Adzine Interview: Data Sources and Targeting Strategies for Cross-Channel Video

In this interview for Adzine, Anton Priebe speaks with Olaf Peters-Kim, Co-Founder of Welect GmbH.

Whether data can truly be called "the new oil" remains debatable, as the comparison falters in some areas. Stricter data protection regulations and various technical developments have led to a situation where information about online users is no longer as widely available as before. In this interview, he evaluates how effectively available data can be used in today’s advertising ecosystem and discusses the differences between good and bad data foundations.

ADZINE: Olaf, data has always been considered the fuel for online advertising campaigns. The better the data foundation, the more precise the targeting, and the more personalized the communication to the desired audience groups. Does this principle of data as "fuel for campaigns" still apply today?

Olaf Peters-Kim: Yes, data remains a central driver for the success of online advertising campaigns, not only in targeting but also in the preliminary strategic planning phase. What will change is the type of data being used. In the past, third-party data and cookies dominated. In the future, first-party data and user signals will take precedence. Choice-Driven Advertising (CDA), as practiced at Welect, relies on voluntary user signals. Advertisers can then work with higher-quality, real-time data that is not only GDPR-compliant but also precisely relevant for targeting—specifically, which ad holds the highest relevance for a person in a given situation. However, amidst all the focus on data, we must not neglect creativity. In particular, user empowerment enhances the role of creative agencies, even in targeting.

ADZINE: What makes a data foundation good or bad for advertising?

Olaf Peters-Kim: A good data foundation is characterized by its timeliness, accuracy, and relevance. Bad data is outdated, incomplete, or collected through questionable methods. Accurately interpreting data signals—often unrelated to advertising, such as my creditworthiness, browsing behavior, or smartphone location data—poses a complex and resource-intensive challenge for our industry. This is especially true in targeting, where describing the ever-changing behaviors of millions of individual people is incredibly demanding. Moreover, experts in this complex discipline face an additional issue: Collecting, storing, processing, analyzing, and applying the vast amounts of data, which need to be updated daily, generates nearly as much emissions as the entire AdTech ecosystem itself.

ADZINE: From overseas, we hear that the available targeting data in the programmatic ecosystem is "garbage." Arielle Garcia stated this during this year's Adzine Connect. How do you assess the quality of the targeting data available on the market?

Olaf Peters-Kim: Arielle and the Check My Ads Institute are successfully combating disinformation and ad fraud in the U.S. In the German-speaking world, Thomas Koch and Michael M. Maurantonio take on this role. Arielle’s statement pragmatically points to the low quality of targeting after testing various data suppliers and AdTech providers using her ID or cookies. She examined which audiences she was assigned to and what demographic characteristics were attributed to her. These deficiencies don’t necessarily result from poor data but, as mentioned earlier, even the best data cannot adequately represent the complex, individual behaviors of people.

ADZINE: One alternative to buying data is building first-party data. According to some experts, we are in the midst of a transition to a first-party advertising world. How can advertisers improve the quality of their own data?

Olaf Peters-Kim: Regarding advertising, I’m skeptical whether this is the right solution. While I’m not a data supply expert, companies can enhance the quality of their first-party data primarily through direct interactions with their customers. However, how do they obtain the first-party data of non-customers, which they rely on to generate growth or compensate for customer losses? In the lower funnel and MarTech sectors, first-party data is highly valuable, but in the upper and middle funnel, it becomes significantly harder to use first-party data effectively for advertising communication.

ADZINE: Some advertisers have no choice but to rely on external data since they naturally struggle to build their own. What can they do to improve their targeting?

Olaf Peters-Kim: If they continue to use external data to reach people, they should only collaborate with established and trustworthy providers and avoid unreliable market players for cost reasons. However, this doesn’t resolve the targeting quality issues Arielle highlighted. Moreover, they’ll only reach users who give their consent, don’t use ad blockers, and don’t browse in privacy mode—this is currently a maximum of 40% of online users. More reach can be achieved by using other external data, such as contextual or semantic data. However, it’s far more scalable and efficient to leverage the interactivity of the internet and obtain voluntary data signals from millions of people in real-time.

ADZINE: Is data quality in contextual targeting as problematic as in user-based targeting?

Olaf Peters-Kim:Contextual targeting offers a valuable alternative in times of stricter data protection regulations since it doesn’t rely on collecting or processing personal data. However, contextual targeting also presents challenges. For instance, Handelsblatt knows which articles I read, Netflix knows which films and series I watch, and Spotify knows my music preferences—all high-quality data. But how accurately can it determine which ad will be most relevant to me in my next online session? User-based targeting, on the other hand, directly asks in each session which ad is most relevant to the user at that moment. This approach not only captures the right timing and context more precisely but also does so for millions of people while being significantly more resource-efficient.

ADZINE: And now the big question: Is programmatic advertising and its targeting still salvageable?

Olaf Peters-Kim: Programmatic advertising and targeting don’t need to be saved—they will continue to exist. Instead, we should ask how we can make the programmatic ecosystem and the data value chain significantly more sustainable. By doing so, we can make a vital contribution to saving ourselves and our future.

ADZINE: Thank you for the conversation, Olaf!