One of the things that was always alluring about third-party cookies was the illusion that they provided a real one-on-one connection with consumers. And that, as we discussed in an earlier
installment of this series, pushed the original connector-context-to the side. But now that the industry is faced with crumbled cookies, the blinders have come off and many marketers are saying that
this shift is not only not a bad thing, but it's an opportunity to rethink their original commitment to context-this time energized by data and powered by technology.
"In the past there
were a few primary use cases for third-party data," says Umberto Torrielli, chief strategy officer of Silverbullet, a marketing transformation company, and co-founder of 4D, the company's Contextual
Outcomes Engine. "The main one was to use data to reach the optimal audience. To do that, you'd purchase cookie-based data for people who have supposedly displayed either an interest in or an
intention to purchase the product you're selling." However, he notes, "at the time these were antiquated fundamental basics and the foundation of the transaction was flawed." For examples, third-party
cookies were tied to a particular device, but that device was often used by a number of different people in a household, not all of whom shared the same interests. "Third-party data is based on
someone else's understanding of a behavior or a transaction online," he says. "At best, it's inferred, it's probabilistic."
Third-party data can also be inaccurate. With context not
considered, ads placed-and often retargeted-using programmatic reasoning based on third-party cookies often appeared-and then reappeared-far from content that was relevant to either the brand or the
consumer. This wasted impressions, annoyed consumers, and played into the hands of privacy regulators.
Now, with the days of third-party cookies numbered, Umberto says, new attention is
likely to be paid to "technology that is built on factual, contextual, and deterministic first-party data," which he defines as "media-derived campaign data that a brand, agency, or publisher has the
right to gather, and therefore the right to use."
This data, by definition, is based on context-which is a factor of where rather than who the audience is. The
Campaign as an Organism
"It doesn't matter who in a household was looking at the ad, it's that someone engaged with it in that context."
A key benefit to basing an analysis of campaign performance on contextual intelligence, Torrielli believes, is that a marketer can
determine "how that campaign is performing in terms of specific contexts, topics, sentiment and keywords. You can analyze the campaign mid-flight and use real-time data to understand deterministically
which contexts are overperforming the average of the campaign." As a result, it doesn't matter who in a household was looking at the ad, it's that someone engaged with it in that context. Moreover,
Torrielli explains, "when you start to aggregate that at scale, you start to draw some very strong correlations. And this allows you, based on the data, to fine-tune the composition of your context
mid-campaign. It's a much stronger version, a much stronger narrative, approached from a different angle," meaning you can determine what it is about the combination of advertisement and content that
is resonating with the consumer.
Writing in MediaPost this past March, marketer Cory Treffiletti noted that "No campaign is ever stagnant. They are living, breathing, organic organisms."
What keeps the campaign organism alive, he suggested, is a focus on context. "Contextual targeting based on relevant content is not only important," Treffiletti wrote, "it becomes the backbone of any
And that holds equally true, Torrielli notes, when the context doesn't work for the campaign. Depending on the data, "you may decide to block what's not working," he
says. "If particular topics are underperforming the average against the outcome you're measuring, then you change the context."
"It all drives toward desired outcomes with contextual deterministic data-more powerful and precise than
Ultimately, Torielli says, "It all comes together,
driving toward the desired outcomes using contextual deterministic data, which is tremendously more powerful and precise than the guesswork involved with and the often-dubious nature of third-party
data." The Layered Approach
However, an advertiser's ability to target is only as powerful as the data that advertiser has access to and the ways in which it
can be analyzed. When focusing on context, as we discussed in a previous installment of this series, it's critical to be able to layer on information that will permit careful suitability targeting.
Going beyond the narrow confines of brand safety, that means understanding the nuances that make or break an environment for a specific campaign. Finding the data that exposes those nuances means
working with a collection of partners that are able to, as Torrielli notes, "find and discover new layers, go beyond keyword targeting, and discover what is possible with the power of dimensional
targeting." That might mean layering in, for example, suitability protection against racism. Or it could mean avoiding domains with excessive popups and redirects. Or perhaps it means ensuring that ad
dollars are being funneled across IP-infringing websites. Different campaigns require different filters, and pulling in a collection of different partners offering different dimensions, helps a
company like 4D cover the context waterfront.
But the focus on suitability targeting, Torrielli contends, can't be constrained by medium. Avoiding Missed Opportunities
"As a brand, as an advertiser, you shouldn't have to worry about what medium the consumer is using to consume content when they're going to see your brand," he says. "You just
want to know that, regardless of whether it's a video ad, a display ad, or even an audio ad, you can set your rules and ensure that the brand will be seen in the right time, the right place, the right
Moreover, Torrielli points out, while earlier technology involved in analyzing digital ads became adept at understanding the context of a page based predominately on
keyword analysis, "context is more than keywords." And with video, he adds, it's important to move beyond the limits of just analyzing speech to text. For example, if a video has no audio track, but
instead has music overlaid on top, or if the narration in the video doesn't match what's happening in the background, Torrielli suggests that "there are likely to be a number of missed opportunities."
The trick, he says, is being able to "analyze videos frame by frame in order to capture the plethora of different information, from logos to celebrities, from emotions to things you might want to
avoid, such as nudity."
Still, despite the opportunities afforded by first-party-data-informed context, some marketers are taking their time shifting their focus. After all, they say,
third-party cookies aren't going away tomorrow. Maybe it's better to reap the old familiar rewards they provide for as long as possible.
In addition, there are many efforts underway, by an
array of players, to find some sort of identifier-based cookie replacement. Maybe it's better to wait and see how that shakes out.
Torrielli, however, says he's a great believer in
"first-mover advantage." As he puts it, "You don't have to wait to see who's going to win the Unified ID wars. The longer you wait, the more difficult It will be, and the further behind you're likely
to fall." This is an opportunity," he adds, "for brands to partner with future-thinking technology providers that have been working on creating solutions that solve for challenges in the post-cookie
world. These are the leaders that are bridging the gap between what has been done in the past and what will power the future of marketing. There's no time like the present to start moving forward."