmatic as they allow environments to be checked at scale: enabling brands to instantly deliver ads en masse and stay safe — at least, in theory.
In practice, however, the scope of keyword lists is either too broad or narrow. On the one hand, their generic nature means specific words that an individual brand may want to target away from are frequently not included, raising a considerable risk of ad misplacement. While on the other, their vast catalogues of terms inevitably contain words that are actually safe for some brands, which creates false positives — in which placements are categorised as unsafe while they are actually appropriate — and means valuable advertising opportunities might be missed.
If CMOs want to maintain audience bonds, profitability and a positive brand image, they must therefore upgrade to smarter tools that can meet their individual security needs.
New tech for a tailored brand safety
To avoid restricting advertising reach and ensure the right risk areas are covered, CMOs need to conduct fast yet tailored analysis, particularly when trading programmatically. And one of the easiest ways to achieve this is deploying tools built for automated pre-bid media evaluation, such as natural language processing (NLP).
A form of semantic technology, NLP combines the efficiency of artificial intelligence (AI) and human capacity to understand linguistic subtleties — reading content the same way our brains do, and noting the changing meaning of words depending on their context. The most sophisticated platforms offer this granular analysis at page level, which means they can identify the meaning and detailed sentiment of every potential ad placement.
For CMOs, the benefits of utilising such tech are twofold. Firstly, accurate and scalable content analysis allows them to leverage programmatic without worrying about putting brand reputation in danger. Secondly, keyword searches can be set aside; leaving CMOs free to maximise the amount of safe inventory they advertise against and reduce the number of false positives created by generic lists.
Though it’s worth noting that smart tech doesn’t negate the importance of human input. Until the industry develops a standardised definition of unsafe content, there will always be a degree of subjectivity when it comes to digital media assessment. Consequently, CMOs themselves will still need to play an active part in guiding AI platforms: identifying which specific terms their brand should target away from and programming accordingly.
Greater public consciousness of brand safety and the motivation it has inspired to better protect ads is undoubtedly a good thing. But so far, action has been impeded by a lack of understanding about what makes environments safe, and how to balance advertising scale with adequate security measures. Clearly, there is a requirement for industry-wide agreement on what appropriate media looks like to ensure safety needs are met. Yet for now, CMOs must focus on keeping their brand reputations secure by adopting advanced tech with the power to uncover the exact context of content before ads are placed; even if there are traded on the high-speed programmatic market.
By Giovanni Strocchi