In the context of the Covid-19 outbreak, an overreliance on ‘negative’ and ‘blunt’ brand safety techniques - such as keyword blocking - is costing advertisers and publishers heavily.
Recent data from Newsworks estimates that news publishers will lose somewhere in the region of £50m in online ads from April to July, due to brands’ over-zealous use of keyword blocking on any content related to Covid-19 – positive, negative and educational.
In a recent Drum panel discussion on brand safety, Harvin Gupta, UK director of solutions engineering at Xandr, and Tina Lakhani, head of ad tech at IAB UK, outlined the key problems that the pandemic has highlighted around issues of brand safety practice and programmatic advertising.
Gupta believes that brands’ ‘broad stroke’ approach to content safety, where a brand may choose not to appear next to content featuring ‘death or disaster’ for example, is simply not sustainable in a world where Covid-19 is “the main story in town”.
Lakhani agrees that brands need to move quickly beyond reliance on blunt technologies such as site exclusion lists and 10,000-word blocklists, pointing to best practice guidelines the IAB UK has produced. Despite a massive spike in traffic during the early days of the lockdown, many publishers were frustrated by the keyword blocking practices of their advertisers that left them unable to monetize their increase in audience.
Safety or Suitability?
Technological leaps in areas like AI and semantic understanding mean that brands can now foster much more interesting and nuanced conversations around brand safety, says Gupta.
Advances in tech are also raising awareness of how the context of any single word – ‘shoot’ for example – can have a range of positive and negative meanings in different settings. Lakhani believes that tech tools which help to contextualise content are becoming increasingly crucial to both advertisers and publishers alike.
Technology alone cannot guarantee success, argues Lakhani. Brands need to work more closely with their content verification tech providers in order to better understand the options these new tools open-up to them.
Gupta highlights Mantis, a semantic solution created by publisher Reach plc and powered by IBM Watson’s Natural Language Processing, that contextualises and verifies news content, marking it safe for brands. Mantis is able to filter content for positive/neutral/negative sentiment and package up positive and neutral content for purchase on Xandr’s platform.
Gupta believes that the role of publishers, as the gatekeepers of a ‘logged-in environment’ and its first-party data, will become more important in the months ahead, leading to a new era of ‘Contextual 2.0’ campaigns, stating these can perform just as well as some of the audience campaigns of the past.
However, Lakhani believes that increased regulation and the prospect of “a cookie-less world” mean that establishing a meaningful feedback loop will become harder to achieve. In the age of connected TV and OTT video, she expects to see technology tools emerge that help to establish context without relying on words at all.
The message is clear: if your brand safety strategy is still based predominately on a keyword blocking strategy, it’s time to talk to technology platforms about the other sophisticated options and tools they can offer.