Collaborative activation system and method for unified audience management and cross-publisher execution

Abstract

A secure multi-party collaborative data activation system enables unified audience management and cross-publisher campaign execution with closed-loop optimization. The system introduces a tiered data clean room architecture comprising a centralized cross-publisher clean room synchronized with individual one-to-one clean rooms for each brand-publisher relationship. This tiered architecture enables both aggregated cross-publisher measurement and granular bilateral collaboration while maintaining strict data isolation and access controls, thus enabling closed-loop systems where brand owners can leverage cross-publisher measurement data from the centralized clean room to identify high-performing audience segments, then create refined audience segments through targeted collaboration in individual one-to-one clean rooms, and deploy new retargeting campaigns across the same group of publishers.

Description

Attorney Docket No. RAMP-00327-WOCOLLABORATIVE ACTIVATION SYSTEM AND METHOD FOR UNIFIED AUDIENCE MANAGEMENT AND CROSS-PUBLISHER EXECUTIONCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of US provisional patent application no. 63 / 777,425, filed on March 25, 2025. Such application is incorporated by reference as if fully set forth herein.BACKGROUND OF THE INVENTION

[0002] The field of the present invention is programmatic networks, and more specifically, secure multi-party collaborative systems that enable unified audience management, crosspublisher campaign execution, and closed-loop optimization within a clean room environment while managing diverse digital identifiers across multiple publishing platforms.

[0003] Programmatic networks operate as automated digital ecosystems where content creators and businesses can connect with a desired audience through strategically placed messaging content. These systems leverage sophisticated algorithms and real-time bidding processes to determine which messages appear to which consumers across websites and applications.

[0004] Programmatic networks function as intermediaries that analyze user behavior, demographics, and browsing patterns to deliver relevant messages at the optimal moment. This technology enables efficient message resource allocation by precisely targeting specific audience segments, while publishers can monetize their digital space through dynamically priced messaging opportunities. The entire process takes place in milliseconds, creating a seamless experience for consumers.

[0005] A programmatic network functions through a sophisticated technological infrastructure that connects the computing resources of multiple stakeholders in a near-instantaneous digital ecosystem. When a user visits a publisher website or application at a publisher server through a consumer electronic device such as a computer or mobile phone, an automated auction begins in the background. The network analyzes available messaging space on the web page or app and sends a bid request to potential buyers. This request includes anonymous data about the user. Participating buyers use demand-side platforms (DSPs) to set parameters for their campaigns ahead of the auction. These parameters may include target audience characteristics, maximumAttorney Docket No. RAMP-00327-WObid amounts, and performance goals. Supply-side platforms (SSPs) represent the publishers who provide the digital space within the automated bidding system. In response to the user's actions, the network's exchange conducts a real-time auction where algorithmic bidding occurs in milliseconds.

[0006] The highest bidder wins the opportunity to display its messaging content in that specific space to that specific user. This entire transaction happens programmatically, i.e., without human intervention, while the page or app loads. The transaction is entirely transparent to the user.

[0007] The programmatic network ecosystem has become increasingly complex, with brand owners needing to execute campaigns across multiple publishers, each with unique identifier requirements, data formats, and measurement methodologies. Current solutions often involve fragmented workflows where brand owners must separately manage audience segments, campaign activation, and performance measurement for each publisher relationship.

[0008] Retail media networks (RMNs) form a specialized class of computational components within a programmatic network. RMNs are owned and operated by particular retailers, and this exclusive nature gives them distinct operational characteristics within the programmatic ecosystem. RMNs typically include various channels, such as web pages, apps, and other digital properties, offered by a particular retail company to third-party brand owners. What makes RMNs particularly powerful to brand owners is their access to their own first-party data, which serves as their primary competitive advantage in the programmatic landscape. First-party data is data that a retailer has collected directly from its own customers.

[0009] RMNs allow third-party brand owners to create messaging across a retailer's digital channels to help customers discover and learn about the retailer's products and services. This functions like a "digital store shelf" that increases visibility to shoppers, similar to how in-aisle specials stand out in physical stores. While initially limited to on-site inventory, RMNs increasingly are integrated as a part of programmatic networks, connecting their inventory to off-site properties and third-party media for multi-channel reach. Advanced RMNs offer multiple programmatic capabilities, including DSP integrations for broader reach and precise targeting. They execute campaigns via preferred DSPs that can access both on-site and off-site inventory, including display, video, connected TV, streaming audio, and digital out-of-home channels.Attorney Docket No. RAMP-00327-WO

[0010] Within the programmatic ecosystem, RM Ns typically offer several promotional formats.One format is on-site promotional placements. These appear on the retailer's website and mobile app, similar to placing a product at eye level on a store shelf. They include homepage banners, category pages, and product detail pages. Off-site promotions leverage a retailer's first-party data to reach consumers across platforms outside of the retailer's owned digital properties, including external websites, apps, social media channels, and streaming platforms. Social media extensions allow RMNs to integrate with social media platforms like Pinterest and Facebook / lnstagram, allowing brands to amplify their messaging through social channels while still leveraging the retailer's data.

[0011] RMNs offer significant operational advantages within programmatic networks. For retailers, they provide a new revenue stream through the provision of digital space. For brand owners, they offer access to first-party data for creating highly targeted message campaigns. And for consumers, the messages appear more relevant, making them feel less intrusive and more helpful. Another key operational benefit is closed-loop attribution. Since all activity occurs on a retailer's site or using retailer data, those wishing to send messages can directly tie spending on their DSP to revenue from that retailer. This provides better opportunities to analyze attribution and return on investment (ROI) compared to traditional programmatic channels.

[0012] The programmatic network landscape is evolving in ways that benefit RMNs of all sizes.Programmatic enhancements led by automation have reduced operational costs and increased targeting efficiencies, allowing even smaller RMNs to run campaigns at scale without investing in large internal operations teams. By tapping into exchanges and SSPs, these smaller networks can access more promotional opportunities without negotiating terms for each avenue individually. As RMNs continue to evolve, they're becoming increasingly integrated with the broader programmatic network landscape, offering brands more sophisticated ways to reach consumers with messaging at various stages of their shopping journey while leveraging the valuable first-party data that only retailers possess.

[0013] A related class of components within programmatic networks is commerce media networks (CMNs). CMNs are typically maintained by retailers or platforms that possess extensive proprietary data from actual purchase histories, browsing behavior, and loyaltyAttorney Docket No. RAMP-00327-WOprograms. Unlike traditional publishers, they own both the relationship with the consumers and possess detailed information about what those consumers actually buy, not just what they view. An important strength of CMNs lies in their ability to leverage knowledge about specific customers into messaging for related or complementary products or services, potentially including products or services from different brand owners. CMNs thus allow brand owners to reach specific consumer segments based on purchase behavior, even if the product that is the subject of the messaging isn't directly related to the initial purchase. CMNs thereby empower brands to better understand customer value and intent beyond an immediate transaction.

