The Washington Post Beefs Up Contextual Targeting in the Post-Cookie Era

The offering is called Washington Post Signal. It’s powered by the Post’s contextual targeting platform, Zeus Insights, which the company created in July 2019. The promise of Signal is more sophisticated ad targeting for buyers who aren’t reliant on third-party cookies. The added layer of first-party reader data is also meant to differentiate the capability from the basic parameters of contextual ad targeting. 

Advertisers, like launch partner Bank of America, can access Signal via its own dashboard. The tool monitors hundreds of consumption data cues, such as identifying where the reader came from. The program can also pick out the topic or article entry point (often a clear “signal” of intent) a reader first engaged with. In addition to highlighting all the articles a reader viewed, Signal gathers audience demographics, specific ad engagement and most frequently read topics.

All those data points are ultimately connected to nearly 2,000 of Zeus Insights’ contextual taxonomies on how it categorizes its content, such as topics like fashion or innovation. 

For instance, a brand running an ad unit featuring a Post editorial, such as the ones that appear under its PostPulse unit, can swap out articles quickly based on consumption data. Signal’s analytics can include whether the reader is more active on Tuesdays, or arrived from Facebook, whether they are politically engaged or interact more with video ads rather than slideshows. 

Signal’s launch represents the Post’s desire to act as a deeper partner with brands. As evidence, the Post notes that Signal audience data profiles can be used outside the publisher’s ecosystem. Still, Signal is a tactic to drive more publishers to join Zeus Insights and integrate with its Arc technology platform, which the Post licenses to publishers for a fee. To attract those outside content companies, the Post says Signal can expand publishers’ abilities to compete with the scale and targeting of platforms like Facebook and Google.       

“There are obvious challenges today, but we think about what will happen tomorrow,” said Jarrod Dicker, vp of innovation and commercial strategy. “[Signal] is built to be universal. We think about how we can enable the open web and premium publishers who have their own strategy. We think about how we can strengthen our products so it can be on [publishers’] owned and operated platforms. We can affect all that through product.”

The pendulum swings to context

Contextual targeting has been heralded as a replacement to third-party cookie based audience targeting, the engine powering digital advertising over the last decade. But basic contextual targeting—soccer ads on a sports site—won’t cut it. Publishers like The New York Times and Insider are making contextual tools smarter.

There’s a chunk of change at stake. The global contextual ad market is set to top $447.9 billion by 2027, according to a July report by market research Global Industry Analysts. It was the fourth most effective programmatic ad targeting tactic in a June study by content marketing research firm Ascend2, with 26% of marketers claiming its effectiveness (still, 73% named audience targeting as most effective).

Growth in contextual targeting is primarily being driven by marketers’ knowledge that cookies are going away, said Joe Root, co-founder of publisher data targeting platform Permutive. Complementing that, publishers like Hearst and Immediate Media are stringing together contextual and user action or consumption data to better understand and access audience insights. “That used to sit in the hands of the ad ecosystem tech vendors; now it’s in the hands of the publishers. That’s the real shift we’re seeing,” said Root.

The concerns hampering contextual growth are that fewer data points linked with inventory makes it less premium, generating lower revenue. As a replacement to third-party cookie targeting, ad frequency and popular but flawed models like last-click attribution are also toast.

To help solve the revenue issue, the Zeus Insights taxonomy is built so all publishers using the platform have access to the same keyword combinations used by the Post. That should make it easier for marketers to apply context and consumption data to execute the buy across the Zeus pool, said Dicker.       

While agencies welcome the ability to have the same taxonomy across multiple sites, the concern is that the creative message becomes broader. “There needs to be a balance between breadth and granularity of taxonomy,” said Chris Lehane, data strategy director at GroupM agency Essence.  

The Post wouldn’t share how much revenue it makes from licensing Zeus Insights. Since February it has signed up 100 new publishers including McClatchy, Tribune, The Seattle Times, Snopes and The Dallas Morning News. According to Dicker, across the board, publisher CPMs increased by 35% with viewability increasing 78%. During the coronavirus-related ad downturn, when news publishers reported reductions in open ad marketplace CMPs of between 30% and 40%, being part of a wider network with tight viewability standards has its benefits. 

Keeping media coupled with data

The company’s aim is for all of the Post’s advertisers to use Signal. The percentage of direct buys that already use Zeus Insights is “insanely high,” according to Dicker. Also, contextual targeting, which the Post has offered in earnest for the last few years, is tied to a “significant portion” of its direct sales efforts, although Dicker wouldn’t say how much.

“In the last six months, conversations have really picked up; everything is ‘context, context, context,’” he said. “We don’t want to generalize the uniqueness of publishers. Consumption is obvious. Advertisers can only get [this valuable first-party data and analytics] through working with us; they can’t get it on other platforms, it’s based on the content or the media business.” 

Keeping it accessible is key. The tool can be used for display, video, native ads either via direct deals or programmatically and will be sold at a premium rate given the additional ad-targeting features, although Dicker didn’t disclose an exact price tag.

“Partners who buy premium products usually want premium targeting,” he said.