Fixing the flaw in data collaboration, a conversation with Permutive’s Joe Root

With third-party cookies disappearing, user authentication rates remaining low, and legacy tools failing to deliver on their promises, it’s fair to say publishers are at a crossroads. Hot on the heels of their latest white paper, Rob sat down with Joe Root, CEO and co-founder of Permutive, to discuss the future of publisher data collaboration in a world where identity alone can’t scale.

In this candid conversation, Joe shares why predictive audiences are the key to unlocking value across the entire web, and why publishers must rethink their approach to collaboration to stay competitive.

Rob: According to your research, less than 30% of the web is addressable with cookies, and fewer than 5% of users authenticate across publisher environments. What does that shrinking reality mean for publishers trying to prove the value of their audiences?

Joe: It’s a wake-up call for the entire industry. For too long, publishers have been forced to rely on third-party cookies and other unstable identifiers to prove the value of their audiences. But with the majority of the web now unaddressable through these methods, publishers are left with a fragmented view of their users, making it impossible to accurately represent their audience's value to advertisers. This isn't just a measurement problem; it's an existential threat to publisher revenue streams.

Rob: Identity can be precise, but it doesn’t scale. For publishers, how should identity fit into the mix, and where does it fall short if it’s the only strategy?

Joe: As you say, while authenticated identity provides a precise view, it doesn't offer the activation at scale advertisers need to reach a broad customer base. The (often single-digit) percentage of users who authenticate might be the most engaged, but they're not representative of the entire audience.

But identity is still a crucial piece of the puzzle and should be treated as the connective tissue behind collaboration. We’re working closely with our customers to shape their identity strategy and have identified five pillars of publisher identity that are critical for overall success: unifying, extending, enriching, activating and analyzing their identifiers.

This gives publishers a foundation to build a more comprehensive understanding of their entire audience, enables successful connection between collaboration participants, and at its core, uses privacy-preserving signals to create scalable, high-performing advertising solutions.

Rob: Publishers are stuck in cycles where the same audiences get targeted again and again, driving up costs and lowering yield. What’s the fix to stop inventory from being undervalued in that system?

Joe: The current system overvalues a small slice of "known" users while undervaluing the vast majority of "unknown" but still valuable inventory. The fix is to embrace the power of cohorts and predictive audiences, creating a 100% addressable universe. Publishers can use their rich, first-party data to create predictive models that identify high-performing cohorts without ever sharing personal data. This allows them to package and sell their entire audience, not just those who log in. It creates a more efficient and equitable marketplace, where publishers are fairly compensated for the value of their entire audience, and advertisers can reach their desired customers at scale.

Rob: Identity isn’t dead, but it needs context. What’s the right way for publishers to blend contextual, cohort, and behavioral data with IDs so they actually get scale and performance?

Joe: Again, think of IDs as the connective tissue that can unlock a more sophisticated model of predictive audience building that harnesses all signals. This approach delivers both the precision of identity and the scale of contextual and cohort-based targeting, leading to better performance for advertisers and higher yield for publishers.

Rob: A lot of clean rooms promise privacy, but in reality they’re clunky and slow; more “privacy label slapped on” than real solution. What’s broken in those legacy models, and what should publishers be pushing vendors to deliver instead?

Joe: Legacy clean rooms are built on an outdated premise: that data needs to be moved and commingled to be useful. This creates massive inefficiencies, privacy risks, and a clunky user experience. Publishers should be pushing vendors for solutions that are built on a foundation of "data-in-place" collaboration - raw data never leaves the publisher's or advertiser's data stack. Predictive models are built on each side, and only the resulting, anonymized audiences need to be shared. Not only is it more secure and privacy-preserving, but it's exponentially faster, allowing for real-time audience activation without the bottlenecks of traditional clean rooms.

Rob: You talk about using predictive models instead of shuffling raw personal data between parties. For publishers, what does that actually change, does it really speed up how they can package and sell their audiences?

Joe: It changes everything. The traditional process of packaging and selling audiences is a slow, manual, and often inaccurate process that relies on guesswork and outdated data. By using predictive models, publishers can automate this entire workflow. They can instantly identify high-performing audiences based on advertiser demand, package them into scalable segments, activate them in real time, and most importantly, optimize for maximum outcomes. This eliminates the need for endless back-and-forth negotiations, cumbersome data transfers, and the risk of data leakage. Publishers can start to anticipate advertiser needs and create custom, high-value audiences on the fly - speeding up the sales cycle and dramatically increasing the value of every impression.

Rob: If I’m a publisher thinking about stepping into data collaboration, what are the first questions I should be asking myself to know if I’m ready, and how would I know I’m not?

Joe: Start with your data and your mindset. Are you collecting rich, first-party data that reflects the unique interests and behaviors of your audience? Do you have a clear understanding of your most valuable user segments? And, most importantly, are you ready to move beyond the limitations of the third-party cookie and embrace a new, privacy-first approach to audience monetization? If the answer to these questions is yes, you're ready.

If you're still clinging to the hope that a silver-bullet replacement for the third-party cookie will emerge - or looking for a quick fix rather than a long-term, sustainable solution - you’re not ready. The future of advertising is about collaboration - publishers who are willing to embrace this new reality will be the ones who thrive in the years to come.

Ready to rethink your data collaboration strategy?

As Joe makes clear, publishers can no longer afford to rely on outdated identifiers or broken collaboration models. Identity still has a crucial role to play, but only as part of a wider strategy that blends contextual, cohort, and behavioral data with predictive audiences. The future lies in privacy-first collaboration, and scalable identity strategies that truly capture the value of the entire audience.

👉 Dive deeper into these first-hand insights and discover how to future-proof your strategy in Permutive’s latest white paper: The hidden flaw in your data collaboration strategy (and how to fix it).

This is content created in paid partnership with Permutive We only feature partners who we believe bring real value to the publisher community.