[0014] Within the programmatic network, CMNs function as specialized publishers with privileged insights. They activate their first-party data to create premium audience segments based on genuine purchase intent and behavior patterns. For example, a retail CMN might identify segments such as "frequent organic product purchasers" or "luxury accessories browsers who convert within 30 days." When a user visits a CMN's digital property or a connected site within the network, the CMN can leverage this rich first-party data to facilitate highly targeted messaging. Brand owners can bid specifically for messaging access to these qualified audiences who have demonstrated concrete commercial intent rather than just general interest.

[0015] The value proposition of CMNs stems from their ability to close the loop between exposure and purchase attribution. They can demonstrate not just impressions, but actual sales impact, thereby showing brand owners how their programmatic investments translate to revenue and provide a more accurate assessment of ROI. This closed-loop measurement capability makes CMN inventory particularly valuable within the broader programmatic network landscape.

[0016] Additionally, many sophisticated CMNs include proprietary algorithms that predict future purchase behavior based on historical patterns, allowing for even more precise audience targeting and higher conversion rates for brand owner partners. They often offer specialized formats that appear natively within the consumer experience, creating less disruptive and more contextually relevant messaging opportunities.

[0017] Retail Media Networks (RMNs) and Commerce Media Networks (CMNs) represent different approaches to leveraging consumer data for messaging purposes. While they shareAttorney Docket No. RAMP-00327-WOsome similarities, they have distinct characteristics and applications. Retail media relies on first- party data gathered via a specific retailer's website, while commerce media augments first-party insights with inputs from other publishers and brand sites. This creates a more holistic view of the customer, offering both quality and scale needed to engage audiences across the digital ecosystem. Commerce media uses first-party data from various sources to create personalized messaging across multiple channels, with the intent of creating better ROI as the shopping intent is higher and messaging is contextually placed. In contrast, retail media focuses on consumers already browsing within specific retail environments, offering a more targeted scope of engagement. Commerce media expands upon the retail media model by including a broad network of publishers across the open internet, in addition to retailers. Commerce media also enables retail, non-retail, and non-endemic messages to use commerce data and artificial intelligence (Al) to more effectively engage audiences on those publishers and retailers.

[0018] Commerce media may offer greater scalability through data aggregation from multiple sources for broader reach; more comprehensive audience insights by examining data across industries; and versatility for businesses that don't operate a retail platform or sell physical goods but want to monetize their audience. RM Ns, by contrast, typically focus on providing messaging opportunities on their owned digital properties, with some extending to off-site opportunities. They excel at influencing purchase decisions at the point of sale. Both offer valuable opportunities for brand owners, with CMNs providing greater scale and cross-industry insights, while RMNs excel at targeting consumers with messaging at the critical moment of purchase decision.

[0019] Existing programmatic networks suffer from several limitations. Individual publisher platforms operate in silos, requiring brand owners to set up and manage separate campaigns with varying metrics and reporting across different publishers. Data Management Platforms (DMPs) focus primarily on data storage and segmentation but lack integrated cross-publisher execution and measurement capabilities within secure environments. While clean room technologies enable secure data sharing and collaboration, existing solutions primarily focus on data matching and analysis rather than active campaign execution and optimization.

[0020] Current technology also struggles with identifier management complexity. Publishers accept different types of identifiers including Mobile Advertising Identifiers (MAIDs), HTTP cookies, proprietary identifier systems, and hashed personally identifiable information (PH).Attorney Docket No. RAMP-00327-WOBrand owners must navigate these heterogeneous identifier requirements manually, often resulting in suboptimal reach and accuracy.

[0021] Furthermore, existing systems lack sophisticated cross-publisher measurement and optimization capabilities. Brands cannot easily measure campaign performance consistently across participating publishers or leverage aggregated insights to optimize audience targeting in real-time across multiple platforms simultaneously.

[0022] Data clean rooms are becoming an important technology for privacy-compliant data sharing. A data clean room is a secure virtual environment where data owners can collaborate and analyze data in a privacy-compliant way, without exposing confidential information or violating privacy regulations. Unlike traditional data exchange methods, data does not need to be moved to an external system, which prevents the exposure of sensitive information such as personally identifiable information (PI I). Clean rooms have emerged as a crucial technology for data collaboration in the privacy-focused era of programmatic networks. As regulations tighten and third-party cookies phase out, these secure clean room environments provide a solution for sharing and analyzing data while maintaining privacy compliance.

[0023] With respect specifically to programmatic networks, data clean rooms are collaborative environments where two or sometimes more participants (brand owners, publishers, message senders, or other entities) come together to share and combine their respective first-party data. First, two or more parties (for example, a publisher and a brand owner) collect, compile, and aggregate their first-party data at the user level. These datasets don't need to be identical, but the parties do need a means of matching. Once in the clean room, datasets can be matched using commonly hashed identifiers like email addresses, phone numbers, or user IDs. Inside the data clean room, both parties virtually load their respective data sets but cannot see each other's raw data. Data analysis and processing takes place with aggregated and anonymized outputs, protecting user privacy while still providing valuable information for segmentation and campaign optimization. Data clean rooms share aggregated rather than customer-level data with collaborating parties, while maintaining strict controls. First-party data from retailers, for example, is poured into the same space to see how it matches up with the aggregated data, allowing the retailers to determine whether they are over-serving messages to the same audiences.Attorney Docket No. RAMP-00327-WO

[0024] Data clean rooms offer several significant advantages for programmatic networks. First, they protect user privacy and ensure regulatory compliance, serving as secure environments where parties can safely merge, analyze, and collaborate on their first-party data without sharing any personally identifiable information. They help break down data silos by bridging gaps across platforms and devices, enabling cross-platform tracking that paints a holistic picture of user behavior. They also facilitate partnerships between those wishing to create messaging and publishers, opening the door to richer insights and new segments in a privacy-first way. For those wishing to create messages, data clean rooms can enable customer acquisition by allowing collaboration with publishers and data providers whose first-party data may include the messaging party's potential customers. By allowing customer audiences to be transformed into lookalikes by a publisher or data provider, messaging parties can use their first-party data to identify new prospects. By analyzing the overlap between conversions and impressions within a publisher's network, publishers can provide closed-loop measurement to help messaging parties understand their return on messaging spend.

[0025] Data clean rooms, despite the power that they offer for privacy-safe data sharing, further complicate already complex programmatic networks, which are already held back by the limitations discussed above. Thus a system and method to utilize data clean rooms for unified audience management and cross-publisher execution is desired.SUMMARY OF THE INVENTION

[0026] The present invention is directed to a secure multi-party collaborative data activation system that enables unified audience management and cross-publisher campaign execution with closed-loop optimization. The system introduces a tiered data clean room architecture that, in certain embodiments, facilitates seamless collaboration between publishers, brand owners, and commerce media network / retail media network (CMN / RMN) owners while maintaining data privacy and security at multiple levels.

