Dstillery, an organization leveraging machine learning (ML) for ID-free audience targeting, is entering the CTV market.
AI continues to reshape the programmatic landscape, from ad buying to contextual programming. And as privacy changes and a cookie-crumbled future are thrown into the mix, more businesses are looking to adapt to an ID-free targeting strategy.
Founded in 2008, Dstillery has always focused on using AI modeling data science to understand audiences and targeting more holistically — according to Mark Jung, its VP of Product — and that focal point has set them up well for a post-cookie delivery world. Its ID-free Custom AI platform, launched last year, looked to reach ad impressions rather than users across all browsers. Now they’re bringing the approach to CTV.
And Dstillery isn’t the only one. Comscore recently made its move into the space as well when its business unit tailored to programmatic targeting, Proximic, dropped a trio of AI-powered tools allowing advertisers to target ID-free audiences contextually across platforms without using cookies.
Not long after, advanced TV ad tech company Cadent announced its plans to integrate the tools into its Cadent Aperture Platform in efforts to “unify the fragmented data ecosystem of converged television,” as Cadent EVP of Business Development and GM of Platform Sales Tony Yi described.
Jung told StreamTV Insider that as Dstillery looked to enter the CTV market, they assessed two predominant problems in talking with potential clients: trust and time.“Audience buying today, and CTV a lot of the time, you're doing technology jumps,” he explained. As an example, if a company looks to buy “DIY-ers” that are using CTV as an audience from a DSP, it needs to move “outside of the walled gardens and the programmatic guaranteed PMPs.” In that technology jump, there isn’t enough clarity around what defines the DIY audience.
In a third-party cookie environment, “an audience is a bucket of devices that I'm trying to target, and I have no real idea how that audience got made,” he continued. “So you have this compounding problem of uncertainty on the source audience leading to uncertainty on the expansion and the reaching of that CTV audience, so it's really [a] black box. That's where that trust issue within CTV really started getting called out by the folks that we talked to.”
The complexity of advanced buys is also saturated, cluttered and hard to differentiate. “It takes a massive amount of people power and people time to try to wade through all that and find the right stuff to target,” Jung added.
The ID-less and AI-powered approach attacks these pain points and looks at the “total consumption patterns” of a consumer, painting a larger and more nuanced picture of diverging values and behaviors within a target audience.
Dstillery’s new CTV tool uses its ID-free technology to comprehensively understand inventory relevance in combination with the genres and networks that resonate most with an advertiser’s targeted audience. In efforts to make the process more data-driven, the tool’s AI automation capabilities also eliminate the need for traditional CTV device graph layering and manual setup of PMP deal IDs.
Migrating away from the “bucket-of-devices” mentality revolving around the cookie, this evolution is bringing back a behavior- and interest-focused angle “mirroring how internet ads started,” Jung concluded. “We’re in the wild west of it.”