How Our New Personas & Predictive Audiences Are Reshaping Ad Targeting

Project Agora recently launched a powerful set of new capabilities to elevate advertising targeting precision. At the core of this advanced approach are Data-Enriched Personas and Predictive Audiences, offering advertisers a new level of targeting sophistication. 

Built on Taboola’s technology, leveraging behavioral and first-party data, and enhanced through AI-driven predictive models, this new approach is designed to scale performance effectively and efficiently. 

George Ntouzos, our Chief Product Officer at Project Agora, explains how this targeting strategy is transforming the open web advertising landscape, delivering measurable and impactful results for advertisers. 

Q: What exactly are these new personas and predictive audiences, and how do they help advertisers reach their ideal audiences more accurately? 

George Ntouzos: Think of personas as the avatars (laughs) of our advertisers’ ideal customers. They’re built using Taboola’s massive data infrastructure. In fact, Taboola has more 1st party data signals and contextual insights than anyone out there, but Google. 

We’ve enriched this data further by incorporating additional behavioral, geographic, and demographic data, as well as interest and purchase intent signals from supplementary sources. This gives us an unprecedented level of targeting precision. All data collection and processing fully comply with privacy standards and regulations. 

These personas empower advertisers to engage the right audience, at the right time, from the very first impression, tailored to any campaign goal. 

Predictive Audiences, on the other hand, add a layer of intelligence. These are AI-powered models that forecast which users are most likely to take a desired action—like a purchase or registration—based on real-time behavior and historical conversion patterns. 

The combination of enriched personas and predictive audiences translates into better budget efficiency, more accurate targeting, higher engagement, and increased conversion rates. In fact, campaigns already running with this approach are showing a 20%+ uplift across all major metrics. 

Q: What’s the ideal way for advertisers to leverage this dual approach? What would you recommend to a client? 

G.N.: The most effective way to utilize both custom personas and predictive audiences is by integrating them across the entire marketing funnel—starting from the first impression. 

At the upper funnel, personas allow brands to engage broader but highly relevant audiences based on real-world interests, intent, and behavior. This ensures each brand message reaches users with authentic relevance, boosting early engagement. 

As audiences move further down the funnel into consideration and decision stages, predictive audiences come into play, targeting users most likely to convert. Trained on conversion signals, these AI models drive higher advertising efficiency and a stronger Return on Ad Spend (ROAS). 

Digital advertising is steadily moving toward full personalization, and with this solution, we enable brands to reach the right user, with the right message, at the right moment—across every step of the customer journey. 

Q: How do these new targeting capabilities differ from what done in the past? 

G.N.: The biggest shift is that we’ve moved away from static audiences to a dynamic, intelligent, and evolving targeting layer. With our enriched personas and predictive audiences, we’ve entered an era of advanced personalization and real-time intelligence. 

Our personas are built on real behavioral, contextual, and first-party data—not on rigid profiles. We analyze granular patterns. Take a user who, in one session, shows interest in vegan beauty products and simultaneously browses travel content. Understanding such multi-layered behavior was previously impossible with traditional targeting tools. 

Meanwhile, predictive audiences go even deeper—analyzing micro-behaviors and past conversion events to pinpoint users with the highest likelihood of converting. 

And most importantly, everything we offer is tailored to our clients’ briefs—enabling advertisers to target exactly the audiences they need and reach potential buyers with unmatched precision and flexibility. 

Q: As third-party cookies are being phased out, do these targeting options offer a solution? 

G.N.: As we move beyond third-party cookies, advertisers need to have their post-cookie strategy in place—if they haven’t already. 

At Project Agora, we’re already in the future. Powered by Taboola’s technology, our ecosystem is cookie-agnostic by design. Our enhanced personas are based on context and first-party signals, and our predictive audiences are trained on conversion events that don’t rely on third-party cookies. 

This means we can offer effective, privacy-safe targeting even in cookieless environments. 

Q: What would you say to advertisers who prefer targeting audiences on Facebook or Instagram? 

G.N.: It’s true that many advertisers lean toward platforms like Facebook or Instagram, often believing they offer richer first-party data and more reliable outcomes. How can the open web compete with these closed ecosystems—especially without third-party cookies? The answer is: not only can it compete, but in many ways, it outperforms them. 

Through Taboola and the Realize platform, we have access to data from over 600 million daily active users. This includes what users actually read, how they engage with content, what they care about, and what truly captures their attention. We don’t rely on “likes” or self-declared interests—we use authentic behavior. 

What’s more, we can reach audiences outside the social media bubble—people who engage with quality content across premium publisher websites. This extends the reach and diversity of your audience, often capturing segments that are otherwise missed in walled gardens. 

And here’s a key point: on the open web, people express their real interests—unfiltered and uncurated—unlike social platforms where users often manage their public image. That makes our targeting more genuine, relevant, and effective. 

Our custom personas and predictive audiences bring the power of audience targeting outside the walled gardens, enriched with signals like scroll depth, active time on page, and intent-based behavior—insights unavailable on social media. 

This not only helps brands connect with the right consumers but also supports our publisher partners by driving more relevant demand to their content. 

According to Hootsuite, users spend 25% of their online time on the open web. With our targeting solutions, we enable large-scale, privacy-compliant reach across a wide range of publishers—complementing and strengthening the overall media mix beyond search and social. 

 —

Data-Enriched Personas and Predictive Audiences are now available across all our markets—offering a privacy-first, performance-driven audience journey for advertisers ready to lead in the post-cookie era.

Reach out here or to your account manager to find out how your campaigns can benefit from the new personas and predictive audiences.