[0027] An innovation in certain embodiments is the implementation of a multi-tiered clean room structure comprising a centralized cross-publisher clean room synchronized with individual one-to-one clean rooms for each brand-publisher relationship. This tiered architecture enables both aggregated cross-publisher measurement and granular bilateral collaboration while maintaining strict data isolation and access controls.Attorney Docket No. RAMP-00327-WO

[0028] The system in certain embodiments provides a centralized platform for brand owners to define, create, and manage target audiences across multiple publishers, eliminating the need to manage disparate audience segments and activation strategies for each publisher individually. The multi-tiered clean room environment enables measuring campaign performance consistently across all participating publishers in the centralized clean room while facilitating subsequent retargeting through individual publisher-specific clean rooms.

[0029] The invention in such embodiments enables closed-loop systems where brand owners can leverage cross-publisher measurement data from the centralized clean room to identify high-performing audience segments, then create refined audience segments through targeted collaboration in individual one-to-one clean rooms, and deploy new retargeting campaigns across the same group of publishers. This iterative optimization process drives continuous improvement in campaign performance and return on investment (ROI).

[0030] One aspect of certain embodiments of the invention enables retailer-specific retargeting with publisher inventory access control, allowing brands to retarget specific audiences on retailer websites by selectively choosing publishers that have authorized access to that retailer's inventory through the tiered clean room structure. The system also introduces intelligent identifier management, handling various types of identifiers within the same campaign and ensuring that publishers receive appropriate identifiers based on optimal match rates or specific goals.

[0031] In certain embodiments, the invention facilitates unified audience / segment management. The system provides a centralized platform for brand owners to define, create, and manage target audiences across multiple publishers. This eliminates the need for brands to manage disparate audience segments and activation strategies for each publisher, streamlining the campaign execution process.

[0032] In certain embodiments, the invention facilitates cross-publisher measurement and optimization. The system provides the ability to measure campaign performance consistently across all participating publishers within the clean room environment. This aggregated measurement data allows brand owners to gain a holistic view of campaign effectiveness and optimize audience targeting and creative strategy in real-time across all publishers.

[0033] In certain embodiments, the invention comprises a secure multi-party clean room environment, ensuring data privacy and compliance with relevant regulations. This environment enables both publishers on the open web as well as "walled gardens" (i.e., closed ecosystemsAttorney Docket No. RAMP-00327-WOwhere a single company controls both the supply and demand sides of transactions, as well as the data and technology infrastructure) to share necessary measurement data with brand owners without compromising sensitive user information, allowing for cross-publisher measurement and optimization. This multi-party clean room is synchronized with individual one-to-one clean rooms for each brand-publisher relationship, facilitating subsequent retargeting.

[0034] In certain embodiments, the invention provides iterative optimization and retargeting.This enables the closed-loop system where brand owners can leverage cross-publisher measurement data to identify high-performing audience segments and create new retargeting campaigns that are then deployed across the same group of publishers. This iterative optimization process drives continuous improvement in campaign performance and ROI.

[0035] In certain embodiments, the invention provides for retailer-specific retargeting with publisher inventory access control, for example, the ability for brand owners to retarget specific audiences on a retailer's website by selectively choosing a publisher that has authorized access to that retailer's inventory. This granular control allows brands to execute targeted campaigns on retailer sites while respecting data access agreements and inventory availability, further refining campaign efficiency and reach. This will also enable retailers to bring their publishers with them into a brand owner relationship so that brand owners can choose from among them.

[0036] In certain embodiments, the invention provides for unified identifier management and activation. The system manages various types of identifiers (online identifiers such as browser cookies, Mobile Ad IDs (MAI Ds), and offline identifiers such as email addresses and telephone numbers) within the same campaign. The system ensures that the publisher or other destination receives one or many of these identifiers based on the best match rate or the specific goal. This intelligent identifier distribution optimizes reach and accuracy while respecting privacy considerations and platform capabilities. This means that a campaign that starts with email can end up with MAID or browser cookies based on the best match rate of the individual user and the destination platform.

[0037] These and other features, objects and advantages of the present invention will become better understood from a consideration of the following detailed description of the preferred embodiments and appended claims in conjunction with the drawings as described following:Attorney Docket No. RAMP-00327-WOBRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 is a system architecture diagram for an embodiment of the present invention.

[0039] Figure 2 is a detailed system architecture diagram illustrating a single data delivery group from Figure 1 being used across multiple types of data, identity types, and multiple publishers according to an embodiment of the present invention.

[0040] Figure 3 is a data delivery group user interface (Ul) showing the selection of audience segments according to an embodiment of the present invention.

[0041] Figure 4 is a data delivery group Ul showing the selection of destination connections according to an embodiment of the present invention.

[0042] Figure 5 is a data delivery group Ul showing setup options according to an embodiment of the present invention.

[0043] Figure 6 is the data delivery group dashboard Ul according to an embodiment of the present invention.

[0044] Figure 7 is the data delivery group dashboard Ul of Figure 6 when the destination connection view is selected according to an embodiment of the present invention.

[0045] Figure 8 is the data delivery group dashboard Ul of Figure 6 when the activated segment toggle is selected according to an embodiment of the present invention.

[0046] Figure 9 is a Ul providing an audit log for a data delivery group according to an embodiment of the present invention.

[0047] Figure 10 is an insights and measurement Ul for a data delivery group according to an embodiment of the present invention.

[0048] Figure 11 is a data activation permissions Ul according to an embodiment of the present invention.

[0049] Figure 12 is the data activation permissions Ul of Figure 11 when "to my destination connection" is selected according to an embodiment of the present invention.

[0050] Figure 13 is a hardware diagram of a computing component of a multi-compute cluster for implementing an embodiment of the present invention.DETAILED DESCRIPTION OF THE INVENTION

[0051] Before the present invention is described in further detail, it should be understood that the invention is not limited to the particular embodiments described, and that the terms used in describing the particular embodiments are for the purpose of describing those particularAttorney Docket No. RAMP-00327-WOembodiments only, and are not intended to be limiting, since the scope of the present invention will be limited only by the claims.

[0052] By way of a system architecture overview, the secure multi-party collaborative data activation system operates within a distributed architecture that facilitates secure data sharing and campaign execution across multiple publishers while maintaining strict privacy controls and compliance with relevant regulations. The system architecture centers around Data Delivery Groups (DDGs), which serve as centralized orchestration units for managing audience data distribution across multiple publishers. Each DDG encapsulates one or more audience segments along with associated metadata, identifier mappings, and distribution rules that govern how data is shared with participating publishers.

