Mastering CTV With A YouTube Playbook: How Pixability is Bringing Its Contextual Targeting To CTV
Mastering CTV With A YouTube Playbook: How Pixability is Bringing Its Contextual Targeting To CTV
“There’s a big misconception that everything on a TV screen is safe and suitable,” says Jackie Paulino, Chief Product Officer at Pixability. “Some people who haven't spent much time with CTV shows think everything is fine just because it's on a TV screen. But I've seen content on TV that many brands wouldn't want to align with. Some think user-generated content issues are only on platforms like YouTube and Facebook. Our data shows that's a misconception.”
ALAN WOLK (AW): How would you explain what Pixability does and where do you sit in the greater TV ecosystem?
JACKIE PAULINO (JP): Pixability has been around for over 10 years, first in the YouTube space, focused on contextual targeting, brand safety, suitability, and performance of YouTube ads. Over the last few years, we've expanded from YouTube to CTV, doing the same things focused on brand safety, suitability, and contextual targeting in the CTV space.
As Chief Product Officer, my role is to oversee our strategy and roadmap. We have two different product lines, one focused on managed service where we do full-suite planning, buying, reporting, and insights for our customers. Then we have a self-serve product around our data licensing. So all the work we do around scoring our contextual targeting, we can license that data for use on YouTube and on CTV. That's working more directly with media buyers, they license our data, log into our platform, download data, and then execute it in DV 360, or in Google ads.
AW: You’ve been doing contextual targeting on YouTube for a while now. Any thoughts on why contextual has become a thing on CTV now?
JP: We learned a lot from the YouTube world. And that was important because the YouTube world is more complex—you’ve got 500 hours of content uploaded every minute. As we shifted into CTV, we saw similar patterns. It's important where you run your ads, both in terms of the content and in terms of when they are served. We've proven that performance is dictated by content and context. We brought that idea to CTV, talking to customers about their pain points on CTV. It overlapped nicely with our YouTube products.
AW: What were those pain points?
JP: The big pain points were all around lack of transparency, not understanding where their ads were running, not understanding the types of programs they're in, and lack of control over where their ads ran. Those are the problems we're focused on solving in the CTV space. We've learned a lot about how to do that in the YouTube world, so bringing that to the CTV world has been exciting.
AW: Can you explain how you solve those problems? How do you help someone who says there's no transparency and doesn't know where their ads are running?
JP: For understanding context on CTV, we've partnered with IRIS.TV. They’ve done the heavy lifting of going publisher to publisher to curate data for us. The data they pass us via the IRIS_ID is similar to the data we get from the YouTube measurement program. We're familiar with how to score and understand video context. So we look at the surrounding metadata, the title, description, and the transcript of the video, as well as a frame-by-frame, scene-by-scene analysis to understand a few things. What type of show is this? We use the IAB taxonomy. Is it about Home and Garden, or is it a news show? Is it safe or suitable for a particular advertiser? For safety and suitability, we've partnered with the Global Alliance of Responsible Media, which is part of the 4As. We've categorized shows as low, medium, and high based on the GARM rubric of safety and suitability.
AW: What makes your approach unique?
JP: On the CTV side, there are a few things. Pixability has more than 10+ years of experience powering brand-safe/suitable YouTube campaigns. We're the only IRIS-enabled partner that is looking at YouTube data and then carrying that data into the CTV space. As members of the YouTube Measurement Program (YTMP), we have access to more data than other third-party providers and are the only company certified by Google for Brand Suitability & Contextual Targeting + Content Insights. Our insight into what works best on video on TV screens and our ability to apply that to the CTV world is a differentiator. We're also using generative AI to contextualize some of that data. We're using open AI, combined with our custom machine-learning models. The accuracy and the transparency you'll get are different than what you will get from others. That's a core differentiator.
AW: In terms of where you sit, do you help brands and agencies with the buying process?
JP: It depends which of our two products they are engaging with. With our managed service product, we handle the buying, planning, reporting, insights—full suite, white glove service. On the self-serve side, in the CTV world, we score all the inventory. We work with the partner to understand what inventory they want to run. We create a private marketplace deal, and they activate it. The buying is on them.
AW: So it's a programmatic buying method, not selecting specific spots?
JP: Exactly.
AW: Are you able to share any case studies of someone who benefited from your CTV expertise?
JP: We have some great examples of advertisers who came to us with specific content concerns. For instance, an alcohol brand that wanted to avoid driving content and to stay away from kids' content. An alcohol brand wouldn't want its ads running during a nursery rhyme video or content geared toward children. Those advertisers see success with our contextual product and recognize its value.
Pixability is a founding member of the Alliance for Video-level Contextual Advertising (AVCA), an organization dedicated to funding research into applications of AI for contextual advertising in streaming. Visit www.avca-tv.org to download the AVCA’s recently published consumer research ‘Driving Viewer Attention and Brand Metrics in CTV Advertising: Understanding the Impact of AI in Contextual Targeting.’
