You didn’t return any results. Please clear your filters.
18 December 2020
New research shows inadequacy of brand safety guidelines
Brand safetyUnited States
Accidentally exposed data from three brand safety technology providers reveals the mess of blocklists and risk levels that decide to monetise or block articles.
Why it matters
Revelations made by Adalytics’ Dr. Krysztof Franaszek, and surfaced in the Branded newsletter, means that advertisers and publishers should seriously question the methods through which these technology firms – including Oracle, Integral Ad Science and Comscore – are making decisions that affect who wins and who loses on the internet.
The big problem is that important news, though depressing, is often considered unsafe for advertisers; 30% of New York Times articles were considered by Oracle to be unsafe.
Where they appear to fall down is in filtering extreme far-right, white supremacist and disinformation content.
There’s little consistency: although consensual definitions exist, brand safety systems do not agree.
Funding quality news and entertainment is the collective responsibility of the industry, when automation appears to compromise safety while cutting a supply line to important reporting, brands must look at the detail of their guard rails.