[0053] DDGs function as the primary abstraction for managing complex cross-publisher data activation workflows. Each DDG contains audience segments derived from various data sources including first-party customer data, collaborative audience segments created through secure data partnerships, and enriched segments incorporating third-party data sources. The DDG architecture implements sophisticated identifier translation and distribution layers that dynamically resolve heterogeneous identifier requirements across the publisher landscape. The system autonomously maps and transforms audience data into appropriate identifier formats as mandated by each publisher, including Mobile Advertising Identifiers (MAI Ds), HTTP cookies, proprietary third-party system provider identifiers such as the RampID® system from LiveRamp, Inc., Customer / Platform Identifiers (Cl Ds), and cryptographic hash representations of personally identifiable information.

[0054] The system includes a multi-tiered clean room structure that enables both aggregated cross-publisher / walled garden analytics and granular bilateral collaboration while maintaining strict data isolation. The tiered architecture comprises a centralized cross-publisher clean room that aggregates data from multiple publishers for unified measurement, synchronized with individual one-to-one clean rooms that facilitate targeted collaboration between brands and specific publishers and / or walled gardens.

[0055] The centralized cross-publisher clean room serves as the primary measurement and analytics environment where campaign performance data from all participating publishers is aggregated, deduplicated, and analyzed to provide unified insights across the entire publisher ecosystem. This centralized environment enables calculation of cross-publisher metricsAttorney Docket No. RAMP-00327-WQincluding deduplicated reach, frequency analysis, and attribution modeling that would be impossible within individual publisher silos.

[0056] Individual one-to-one clean rooms operate as dedicated collaboration environments between brand owners and specific publishers, enabling targeted audience refinement and publisher-specific optimization. These bilateral clean rooms facilitate the creation of highly refined audience segments using publisher-specific data and insights while maintaining strict access controls that prevent data leakage between different publisher relationships.

[0057] The synchronization between the centralized clean room and individual clean rooms enables a novel closed-loop optimization workflow where insights derived from cross-publisher analysis in the centralized environment can trigger targeted refinement activities in specific one- to-one clean rooms, with the results then being propagated back to the centralized environment for subsequent campaign iterations.

[0058] The system also includes a closed-loop optimization capability that leverages the multitiered clean room architecture to enable iterative campaign refinement based on crosspublisher performance insights. The system continuously analyzes aggregated performance data in the centralized clean room to identify high-performing audience segments and automatically generates optimization recommendations.

[0059] When performance patterns are identified in the centralized clean room, the system initiates targeted data collaboration sessions with individual publishers within their respective one-to-one clean rooms. These dedicated bilateral environments enable the creation of highly refined audience segments using advanced techniques including implementation of new segmentation rules, execution of complex SQL queries, and deployment of sophisticated data science models that leverage both cross-publisher insights and publisher-specific data.

[0060] The tiered clean room structure ensures that sensitive publisher-specific data remains within individual one-to-one clean rooms while enabling the refined audience segments to be activated across the broader publisher network through the centralized coordination layer. This architecture enables continuous optimization cycles that improve campaign performance while maintaining strict data governance and privacy controls.

[0061] The system implements novel identifier management capabilities that intelligently distribute different identifier types based on optimal match rates and platform-specific requirements. For each audience segment and publisher combination, the system determines the most appropriate identifier types to maximize reach and accuracy while respecting privacyAttorney Docket No. RAMP-00327-WOconsiderations and platform capabilities. This intelligent identifier distribution means that campaigns can dynamically shift between different identifier types (hashed email to MAID to cookies, for example) based on individual user match rates and destination platform capabilities, providing flexibility and efficiency not found in existing systems.

[0062] An overview of the operation workflow may now be described, beginning with the initial campaign setup and data consolidation. Brand owners initiate the process by consolidating first-party customer audiences from disparate data sources including Customer Relationship Management (CRM) systems, loyalty program databases, Point-of-Sale (POS) transactional records, and customer data platforms (CDPs). These baseline audiences are encapsulated within initial DDGs that serve as the foundation for subsequent cross-publisher activation.

[0063] The system supports integration of additional data sources including second-party collaborative data from RMN and CMN owners and third-party audience data from certified data providers. Each data source integration maintains strict access controls and usage restrictions that are enforced throughout the campaign lifecycle.

[0064] The system maintains detailed profiles for each integrated publisher that specify supported identifier types, data format requirements, and platform-specific constraints. When DDGs are configured for distribution, the system automatically validates identifier compatibility and performs necessary transformations to ensure seamless data delivery.

[0065] For publishers supporting multiple identifier types, the system implements intelligent selection algorithms that choose optimal identifier combinations based on historical match rate performance, campaign objectives, and real-time availability assessments.

[0066] Once DDGs are activated, the system orchestrates simultaneous campaign execution across all specified publishers and / or walled gardens while maintaining detailed logging of all data distribution activities. Each publisher collects granular performance metrics including message exposure logs, user interaction events, and conversion tracking data.

[0067] Performance data collection implements standardized measurement frameworks that enable consistent metric calculation across heterogeneous publisher platforms. The system accounts for different measurement methodologies and normalizes metrics to enable accurate cross-publisher comparison and aggregation.

[0068] All campaign performance data is consolidated within the centralized cross-publisher clean room where sophisticated deduplication algorithms eliminate duplicate user countingAttorney Docket No. RAMP-00327-WOacross publishers. The system implements advanced statistical techniques to provide accurate reach and frequency calculations that account for user overlap across different platforms.

[0069] The tiered clean room architecture enables this aggregation while maintaining strict data isolation, whereby raw publisher data remains within individual one-to-one clean rooms, while only aggregated and anonymized performance metrics are shared with the centralized environment. This structure ensures compliance with data privacy requirements while enabling comprehensive cross-publisher analytics.

[0070] Aggregated performance metrics are made available through comprehensive analytics dashboards that provide both high-level campaign summaries and granular publisher-specific and / or walled-garden specific performance breakdowns. Users can analyze performance across multiple dimensions including audience segment, publisher platform, identifier type, and temporal patterns.

[0071] Based on aggregated performance analysis in the centralized clean room, the system identifies high-performing audience segments and generates optimization recommendations. When optimization opportunities are detected, the system initiates targeted data collaboration sessions with specific publishers in their respective one-to-one clean rooms to create refined audience segments.

[0072] These bilateral optimization sessions leverage the full spectrum of available data including cross-publisher performance insights from the centralized clean room, publisherspecific behavioral data within the one-to-one clean room, and advanced analytics techniques. The tiered clean room structure ensures that sensitive publisher data remains isolated while enabling refined audience segments to be created and then activated across the broader publisher network. Refined audience segments created in individual clean rooms are then incorporated into new DDGs that can be activated across the same publisher network through the centralized coordination layer, creating a continuous optimization loop that spans both tiers of the clean room architecture.

