Is there a way to resolve conflicting needs and make the ecosystem a safer place? Nick Welch investigates
YouTube is at last taking action on brand safety. After Logan Paul’s controversial video became the latest in a series of scandals, the media giant has promised to implement tighter content vetting, stricter monetisation policies, and more transparent advertising options.
But while YouTube’s move marks a positive change, the wider issue of brand safety persists. For CMOs and publishers in particular, it’s difficult to find a solution that offers robust brand protection and maintains ad revenue flow for content producers. And this is driving tensions.
So, is there a way to resolve these conflicting needs and make the ecosystem a safer place?
The CMO’s perspective
For many CMOs, the last year has proved a vivid demonstration of what happens when one of their worst fears is realised, as it was revealed ads for big names such as Cadbury, Land Rover, and Sandals had unwittingly appeared beside extremistand inappropriate content. The damaging effects on brand reputation from poorly placed ads is all too clear; with a CMO Council study finding that more than one-thirdof consumers rethink purchases when they see brands advertised next to objectionable content.
Consequently, it’s no surprise that frustration with current brand safety efforts has increased and key players are advocating affirmative action. For instance, Alicia Tillman, CMO at SAP has encouraged reducing ad spend until better controls are applied and marketing chief at Abode, Ann Lewnes, has called for stricter overall governance. But these are far from isolated views. According to research from Teads, 95% of CMOs have already adjusted digital strategy amid concerns about brand safety, fraud, and transparency. The majority (93%) also intend to select future suppliers based on their brand safety abilities, and 41% are contemplating in-house media buying.
To tackle this issue, many CMOs are reliant on keyword lists — a list of terms or words in isolation that brands either do or do not want to be linked with. In principle this provides a blueprint for secure ad placement, but in practice it can have a detrimental impact on monetisation.
The publisher’s perspective
Publishers are concerned about brand safety too. After all, revenue from the ad space brands purchase is integral to funding content creation. But some protection methods — such as URL or site blocking based on flagged keywords — are proving increasingly restrictive for publishers, and their audiences.
The problem is that as fear of ad misplacement grows, so does the list of potential keywords that pose a threat. And each time the catalogue of terms brands want to avoid expand, the volume of content publishers can offer as ‘safe’ digital advertising inventory shrinks. False-positives are the latest threat derived from both negative and positive keyword targeting. Here, entire URLs are automatically blocked due to one keyword being found, irrespective of context or sentiment. So, for example, headlines such as “IKEA Kills It in Retail”, which are positive in context, are not properly understood by keyword engines and as a result the web page is blocked, despite being perfectly safe.
More troublesome still is the fact that brand safety is such a subjective concept: what counts as acceptable for one company, could be risky for another. An ad requesting charitable donations might be appropriate next to editorial on a humanitarian crisis for example, while ads offering cheap holidays would not. Ultimately, blanket lists of keywords are getting longer, decreasing monetisation options for publishers and creating false positives for brands whose messages are targeted away from content that in reality pose no threat to them.
Plus, this trend is also driving publishers that rely on ads as their main source of income to cease production of content containing banned keywords entirely. And, in turn, this means audiences who visit certain sites for hard-hitting news or insightful articles on favoured topics will no longer find the content they are searching for – it simply won’t exist.
How can brand safety be achieved?
In short, it’s vital to put a greater emphasis on written comprehension and understanding the true context of content. The meaning of words change depending on how they are used, for example: ‘cat’ could be an animal or leading manufacturing company. So, simply compiling lists of words that are deemed generally unfit for advertising is too broad an approach – this lack of ability to understand context (the very thing they are employed to deliver) being the main problem with keywords.
Instead, both CMOs and publishers need to define what terms mean in their specific context. And, at the moment, the best way to do that is deploying semantic technology designed for granular analysis. By conducting in-depth content assessment at a page level — using techniques such as Natural Language Processing (NLP) — such platforms can assess words in the same way the human brain does; picking up subtle differences in meaning.
Primarily, this means that these tools can evaluate content and judge whether its sentiment and subject matter make it safe for a particular brand; thereby providing the insight needed to prevent ad misplacement. Yet it also has other benefits. By lifting restrictions previously imposed by large keyword lists, CMOs can extend their advertising reach in the knowledge they are not risking their reputation, while improving both yield and freedom for publishers. Plus, with greater insight into the sentiment of content and the audiences it is likely to attract, the precision of ad targeting can improve too.
Of course, semantic technologies are not in themselves a panacea for brand safety. To clean up the industry, all members of the digital ecosystem need to keep striving to enhance media standards. But in utilising such tools, instead of exclusively depending on limited keyword lists, CMOs and publishers can create a safer and more sustainable advertising system.
By Nick Welch, VP Business Development UK and Northern Europe, ADmantX