LONDON – Contextual advertising may be one of digital media’s oldest ideas, but advances in artificial intelligence are pushing it into new territory, said Brian Gleason, chief executive of contextual advertising platform Seedtag.
Speaking with Beet.TV contributor Robert Andrews at the Beet.TV and WPP Media Leadership Summit, Gleason traced the evolution of contextual advertising from its roots in niche publishing to what he described as “neural contextual,” a new approach that aims to understand not just content, but the emotions and mindsets connected to it.
Gleason pointed to his early career at magazine publisher Ziff Davis as an example of how contextual advertising began.
“Bill Ziff founded his businesses on very specific or bespoke magazines on different topics,” he said. “That was contextual at its core.”
Over time, the industry moved toward semantic models that categorized pages by topics or URLs, such as tagging a website as “running” or “travel” using taxonomies like those developed by the IAB.
Targeting personas and mindsets
The next major shift, Gleason said, came with the realization that brands don’t target topics. Instead they target people, and more specifically, the personas and mindsets of those people. At Seedtag, that led to the development of technology capable of deeply understanding pages and videos using embeddings connected to media supply. But even that approach had limits.
“What [our technology] didn’t understand was the emotion connected to that,” Gleason said. Over the past two years, Seedtag has worked to build what he called an “emotional graph,” allowing its AI system, internally known as “Liz,” to assess emotional resonance alongside content. The result, Gleason said, is advertising that moves beyond organizing topics to understanding how people feel when they encounter content.
Insights into emotional states
This emotional layer is central to Seedtag’s concept of neural contextual advertising. Rather than focusing on what users did in the past, Gleason said the goal is to understand the mindset someone is in when they are consuming content (whether they feel inspired, nostalgic, curious or reflective) and align advertising accordingly.
To illustrate the point, Gleason referenced brand storytelling.
“We’ve been telling stories forever in advertising,” he said, arguing that heavy reliance on behavioral targeting caused the industry to drift away from emotional connection. A campaign like Nike’s “Just Do It,” he said, performs best in moments of inspiration, effort and achievement. Seedtag’s AI allows those emotional moments to be identified at scale and used as part of media planning.
Seedtag has begun testing this approach academically. Gleason said a recent study conducted with a Columbia University professor found that neural contextual advertising delivered a roughly 350% lift compared with traditional contextual methods. While he emphasized that the work is still early, he said the company is encouraged by the results.
Respect for consumer privacy
Privacy, Gleason added, is a key distinction between neural contextual advertising and traditional behavioral targeting. Seedtag does not use personally identifiable information, instead analyzing the environment in which ads appear.
“It’s advertising that’s welcoming rather than advertising that’s chasing,” he said, describing the approach as fundamentally different from surveillance-based models.
For publishers, Gleason argued neural contextual offers a chance to reassert the value of high-quality content. He said the industry’s focus on clicks has often failed to reward emotional resonance, contributing to the rise of clickbait. By identifying which articles genuinely connect with audiences, whether through happiness, nostalgia or curiosity, publishers can better surface and monetize meaningful content.
Tackling mobile challenges
The approach also addresses long-standing challenges on mobile devices, where behavioral targeting often performs poorly. Because contextual advertising is based on content rather than user identity, Gleason said it works consistently across mobile, desktop and connected TV, helping publishers earn value regardless of platform.
Looking ahead, Gleason said neural contextual advertising could become a foundational layer for media planning as early as the next year or two. While personalization has long been a goal for advertisers, he said it has been difficult to achieve at scale without high costs or privacy trade-offs. Neural contextual, he argued, can act as an “equalizer,” offering insight into how audiences think and feel without relying on personal data.
“The advancements that we’ve seen in AI and large language models over the last two years has been astonishing,” Gleason said. While he acknowledged that future capabilities may feel “frightening in some ways,” he added that they are also “magical,” and that companies able to adapt to this new reality will be best positioned to succeed.
Seedtag Wants Its AI to Understand Advertising Context ‘Like a Brain’
You’re watching coverage from Beet.TV’s Global Leadership Summit with WPP Media, filmed in London, presented by Criteo, Index Exchange, Seedtag & The Trade Desk. For more videos from this series, please visit this page.