[0073] The system enables sophisticated messaging scenarios including multi-brand owner collaborative campaigns where non-competing brand owners can partner to reach shared audiences while maintaining individual campaign measurement and optimization. The system also supports advanced holdout testing capabilities that maintain consistent control groups across multiple publishers to enable accurate campaign effectiveness measurement.Attorney Docket No. RAMP-00327-WO

[0074] Cross-publisher frequency capping prevents over-exposure of messages to the same individuals across multiple publishers by tracking impression frequency within the clean room environment and enforcing frequency limits across all participating platforms.

[0075] The system implements comprehensive data security measures including encryption of all data in transit and at rest, role-based access controls that limit data access based on user permissions and organizational policies, and detailed audit logging of all data access and modification activities. Privacy controls ensure compliance with relevant regulations including GDPR, CCPA, and other applicable data protection frameworks. The system maintains detailed records of data lineage and usage permissions to support compliance reporting and user rights requests.

[0076] The system architecture is designed for horizontal scalability to support large-scale campaign execution across numerous publishers and audience segments. Distributed processing capabilities enable parallel execution of data transformation, identifier matching, and performance aggregation tasks. Caching mechanisms optimize frequently accessed data and pre-computed analytics results to ensure responsive user experiences even with large datasets. The system implements intelligent data partitioning strategies that distribute computational load across available resources while maintaining data consistency and integrity.

[0077] The system provides comprehensive application programming interfaces (APIs) that enable integration with existing message technology stacks including demand-side platforms, customer data platforms, and programmatic network automation systems. Standardized data formats and protocols ensure seamless interoperability with third-party systems.

[0078] Real-time event streaming capabilities enable immediate propagation of campaign performance data and optimization signals across the system architecture. Webhook mechanisms provide flexible notification systems that can trigger external system actions based on campaign events and performance thresholds.

[0079] Referring now to Fig. 1, the architecture of a system according to a preferred embodiment may be described in greater detail. The system orchestrates diverse messaging identifiers (including, but not limited to, MAI Ds, Cookies, proprietary ID systems, and hashed PH such as email addresses) in a unified campaign activation and measurement framework. This system employs a closed-loop architecture facilitated by centralized data delivery groups (DDGs), specifically data delivery groups 10a and 10b, which ensure seamless, privacypreserving, and efficient campaign management across a complex ecosystem of publishers.Attorney Docket No. RAMP-00327-WO

[0080] Initially, brand owner A consolidates its first-party customer data 12 from disparate data sources, encompassing customer relationship management (CRM) systems, loyalty program databases, and point-of-sale (POS) transactional records. These audiences, denoted as 14a, 14b, and 14c, are then encapsulated within data delivery group 10a. These audiences are derived solely from brand owner A's first-party data 12, so they are inherently exempt from initial data usage or publisher-specific restrictions. In this exemplary embodiment, brand owner A initiates a preliminary campaign activation by distributing these baseline audiences (14a, 14b, and 14c) to three distinct publishing entities: publishers 16a, 16b, and 16c. Each of these publishers operates within a unique identifier ecosystem, characterized by distinct acceptance criteria for identifier types and encoding schemes.

[0081] Publisher 16a in this example exclusively accepts Mobile Advertising Identifiers (MAIDs) for audience targeting and campaign activation. Publisher 16b supports a hybrid identifier model, accepting a proprietary third-party ID set and HTTP browser cookies. Publisher 16c mandates using personally identifiable information (PH), specifically, cryptographic hash representations of email addresses.

[0082] Data delivery group 10b implements an identifier translation and distribution layer. This layer dynamically resolves the inherent heterogeneity of identifier requirements across the publisher landscape. It autonomously maps and transforms brand A's audience data 12 into the appropriate identifier format as mandated by each publisher. This automated and intelligent routing ensures seamless delivery of audience segments 14a, 14b, 14c, and enables the initiation of the messaging campaign across all three publishers. Each publisher, in turn, collects granular performance metrics, including, but not limited to, ad exposure logs and conversion event logs, for each of brand owner A's delivered audience segments 14a, 14b, and 14c.

[0083] All campaign performance data from the three publishers is subsequently aggregated and consolidated within a secure data collaboration cross-publisher clean room 18. This clean room environment performs important data processing operations, including deduplication and cross-publisher aggregation of performance metrics. This data aggregation enables the generation of comprehensive and unified reports on audience performance across the entire publisher ecosystem. The resulting reports provide rich insights, encompassing, without limitation, lift and attribution analysis, quantifying the incremental impact of the messaging campaign; deduplicated reach and frequency metrics, providing an accurate count of unique consumers reached and the frequency of message exposures; projected scaled reach frequencyAttorney Docket No. RAMP-00327-WQestimates, forecasting the potential reach and frequency of the campaign at a larger scale; incremental reach calculations, measuring the additional reach achieved through specific campaign optimizations; and return on ad spend (ROAS) calculations, assessing the overall financial effectiveness of the campaign.

[0084] These granular performance insights are then accessible to Brand A's campaign managers, data analysts, and / or Al agents, as applicable in particular embodiments. These stakeholders can leverage the insights to formulate data-driven optimization strategies, including adjustments to existing audience segments and creating new, publisher-specific audience segments based on the observed performance patterns.

[0085] In a specific optimization scenario 20, brand A engages in targeted, one-to-one data collaboration sessions with each publisher within dedicated, isolated clean room environments. This allows for the creation of highly refined audience segments using advanced techniques, such as the implementation of new segmentation rules, the execution of complex SQL queries, or the deployment of sophisticated data science models. This publisher-specific refinement is predicated upon the insights derived from the initial cross-publisher / walled garden clean room analysis. In clean room 22a, audience 24a is created through a direct data collaboration between brand owner A using first-party data 12 and publisher 16a, with the usage of this new audience being explicitly restricted to the publisher 16a platform and exclusively using MAID identifiers. Audience 24b is created through a data collaboration between brand A and publisher 16b, with the usage limited to the publisher 16b platform and the identifier space restricted to third-party proprietary IDs due to suboptimal performance observed with HTTP cookies on this platform. Audience 24c is created through a data collaboration between brand A and publisher 16c, with the usage limited to the publisher 16c platform and exclusively using Pll in the form of cryptographic hash representations of email addresses.

[0086] Subsequently, brand owner A establishes a new data delivery group 10b. This group manages the identifier heterogeneity and ensures the accurate and compliant transmission of the appropriate identifier type for each refined audience segment 24a, 24b, and 24c.Furthermore, it enforces the publisher-specific usage restrictions imposed during the one-to- one data collaboration phase. This updated data delivery initiates a new iteration or retargeting of the campaign across the three publishers, which updates the performance data collection, including exposure and conversion logs, for each of brand owner A's optimized audience segments.Attorney Docket No. RAMP-00327-WO

[0087] Cross-publisher data collaboration clean room 18 is then updated with this additional performance data, encompassing all campaign performance data for brand owner A's refined audiences across the three publishers. The performance data is again subjected to deduplication and aggregation across all publishers, and the optimization process can be iteratively repeated as required to refine campaign performance further and achieve desired outcomes.

