sonobi logo in blue with transparent background

Contextualization and Its Promise for Digital Advertising

Photo by Andrew Neel from Pexels

Digital advertisers have long depended on third-party cookies for consumer targeting and ad personalization. However, even though cookies show no sign of ever being phased out, the digital advertising industry has been working diligently to discover alternative methodologies that will future-proof their advertising and allow it to remain relevant regardless of how consumers are identified. Sonobi, a pioneer in digital advertising technologies, is expanding its signature holistic identification offering by adding contextualization as a complementary targeting method. Contextualization, which does not involve third-party cookies, is the automated analysis of websites in order to accurately identify the subjects and categories of their content. Such information greatly enhances user targeting that could improve cookie functionality, or work without cookies entirely.

Digital advertising technology supports brands by enabling them to reach the right consumers, in the right context, at the right time.  Marketers and publishers have relied heavily on third-party cookies as a means of identifying consumers on the open web, even building behavioral profiles for them. That historical ability to identify the unique user is becoming increasingly difficult, both by the limited availability of cookies, as well as the introduction of new laws, regulations, and browser policies centered around privacy concerns.

Removing PII

With the advancing removal of personal user identification from internet browsers and device platforms, brands are expanding the techniques they employ to ensure that they reach their intended audiences. A possible solution is contextualization since it allows brands to relate to consumers in a manner that better respects privacy. An often overlooked, yet relevant, piece of targeting information is the context of the media surrounding an advertisement. Instead of targeting users based on tracking or some form of identifier, brands can reach consumers based on their interests, which can be determined by the topics of the web page they are visiting. For example, a user visiting a web page about travel would likely be a good candidate for hotel or car rental advertisements. Furthermore, contextualization respects a user’s privacy while simultaneously supporting brand safety. When a brand is aware of the content that surrounds adjacent advertising inventory, they can prevent their advertisement from appearing next to subjects that may be negative or damaging to the brand’s image. For example, an airline company would be able to keep their advertising from being placed next to articles about airline disasters. An additional benefit of contextualization is that it doesn’t preclude the use of alternative or existing identification methods. In fact, it can be used in concert with user cookie targeting as well as other user identification solutions.

Resilience to increasing privacy regulations is a major advantage of a strategy that does not rely on user identification. Measures such as Europe’s GDPR or California’s CCPA already inhibit the collection and tracking of personal identifiable information; yet the collection of user data is essential in user identification. Some other alternatives to third-party cookies, such as Google’s FLoC, have been met with considerable opposition from publishers and users alike (Citation: Google’s FLoC Is a Terrible Idea). In contrast, the contextual classification of websites is unaffected by privacy regulations – a website’s content belongs to its publisher as opposed to the legally protected viewer’s data.

Establishing Standards

A website’s contextual information can relate to the site as a whole, or to a specific page within that site.  It can take the form of a keyword summary of content, or pre-specified categories (e.g. sports, nutrition, entertainment, fashion).  The Interactive Advertising Bureau (IAB), responsible for establishing ad-tech industry standards, has defined a multi-tiered set of content categories, such as “Food & Drink”: “Desserts and Baking” that can be used in describing web context. Sonobi has developed its own algorithm for extracting a list of categories and/or keywords that best capture the essence of a website or web page. This information can be communicated to brands and buyers providing them a way to deliver relevant ads to the right audience in a brand-safe environment. For example, if a website publishes articles about sports, and the advertising partner is made aware of it through Sonobi’s technology, a brand can choose to show advertisements related to sports equipment or clothing to the viewer. Furthermore, Sonobi’s detailed information on the specific content of the sports articles would help a brand avoid placing its advertisement alongside brand-inappropriate content, such as an article about a violent incident in a soccer game. This level of detail may not be captured through IAB’s relevant categories of “Sports”: “Soccer”.

Where Sonobi Comes In

Sonobi sees contextualization as one aspect of its holistic targeting solution. It’s not only an effective alternative to third-party cookies but also a more reliable and less intrusive solution for delivering relevant ads while respecting consumer privacy. The digital advertising community is at a stage where it is still exploring viable options for a cookie-lean, privacy-protected future. Contextual targeting is among the strongest and most promising approaches.  Extracting contextual information at scale and in real-time has recently become possible due to the wide availability of distributed processing and powerful new algorithms. Even now, machine learning approaches are bolstering new techniques for audio and video contextualization. Methodologies like the ones developed by Sonobi, that can extract keywords, standard IAB as well as custom categories, provide content-targeting flexibility. Ultimately, contextualization offers a sound path to effective targeted advertising no matter how ID technologies end up evolving in the future.

Latest News