MarTech Interview with Jana Jakovljevic, Senior Vice President for Partnerships at Cognitiv

Jana, it’s a pleasure to have you with us at MarTech Cube. To start off, could you give us a brief overview of your professional journey and what led you to your current role?

It’s a pleasure to be here! Yes of course, my first job out of college was working for a record label and while there I quickly took an interest in their digital department which led me to the path of digital advertising. After gaining valuable experience working on the publisher side, I ventured into the ad tech and programmatic space in 2009, during its formative years. Six years ago, I joined Cognitiv, a pioneering Deep Learning Advertising Platform, where I now hold the role of SVP of Partnerships. Our goal is to help brands identify and connect with consumers through advanced AI technologies.

How does the collaboration between Cognitiv and PubMatic enhance advertising efficiency for your clients?

Our partnership with PubMatic makes it possible for buyers to create and target content that is relevant to their brand, powered by the latest advancements in Large Language Models (LLMs). Our integration utilizes a dynamic deal ID enabling buyers to easily access content that aligns with their campaign goals.

Cognitiv provides the cutting-edge AI and intelligence, while PubMatic ensures seamless integration and scale. Since most buyers already have access to PubMatic supply through their DSPs, they can effortlessly target their custom ContextGPT deal ID through their DSP of choice, making the process simple and highly effective.

Can you walk us through how Cognitiv’s advanced ContextGPT product helps advertisers achieve positive sentiment and meet DE&I media buying goals at scale?

We designed our product where by default all deals are set to neutral sentiment and above, but advertisers have the option to adjust a slider to target only positive sentiment content. This feature allows brands to appear on premium news publications while avoiding content that may discuss violence, or mention terms that could be misinterpreted as unsafe for brands out of context, providing a more practical solution than blocking entire news domains. You can also filter for URLs that resonate with specific demographic groups, such as Hispanic or Asian communities. This ensures that campaigns can effectively reach and engage diverse audiences, aligning with DE&I objectives and enhancing brand presence in a positive and inclusive manner.

What are the standout benefits of the Inclusivity product feature built using GPT-4, and how does it help advertisers identify relevant and brand-safe media?

In our commitment to fostering inclusivity, we meticulously defined inclusive and non-inclusive language, guided by the standards of the Linguistic Society of America. Our advanced models are trained to capture meaning, tone, and sentiment with human-like precision, adept at identifying even the most subtle nuances of non-inclusive language.

For instance, while the sentence “only women can be astronauts” would be flagged as non-inclusive, our models are sophisticated enough to understand the context—recognizing it as an example rather than a statement promoting exclusion. From a brand safety perspective, our system allows you to set parameters that ensure only appropriate inventory is added, eliminating the need for manual targeting of negative keywords. This streamlined, “set it and forget it” approach ensures a seamless and secure advertising experience.

How does the integration with PubMatic provide more nuanced and precise media buying opportunities compared to the usual contextual segments?

Our integration with PubMatic offers a distinct advantage because our tool can “read” and comprehend language, moving beyond the limitations of a keyword-based approach. Traditional tools often rely on counting keyword appearances, such as tracking how frequently “basketball” is mentioned on a page. However, this method lacks the nuanced understanding that a human mind would possess. For example, does the page discuss the best sneakers for playing basketball, or is it a news article reviewing a basketball game? ContextGPT captures these subtleties, empowering buyers to create custom segments tailored to the specific product they wish to market and the precise content they want their ads to appear alongside. This leads to more precise and effective media buying.

In addition to this our solution operates in near real-time. Newly published content becomes available for targeting within 15 minutes, whereas most other solutions have a 24-hour lag. This rapid turnaround allows buyers to capitalize on the fast-paced nature of content consumption, ensuring their ads remain relevant and timely.

I’m curious, what personal strategies do you use to stay ahead in the rapidly evolving field of programmatic advertising?

Having worked in digital media for such a long time, I’ve been fortunate to build a vast network of professionals from whom I can learn, share ideas, and seek advice from. I maintain a strong sense of curiosity, engaging with as many companies as possible and taking an interest in their business, even when I know there are no immediate opportunities for collaboration. You never know how these connections might evolve in the future!

Jana, could you please elaborate on the importance of robust cookieless solutions in today’s advertising landscape?

Despite the recent news that Chrome cookies are not going away just yet, it remains crucial for advertisers to leverage their first-party data and start experimenting with cookieless solutions. There are two main reasons for this. First, we don’t know exactly how Google’s opt-in mechanism will unfold. Second, a significant portion of users are already browsing without cookies. As we’ve just seen, we don’t know what the future holds and things can change unexpectedly. Ensuring your first-party data is robust and well-organized will give you a step up from your competitors. For example, if you’re an e-commerce business, ensure your media partners are working with alternate ID solutions and that you’re collecting the right first-party data, such as order IDs, this will help create a closed feedback loop for where cookies are not available. It’s probably a good time for brands to reduce their reliance on them if they haven’t already.

How can advertisers leverage Cognitiv’s tools to effectively target diverse audiences, special interest groups, and multicultural communities?

Cognitiv’s tool was designed to create custom contextual segments, allowing advertisers to target a virtually limitless range of special interest groups. For example, you could target content specifically tailored to people interested in the most niche topics, such as turtles riding skateboards (which, by the way, is highly entertaining and worth a watch!). This flexibility ensures that advertisers can reach diverse audiences, special interest groups, and multicultural communities with precision and relevance.

What advice would you give to advertisers looking to boost their campaign relevance and performance while respecting online privacy?

First, embrace contextual targeting and utilize solutions that leverage the latest advancements. This ensures your ads are relevant to the content being consumed while respecting user privacy. Second, make the most of your responsibly collected first-party data. This data is not only more reliable but also helps create personalized experiences without compromising privacy. Lastly, focus on high-quality, engaging creative. When combined with the right context, compelling creative will resonate positively with your target audience.

Finally, Jana, what insights or “golden rules” would you like to share with our readers about the future of advertising and the importance of contextual targeting?

We’ve seen from our own data that context can be a better predictor of conversion than user behaviour. Understanding the context in which your ad appears in real-time is vital for enhancing user engagement without relying on personal data. Invest in high-quality, creatives, from our experience we’ve found rich media tends to have a higher engagement rate. Lastly, don’t ignore the quality of placement. This year, optimizing towards higher attention units and reputable publishers will be a major focus.

For more expert Interviews, articles and industry updates, follow Martech News