[0088] Referring now to Fig. 2, the process for using a data delivery group across multiple types of data (first party, second party, and third party), multiple identity types, and multiple publishers is illustrated, thus creating a system and / or model for managing complex campaigns from a single end-user representation or business object.

[0089] Brand owner B initiates the process by consolidating its proprietary first-party customer data 28. These audiences, 26a, 26b, 26c, and 26d, derived from many internal data repositories, including but not limited to CRM systems, loyalty program databases, and POS transactional records, serve as the foundational baseline for subsequent data activation. Inherently, these audiences are free from third-party data usage restrictions, as they originate exclusively from brand owner B's controlled data environment.

[0090] Simultaneously, third-party data seller 30 provides a valuable source of supplementary audience data 26e, characterized by rich demographic profiles and inferred buying intent signals. This data, distinct from brand A's first-party data 28, augments audience coverage and provides insights into broader segments, demonstrating the system's ability to incorporate diverse data sources and enrich audience profiles.

[0091] In parallel, brand A engages in a strategic data collaboration with commerce media network owner A, leveraging the CMN's unique insights into audience behaviors within its network. (A CMN is used in this example, although an RMN could also be used similarly.) Through a secure, dedicated data collaboration clean room 32, brand B and CMN owner A jointly derive a new audience 36a. CMN owner A retains exclusive control over the distribution of this audience segment to a specific publisher A, via its proprietary publisher seat / configuration. Furthermore, CMN owner A imposes a commercial condition, precisely a fee or percentage of media cost 38, on the utilization of this segment. This exemplifies the system's novel ability to manage complex commercial agreements and granular access controls within a collaborative data environment.Attorney Docket No. RAMP-00327-WO

[0092] Further, brand owner B engages in a data collaboration with publisher B in brand / publisher data clean room 34, leveraging publisher B's proprietary insights into audience behavior on its platform. Within the collaborative clean room, brand B and publisher B derive a new audience 36b. Additionally, brand owner B integrates third-party data from data seller 30, augmenting its first-party audience to create another new audience 36c. Publisher B imposes a platform-specific data usage restriction, limiting the application of all collaboratively created audiences exclusively to its platform, or alternatively to another publisher with which it has a relationship with when a CMN or RMN is acting as an agency within its own organization to serve brand owners. Data seller 30 imposes a commercial condition, precisely a fee or percentage of media cost 40, on the utilization of this augmented segment. This demonstrates the system's ability to facilitate multi-party data enrichment and manage granular, platformspecific data permissions.

[0093] Consequently, brand owner B establishes a comprehensive data delivery group 42 that encapsulates all previously defined audiences, including direct audiences from its first-party data 26a, a direct audience from data seller 30, audience 26e, and all collaboratively derived audiences 26b, 26c, and 26d. It is crucial to note that various publishers and platforms exhibit divergent requirements and privacy controls about first-party and third-party data, which adds a significant layer of complexity to the data activation process. Brand owner B's data delivery group 42 manages the data distribution and permissions with precise control, ensuring that each audience segment is delivered to the appropriate publisher with the correct permissions and usage restrictions.

[0094] Data delivery group 42 further supports a diverse array of identifier types specific to each publisher, including but not limited to third-party proprietary IDs, MAIDs, HTTP browser cookies, customer / platform identifiers (CIDs), and Pll in the form of cryptographic hash representations of email addresses. This intricate mapping, switching, and orchestration necessitate a robust status tracking mechanism to manage the complex allowed and possible configurations matrix. Each audience is subjected to rigorous validation against the specific permissions, data governance policies, and privacy regulations (first-party and third-party) supported by each publisher. Specifically, publisher A receives audience 26a and 36a at activation 44a; publisher A receives audience 26e at activation 44b; publisher B receives audience 26a and 36b at activation 44c; and publisher B receives audience 36c and 26c at activation 44d. The system meticulously ensures that each audience is delivered with theAttorney Docket No. RAMP-00327-WOcorrect identifier type and permissions, optimizing reach and accuracy while respecting privacy considerations and platform capabilities.

[0095] With reference now to Figs. 3-12, user interfaces (Ills) and user experience (UX) will be described. This system provides the visibility to track such a complex matrix of possibilities without confusing the end-users. In particular, the following section will detail the key Ul components and interaction paradigms that enable intuitive, efficient, and secure management of cross-publisher campaigns. The design focuses on providing a "single pane of glass" for complex operations, ensuring that users, whether brand owners, publishers, or Al agents, can seamlessly navigate and leverage the system's powerful capabilities.

[0096] Referring now to Figs. 3-6, the data delivery group Ul component is the central hub for managing audience data distribution across multiple publishers. This component provides a comprehensive overview of audience segments, associated identifiers (MAIDs, cookies, proprietary third-party IDs, hashed PH), and the intended recipients (publishers). Functionality includes creating, modifying, and deleting data delivery groups, with explicit visual representations of data flows and identifier mappings. A detailed view allows users to inspect the specific data points included in each segment, the transformation rules applied for identifier matching, and the access permissions granted to each publisher. This component ensures transparency and control over data sharing, allowing users to meticulously manage the activation of their audience segments in alignment with privacy regulations and publisher requirements.

[0097] The first step in crafting a data delivery group involves selecting the specific audience segments that will be activated and distributed, as shown in the data delivery groups Ul 50a of Fig. 3. This decision is made through the data delivery groups Ul 50a selector interface 52, a user-friendly tool that allows the user to choose one or multiple audience segments. To aid in this crucial selection, the system provides rich metadata for each segment, including its current status (active, paused, or draft); an estimated segment size (vital for planning reach and budget allocation); and the dates it was last modified or created.

[0098] Additionally, fee indicator 54 of data delivery groups Ul 50a provides transparency regarding any additional fees associated with a segment. These fees might stem from collaborative data activities or the use of third-party data. Once the user has chosen the audience segments, the system performs real-time validation checks to ensure their eligibility for distribution, adhering to data governance policies and privacy regulations. Real-timeAttorney Docket No. RAMP-00327-WOvalidation prevents accidental or intentional data breaches. The system actively enforces compliance at every point.

[0099] This high-level example shows the system's ability to manage a highly complex ecosystem of data sources, collaborations, permissions, commercial fees, and identifier types within a unified data delivery group framework. This centralized approach significantly reduces operational overhead and complexity while ensuring data privacy, compliance, and enhanced campaign performance.

