Resolve and PubMatic Partner for Privacy-First Ad Solution

Resolve, Choreograph’s privacy-first audience builder for brands and media partners, partnered with PubMatic on a six-week campaign to use machine learning to create new audience segments. These segments regained 80 percent of the impressions that had otherwise been lost.

Finding target audiences across a complex media landscape and fragmented attention is getting harder to manage. In addition, reaching potential customers has become more restricted with user data now vigorously safeguarded through anti-tracking initiatives like Apple’s Intelligent Tracking Prevention (ITP) and data protection legislation like the California Consumer Privacy Act (CCPA) or Europe’s General Data Protection Regulation (GDPR).

The Challenge: Unaddressable Inventory in a Tightly Regulated World

A 2022 IAB study revealed that, because privacy is now protected through techniques like machine learning and new regulations, advertisers can no longer reach 80 percent of their targeted population. As evolving data signals have impacted advertisement accessibility across the internet, PubMatic has made dedicated investments to address this challenge. Meanwhile, Resolve, a Choreograph company, has embraced the latest in federated learning, a method that trains models from a publisher’s inventory in a privacy-safe way without extracting or moving user data.

The partnership, bringing together PubMatic’s data with Resolve’s technology, sought measurable outcomes across previously unaddressable inventory while maintaining robust protection of user data.

Our Approach: Federated Learning for New Audience Segments

Resolve uses a decentralized approach. AI models live on the advertiser’s or publisher’s cloud, allowing them to process data where it’s stored, rather than taking the data out and storing it on their own servers. This is better for privacy as the data never leaves the publisher, and better for advertisers as Resolve builds models on deeper audience insights than ever before. These models learn from every interaction that audiences make, allowing them to track patterns and provide fresh insights into customer behaviors. The sharing of the resulting insights, rather than the data, gives brands and publishers control over how to curate better content experiences without needing to move sensitive data or compromise privacy.

Resolve then uses publisher data and traffic to create audience segments that can be applied to the total inventory, including previously untrackable environments. This gives advertisers access to over 95 percent of a publisher’s inventory, reducing total cost expenditure instantly.

With publisher integrations like the partnership with PubMatic, Resolve can use far more page and URL-level data to build deeper and richer audience segments. These new segments, and the audiences they represent, are owned by the publisher and Resolve. These segments can be made available for purchase, which provides immediate scale and cost benefits.

Data-Driven Results: Lowering Costs, Regaining Lost Impressions

Resolve drove more clicks for the same investment and advertisers, increasing the working media spend while ultimately reaching 95 percent of the targeted population. In addition, Resolve installed their models in PubMatic environments, helping to maintain privacy and allowing PubMatic to regain 80 percent of impressions that would be otherwise lost.