55% Net New Audience: How Horizon Media and Intent IQ Are Turning Signal Loss Into Incremental Growth | AdExchanger

How Horizon Media and IntentIQ Drove 55% Incremental Reach for a National Retailer | Cannes 2026. Incremental growth is the metric every marketer is chasing and few can prove. A joint pilot between Horizon Media and IntentIQ is making the case that the gap between what brands know about their audiences and where those audiences can actually be reached is far larger — and far more valuable — than most realize.At Cannes Lions 2026, AdExchanger Editorial Director Sarah Sluis sat down with Fabrice Beer-Gabel, SVP of Strategy and Partnerships at IntentIQ, and John Koenigsberg, EVP and Head of Platforms at Horizon Media, to break down the results of their collaboration and what it signals for the future of identity in an AI-driven advertising environment.The Problem: Known Audiences in Unknown EnvironmentsMore than half of digital touchpoints in the US — across iOS, CTV, Chrome WebViews, and cookie-less browsers — lack reliable identity signals. For a sophisticated agency like Horizon, that means carefully built audience segments can't be activated across a significant portion of the addressable landscape. The audiences are known. The environments aren't.The Pilot: Amplification Without DisruptionWorking with a major national retail client, Horizon and IntentIQ applied audience amplification to Horizon's existing Blue audience segments — without replacing any part of the existing tech stack. The same DSP, the same media buying workflow, the same audiences. IntentIQ's amplification layer extended those segments into identity-constrained environments with over 90% accuracy, then tracked performance through the funnel with a baseline group and an exposed group running in parallel.The ResultsThe outcomes went beyond what the team anticipated going in:98% of the audience reached through amplification was net new — users Horizon's Blue audiences would not have found through existing activation alone55% net new users reached from the same audience segments, sustained over months of measurement — not a short-term test45% of those incremental users converted to purchases on the retailer's site, proving the reach extended through the full funnelWhy AI Makes Identity Accuracy More Critical, Not LessBoth Beer-Gabel and Koenigsberg are clear on a point that cuts against some of the optimism around AI in advertising: in a rules-based system, bad identity infrastructure wastes impressions. In an AI-driven connected system, inaccurate signals corrupt the entire system — and errors compound at the same speed as good results scale. The implication for brands building AI into their marketing infrastructure is significant. Identity accuracy isn't a nice-to-have. It's foundational.Beyond Deterministic vs. ProbabilisticThe conversation also addresses one of the industry's longest-running debates. Beer-Gabel argues the deterministic versus probabilistic frame is no longer the right question — deterministic approaches can be inaccurate, and probabilistic approaches can be highly accurate. What matters is validated accuracy, outcomes that hold up to measurement, and scale that doesn't introduce risk.The Takeaway for BrandsKoenigsberg closes with a point worth sitting with: even a brand with an exceptionally sophisticated first-party data strategy has significant latent potential sitting untapped in identity-constrained environments. The opportunity isn't to rebuild the data foundation — it's to make what you already have work harder, in more places, with results you can measure all the way through the funnel.Watch the full interview above to hear Fabrice Beer-Gabel and John Koenigsberg break down the pilot, the results, and what accurate identity amplification looks like at scale.