[0100] As illustrated in Fig. 4 showing data delivery groups Ul 50b, the user next selects the publishers or platforms to receive the audience segments. This is accomplished through selectors 56, which present a comprehensive list of all configured and available destination connections. To guide the user's choice, the system provides useful details about each publisher, such as the type of data they accept at data type 58, indicating whether they handle first-party data, third-party data, or both, and a visual representation of the various identifiers they support at targeted identifier types 60, including MAIDs, Cookies, proprietary third-party identifiers, and hashed PI I. Each publisher has unique policies and restrictions regarding data acceptance and identifiers; the system's ability to manage diverse identifiers is therefore paramount. Inaccurate mapping can lead to wasted impressions and missed opportunities.

[0101] Furthermore, targeted identifier types 60 offers an interactive element, allowing the user to toggle on or off the specific identifier types sent to each publisher. Granular control ensures that only the most appropriate identifiers are used, respecting campaign requirements. The system also checks for publisher-specific restrictions or commercial conditions, such as fees for using specific segments. The system maintains an up-to-date registry of these requirements, ensuring the user has the latest information.

[0102] Once audience segments and publishers are selected, the final step involves configuring additional parameters that apply to the entire data delivery group, as shown in data delivery group Ul 50c of Fig. 5. The user begins by giving the data delivery group a descriptive name, which aids in tracking and organization. Then, the user set the delivery or data distribution cadence using delivery cadence 62. This can be configured for a one-time delivery, or set to recurring intervals (daily, weekly, monthly, or custom). The delivery cadence directly impacts how quickly campaign performance data feeds into the system, enabling rapid optimization. The user can also specify a date to "turn off" the data delivery group at end date 64, which is especially important for time-sensitive campaigns.Attorney Docket No. RAMP-00327-WO

[0103] Finally, the user defines the data delivery precision level 66, choosing between individual-level targeting (with precise identifiers) or household-level targeting (based on aggregated identifiers). Throughout this process, the system validates the entire configuration against data governance policies, privacy regulations, and publisher-specific requirements.

[0104] Before activating the data delivery group, the system performs comprehensive validation checks to ensure identifier compatibility (that the audience segments contain identifiers accepted by the selected publishers), permission compliance (that the user have the necessary permissions to distribute the segments), and data governance adherence (that the distribution complies with all relevant policies and regulations). The system must intelligently translate and map identifiers, which prevents unauthorized data sharing.

[0105] The system logs every data delivery group creation and modification for compliance.The data delivery group can be activated upon successful validation, initiating the data distribution process. The system provides real-time status updates and detailed logs to monitor the delivery. Performance data collected from each data delivery group's activation feeds into the system, informing future creations and optimizations.

[0106] The data delivery group status Ul component 50d, as shown in Fig. 6, is a useful control center providing real-time, granular visibility into the activation and performance of each data delivery group. This component is not merely a monitoring tool but a sophisticated diagnostic and management interface designed to empower users to actively oversee and optimize their cross-publisher campaigns, particularly within the complex environments of RMNs and CMNs.

[0107] The core of this component is a dynamic dashboard that delivers a comprehensive, up- to-the-second view of data delivery status across all connected publishers and platforms. The data delivery group name 68, as set in data delivery group Ul 50c for each data delivery group, provides a list of active data delivery groups. Clear, visual indicators 72 are provided for each publisher, showing whether data has been successfully transmitted (e.g., green checkmark), is pending (e.g., yellow clock icon), or has encountered errors (e.g., red exclamation point). These indicators are color-coded for quick recognition and understanding.

[0108] For ongoing deliveries, progress bars 70 provide a visual representation of the transmission status, indicating the percentage of data sent and the estimated time remaining. This is particularly crucial for large audience segments or complex identifier mappings, as well as to many publisher / walled garden mappings.Attorney Docket No. RAMP-00327-WO

[0109] Comprehensive, searchable logs are available for each delivery attempt. These logs record every step of the process, from data extraction and transformation to transmission and confirmation. They include timestamps, data volumes, identifier types, and any error messages encountered. This level of detail is useful for troubleshooting and auditing purposes. In case of errors, the logs provide clear, specific error codes and descriptions, categorized for efficient troubleshooting and integrated with an intelligent alerting system. A number of types of errors may be indicated. One type is authentication failures (OAuth publisher / platform / walled garden errors, API key issues) at the publisher or platform, including detailed authentication logs with token exchange details and error responses from the authorization server. Another type is permission restrictions or access control violations, with specific details on the violated policies (e.g., ABAC / RBAC rules) and the affected audience segments. Another type is activation fee support required, including details on the pending fees, approval workflow status, and integration with financial systems. Another type is delivery delays due to publisher data rate limiting or API throttling, with real-time rate limit status at the receiving platform / publisher / walled garden, retry attempts with exponential back-off, and integration with circuit breaker patterns to prevent cascading failures. Another type is temporary network outages or connectivity issues, with network latency metrics (RTT, jitter), error codes from the transport layer (TCP / IP), and integration with network monitoring tools. Another type is data format incompatibility or schema mismatches, including schema validation errors (using JSON Schema or similar), data transformation logs with input / output samples, and integration with schema registries. Another type is low match rates for provided identifiers, with detailed match rate statistics for each identifier type (MAIDs, Cookies, third-party proprietary IDs, hashed PH), diagnostics for identifier mapping algorithms, and integration with identity resolution services. Another type is audience segment size below minimum thresholds required by specific publishers, with real-time segment size calculations, validation results against publisher policies, and integration with audience estimation tools. The alerting system can trigger notifications via email, SMS, or in-platform alerts, and can be configured based on severity levels, error types, and user roles. This proactive alerting is a novel feature that enables rapid response to issues.

[0110] Users can drill down into specific data points to access more detailed information. One type of available information is performance of individual audience segments on each publisher platform, including impression counts, click-through rates, conversion rates, viewability metrics, and attribution models. Another type of information is specific identifiers used for each user,Attorney Docket No. RAMP-00327-WOalong with their match status, associated attributes (e.g., demographic data), and identifier lifecycle information (e.g., creation timestamp, last seen timestamp). Another type of information is exact error messages generated during transmission, with links to relevant documentation, troubleshooting guides, and knowledge base articles. Another type of information is contextual data, such as the user's role and permissions, the specific data governance policies in effect during the delivery attempt, the data lineage (source of data, transformation steps), and the API request / response payloads. Also available are real-time diagnostics tools, such as network trace routes, API call latency graphs, and data transformation debugging interfaces, which are useful for pinpointing issues in complex, distributed systems.

[0111] Referring more specifically to data delivery group Ul 50d, the Ul provides a high-level overview of all data delivery groups, displaying key information such as data delivery group name 68 and ID, the number of audience segments included, the number of destination connections (publishers / platforms); the data delivery cadence 62 (one-time, daily, weekly, etc.); and creation and modification timestamps.

