Fabio José Arnau Gallardo, VP of Performance at Seedtag, looks at how AI-powered intention models are reshaping advertising, exploring their real-world impact across industries, and the future of advertising in a privacy-first world.

Contextual targeting has transformed remarkably over the years. Once reliant on keyword-based techniques and category pages, more sophisticated systems now use embeddings—a more comprehensive user profiling method that places numerical representations of semantically similar words, phrases, and images in proximity to each other. While these advancements have significantly enhanced the precision of context-driven ad placements, we're now entering a new phase in contextual technology: one led by user intention.

For years, advertisers have tried to understand how different types of content are created across the internet. Some content becomes outdated almost immediately, while other pieces remain valuable over time as "evergreen" content. But beyond understanding content creation patterns, the next step is deciphering what users are hoping to get from the content they consume: are they trying to learn (informative content), looking for another website (navigational content), or interested in purchasing something (transactional content)? Solving this puzzle involves examining how users feel when interacting with different content types and how they behave during their search journeys.

Traditional ways of targeting intention have leveraged behavioural targeting, search intent, shopping signals, or location-based intent, but these methods either rely on third-party data, are vulnerable to signal loss due to increasing privacy regulations, or often fail to capture nuanced shifts in user mindset during multi-touch journeys.

This is where advanced predictive AI models pose a more innovative approach. These systems can be trained to replicate how people search for content, based on their intent, during a product decision process; distinguishing meaningful content from noise across the open web and optimising every available impression without relying on third-party data. Let’s dive into how these systems can work, alongside the impressive results advertisers are beginning to see with them.

AI can add intention to the context mix

Conventionally, contextual targeting has focused primarily on what content is about. Intention-based approaches add another crucial dimension to the mix: understanding whether content is more informative, navigational, or transactional in nature, and how likely it is to influence a user's decision-making process.

AI models trained on intent-labelled datasets can leverage advanced natural language processing to classify content along a spectrum from purely informational to highly transactional, assigning scores that reflect how close a user might be to taking action. For instance, a product review, checkout guide, or pricing comparison would receive a higher intention score than a general interest article on the same topic.

However, intention alone isn't enough. The most effective systems combine intention scoring with campaign relevancy assessment, ensuring that even highly transactional content aligns precisely with a campaign’s specific objectives, messaging, and target audiences. This combination allows marketers to activate specialised campaigns at the intersection of high intent and high relevance—delivering ads not just to the right person, but at precisely the right moment in their journey, and within the most effective context. Such an approach reshapes the traditional customer journey by allowing marketers to focus on mindset over demographics, therefore also making media buying more strategic, efficient, and impactful. 

Real world impact across industries

The power of intention-based contextual advertising reveals itself most clearly when examining specific industry applications. For example, in the automotive sector, consumers in late evaluation stages consistently engage with highly technical comparison content, especially when deciding between vehicles in similar price ranges.

These users aren't casually browsing; they're close to making decisions and are comparing specific details like engine performance, fuel efficiency, handling characteristics, and safety features. They're consuming in-depth articles that traditional keyword systems might view as simply another automotive piece, despite the content being much more focused and conversion-oriented.

Similar patterns emerge in the telecommunications sector, where consumers approaching purchase decisions engage with detailed content comparing mobile plans, data limits, coverage, and contract terms, especially when choosing between similarly priced providers. In travel, also, consumers close to booking consistently interact with practical comparison content like itineraries, cost breakdowns, and hotel recommendations when deciding between destinations or trip types.

These intention signals, often missed by traditional targeting methods, enable marketers to capture high-value users more efficiently and significantly improve campaign performance. A case study from Seedtag, employing intention-based segmentation in a campaign for Nissan, confirms this—a 68% reduction in cost per quality visit (CPQV), 35% reduction in cost per lead (CPL), and three-fold increase in qualified visits. The shift from demographic proxies and broad categories to content-level intent signals tied to specific comparison behaviors was a key driver behind these impressive performance uplifts.

The future of advertising in a privacy-first world

With much of the ad industry now favouring privacy-first solutions, intention-based approaches take on even greater importance.  

Rather than replacing existing metrics or strategies, intention-based approaches add a powerful new dimension to the marketer's toolkit in addressing privacy. These technologies help advertisers identify and activate users with clear intent at precisely the right moment, and reach them with relevant, timely messaging to maximise their engagement, thereby delivering stronger mid- and lower-funnel performance. As dynamic technologies, they can run continuous optimisations, ensuring ad placements adjust in real-time to capture peak user intent.

By focusing on the content that users are engaging with, rather than tracking their personal data, intention-based strategies also enable brands to fully respect user privacy while building out their first-party database. This can then be used to guide users further down the funnel toward conversion in multi-touch journeys. This approach isn't merely a workaround for signal loss—it's an evolution in advertising strategy that places user motivations at the centre of the targeting equation.

As we look to the future of digital advertising, one thing is clear: understanding and activating on user intention will be a key differentiator for brands seeking to find their audiences and create meaningful connections with them in an increasingly privacy-conscious world. Those that align their contextual targeting techniques—whether it’s keywords, categories, or embeddings—with the user intent, user journey, and specific campaign objectives will create more impactful connections with their audiences and achieve outstanding results.