Home ▸ Opinion ▸ How AI Predictive Modeling Can Help Publishers Escape Third-Party Cookie Uncertainty

How AI Predictive Modeling Can Help Publishers Escape Third-Party Cookie Uncertainty

March 22, 2021

Fact checked by Jürgen Galler Edited by Jürgen Galler

Everyone is still talking about the death of third-party cookies, but the conversation is shifting. While publishers, advertisers, and tech providers initially felt in the dark about how targeting – and digital marketing in general – would function going forwards, the focus has now moved to the implementation of solutions that bring benefit to all sides. As Google tests various bird-themed proposals in its Sandbox project to address an estimated 52% drop in publisher revenue, industry players have set their sights on first-party data and contextual targeting as ways to help them navigate cookieless uncertainty. This shift bodes well for a future that allows the ongoing delivery of effective digital marketing while preserving data access, control, and privacy compliance on the open web. It will also ensure scalability and sustainable monetization. However, for this shift to be truly successful, it will also require some help from technology – specifically artificial intelligence (AI) and predictive modeling.

Cracking the first-party data nut

Publishers know that one sure road to post-cookie salvation is already laid at their feet: first-party data. With a direct connection to their audiences, publishers have a better chance of gaining user consent and collating the data needed to fuel tailored content and monetization strategies and, in return, protect their bottom line. However, leveraging first-party data requires a holistic approach. Information from user interactions with the web is frequently unstructured and difficult to manage, especially for under-resourced publishers. Some users may be logged-in while others are anonymous, meaning the data coverage and the understanding of user activity is often inconsistent and incomplete. For example, our data shows only 2-10% of users share details such as age and gender, leaving the remaining 90% unknown. To make the most of valuable audience data, publishers need a way to organize, expand, and harness it effectively. This is where AI can help. Firstly, AI-powered tools with high processing and orchestration capacity can consolidate vast pools of unsorted data into a single insight store that’s easier to understand and activate. Secondly, they can fill in vital missing pieces to give publishers the all-important unified picture of the user journey, which opens the door to precise segmentation and activation, even in the absence of hard fact data. For example, machine learning algorithms can automatically analyze the engagement of consenting users based on contextual signals to provide a real-time window into unique interests and preferences that keeps profiles up-to-date and accurate. This not only saves days, or weeks, of manual processing, but also improves the advertising experience – driving more high-value publisher-driven ad formats that match context and user experience. Moreover, advanced AI modeling technology can plug the gaps for untraceable users. For instance, by including data from different environments – web, app, CRM, and CTV – patterns among users with certain attributes can be uncovered, fueling profile enrichment of similar users to maintain targetability across the whole audience. These technologies focus on logical, rather than declared attributes, which directly address the privacy-concerns that caused the third-party cookie deprecation in the first place.

Taking contextual to a new level

Contextual targeting has also regained popularity as the industry continues its pursuit to find effective, yet privacy-conscious and compliant solutions to target consumers in the post-cookie era. Technology in this area has come a long way in the past 10 years, now allowing the development of more accurate and agile contextual targeting tools. On their own, publishers have the extensible knowledge and ability to personalize content and build audience segments that form a workable foundation for context-based advertising. Yet, with more sophisticated tools in tow, they can now offer much greater targeting precision. For instance, today’s new generation of AI-powered technology allows publishers to move beyond traditional contextual limits. By using real-time signals, and comprehensive evaluation of their digital properties, they can collate accurate and scalable audience insight that can be made available to brands as well as to the publisher’s own marketing department. In short, it delivers the incremental addressability required to facilitate personalization that is not just highly appealing for advertisers but also ensures a better experience for users – ultimately strengthening audience bonds and boosting the likelihood of long-term loyalty. Honing in on first-party data and advanced contextual targeting is certainly a step in the right direction for the digital mediascape. On the publisher side, simply swapping one cookie (third-party) for another (first-party) might not be enough to entirely escape cookie uncertainty. The key is to build an agile and scalable first-party strategy by testing alternative technologies. This will create further opportunities for publishers to increase the value delivered to both users and brands and, in turn, strengthen their position in the market.

0 0 votes

Article Rating

Subscribe

0 Comments

Inline Feedbacks

View all comments

Contributor?

Join our community now! Log in to comment or become a contributor to share your unique insights and expertise. Your voice matters—get involved today!

Check Our Contributors

How AI Predictive Modeling Can Help Publishers Escape Third-Party Cookie Uncertainty

Jürgen Galler

Jürgen Galler

Cracking the first-party data nut

Taking contextual to a new level

8 Best Email Newsletter Platforms for Publishers in 2025

What Is a Content Creator?

Google News SEO Guide 2024: Best Practices for News Publishers