Ahead of a contentious 2020 U.S. presidential election, two trade bodies representing media agencies and publishers are working toward more sophisticated brand safety standards and how to apply them to news content.
These organizations—the American Association of Advertising Agencies and Digital Content Next—are devising a standardized means of analyzing sentiment on recognized news websites to better assure advertisers their ads aren’t served next to hot-button political content.
In particular, the group wants to establish a consensus whereby the industry’s major content verification providers can identify the context, sentiment and tone of news items, then use software to help advertisers decide if such stories are suitable for their brand.
Discussions first began in Q1, with subsequent meetings taking place throughout the year. The group’s next meeting is slated for early Q4, when they will attempt to gain agreement from publishers to implement their proposals, sources tell Adweek.
This initiative is jointly conducted with oversight by the 4As and DCN, and with participation by three of the industry’s major content verification providers: Oracle’s Grapeshot, Integral Ad Science and DoubleVerify.
The group doesn’t have a formalized name yet, although sources told Adweek that “Trusted Media Sustainability Initiative” is a working title—with TrustX, a private marketplace where DCN’s publishers can trade inventory programmatically as a collective, looking to lead its implementation.
TrustX CEO David Kohl tells Adweek the scheme was geared toward increasing confidence in using software to automate placing ads next to the finance, health and politics sections of its members’ websites.
These new tools would gives verification companies “a better way to align content with ad revenue,” Kohl said, adding that many advertisers rely heavily on keyword blacklists to err on the side of caution.
“It’s a complicated, big topic, and while safety is a floor, suitability is subjective,” he said. “But an over-reliance on keyword ad domain blocking is just an under-use of the A.I. that’s now available.”
In total, more than a dozen entities are involved, with each party hoping it will help infuse quality news organizations with much-needed ad revenue as the U.S. general election—historically, a boom time for news brands—draws near in 2020.
However, this cyclical bounty has become less reliable in recent years, particularly as U.S. politics have become increasingly polarized and certain brands employ a risk-averse approach to their media strategies.
Brands don’t trust the most trusted news brands
The topic was discussed at length at a conference session as part of Advertising Week New York, where panelists representing the three major tiers of the industry debated the evolving nuances of brand safety in a polarized political climate.
Christine Cook, senior vice president at Warner Media and chief revenue officer of CNN Digital, said her organization faces business challenges despite record audience growth because of advertiser concerns about being seen as overtly political.
“Advertisers say to us that even though we’re premium and the No. 1 trusted brand for the people that come to us, it isn’t safe for their brands to be there,” Cook said. “That is definitely a challenge for us.”
Blocklists hurt legitimate news sites
Fellow panelist Susanne Mei, senior vice president of CBS News Digital, went on to “echo the disconnect” of the increased audience numbers not necessarily equating to more ad revenue because some media buyers avoid placing ads next to news stories without knowing what their content will be.
This is a tactic that often involves employing keyword blacklists, such as “gun control,” and is often used as a means of steering brands away from disreputable “fake news” sites, but often also sacrifices placements on trusted news channels.
“On the one hand, we have this great story around consumption and being America’s No. 1 trusted news organization … but for whatever reason, [marketers] are afraid of the whole [news] category,” Mei said.