[0112] Progress bars 70 provide a column indicating the overall "health" of each data delivery group, including current process step (taxonomy creation, ID matching, data delivery, etc.); overall status (active, paused, completed, error); and progress indicators for each process step.

[0113] Visual indicators 72 provide a summary of any errors or warnings associated with each data delivery group, providing a quick overview of potential issues without requiring users to drill down into details. Examples include: no delivery due to authorization failure; no delivery due to permission restriction; no delivery due to activation fee support required; delivery delays, publisher data rate limited; delivery delays, temporary network outage; no delivery due to incompatible data format; low match rate for provided identifiers; and audience segment below minimum size threshold. These errors and warnings are just a first step on how the system can help user or Al agents understand, correct and debug how their data delivery groups are running across multiple publishers and across multiple identifiers.

[0114] Each data delivery group has another set of very detailed status data that are specific to the publisher, the audience segments and the identifier. Based on the use case it may be favorable to review and read the data delivery group activity from the lens of the audience segments that are part of a given data delivery group or a different way to look at a given data delivery group through the lens of the publisher or platforms. The following Ul components will traverse these two different lenses.Attorney Docket No. RAMP-00327-WO

[0115] Data delivery group status Ul 74a, as shown in Fig. 7, provides a highly detailed, publisher or destination-centric view of the data delivery group status. This perspective is useful for understanding and managing the complexities of cross-publisher data activation. By focusing on each individual destination connection, whether a publisher, platform, or even an RMN or CMN owner's specific seat, users gain the granular visibility needed to effectively monitor, troubleshoot, and optimize delivery issues that may be unique to each platform. This capability is accessed by toggling the destination connection view 76 within the data delivery group status interface, shifting the focus from a general data delivery group overview to a publisher-specific breakdown.

[0116] Upon selecting the destination connection view 76, the detailed status for a given data delivery group is meticulously organized by publisher or platform at publisher identity 78. This organization is not merely a list but a comprehensive dashboard, providing insights into the intricate processes of taxonomy building, identifier matching, data delivery, and more. Each publisher's entry further breaks down into detailed status reports for each identifier type 80 that the platform supports. This is useful because publishers have diverse identifier acceptance criteria (MAIDs, cookies, third-party proprietary identifiers such as RampID®, hashed PI I, etc.), and successful delivery hinges on accurately mapping and transmitting the appropriate identifiers.

[0117] The detailed status for each publisher and identifier combination includes a wealth of information, meticulously designed to provide actionable insights and facilitate quick problem resolution. This information includes connection status, which is an overarching indicator signaling the real-time connectivity and operational state of the connection (e.g., 'Connected,' 'Disconnected,' 'Error'). This reflects the system's active monitoring of API endpoints and communication channels with each destination.

[0118] This information also includes last successful delivery, which is a precise timestamp marking the last successful data delivery to this specific destination. This helps users track delivery cadence and identify potential delays or interruptions.

[0119] This information also includes last attempted delivery, which is a timestamp of the most recent delivery attempt, providing context for troubleshooting failed deliveries or pending statuses.Attorney Docket No. RAMP-00327-WO

[0120] This information also includes delivery status, which is a granular status indicator detailing the outcome of the last delivery attempt (e.g., 'Success,' 'Pending,' 'Failed'). This status is often accompanied by real-time progress updates, especially for large data transfers.

[0121] This information also includes error messages, if any; in case of delivery failures, the system provides specific error codes and detailed messages. These messages are not just generic alerts but are designed to provide actionable insights into the root cause of the failure. They might include authentication errors (authorization failures, API key issues), permission restrictions, data format incompatibilities, rate limiting issues, or network connectivity problems. This aligns with the system's focus on robust error handling and diagnostics, as detailed in the broader document.

[0122] This information also includes data volume delivered, which is the precise amount of data (e.g., number of records, file size in bytes) successfully delivered. This metric is useful for auditing and ensuring that the intended volume of data reached the destination.

[0123] This information also includes the identifier types sent, which is a comprehensive list of the specific identifier types (MAIDs, Cookies, third-party proprietary identifiers, hashed PH, customer / platform identifiers (CID), etc.) transmitted to this destination. This list is dynamically generated based on the audience segments selected and the publisher's accepted identifiers, reflecting the system's intelligent identifier management capabilities.

[0124] This information also includes the identifier match rate for each identifier type, indicating the percentage of identifiers successfully matched by the destination platform. This is a useful metric for assessing data quality and the effectiveness of identifier mapping. Low match rates can indicate data quality issues or the need for refining identifier translation rules.

[0125] This information also includes data governance policies applied, which is a summary of the specific data governance policies and restrictions applied to the data delivered to this destination. This includes details on data usage limitations, privacy regulations, and any commercial conditions (fees, usage limits) imposed by the publisher , walled garden, or RMN / CMN owner.

[0126] This level of detailed, publisher-specific status reporting is useful for several reasons.First, it acknowledges the heterogeneity of the programmatic ecosystem, where each publisher or platform has unique requirements and constraints. Second, it empowers users to proactively identify and address delivery issues, minimizing downtime and ensuring campaign effectiveness. Third, it provides a clear audit trail of data activation activities, facilitating transparency andAttorney Docket No. RAMP-00327-WOaccountability. Finally, it enables iterative optimization, as users can leverage the granular status information to refine their targeting strategies, identifier mapping rules, and data governance policies for each destination. This aligns with the invention's overall vision of a closed-loop system for continuous campaign improvement.

[0127] Data delivery group status Ul 74b, shown in Fig. 8, provides a useful alternative view of the data delivery group status from Ul 74a, shifting the focus from publisher or destination connections to individual audience segments within the data delivery group. This segmentcentric perspective is useful for understanding how specific audience segments are performing across different publishers and for identifying issues related to particular segments. By switching to the activated segments view 82 within the status interface, users can analyze the activation and performance of each segment in detail.

[0128] Upon selecting the activated segments view 82, the detailed status for a given data delivery group is meticulously organized by audience segment, again sorted by identifier types 80. This organization is not merely a list but a comprehensive dashboard, providing insights into the intricate processes of taxonomy building, identifier matching, data delivery, and more. Each segment's entry further breaks down into detailed status reports for each identifier type that the publisher and programmatic platform support for this given data delivery group.

[0129] The detailed status for each segment and identifier combination includes a wealth of information, meticulously designed to provide actionable insights and facilitate quick problem resolution One type of information is connection status, an overarching indicator signaling the real-time connectivity and operational state of the connection (e.g., 'Connected,' 'Disconnected,' 'Error'). This reflects the system's active monitoring of API endpoints and communication channels with each destination platform / walled garden / RMN / CMN.

[0130] Another type of information is last successful delivery, which is a timestamp marking the last successful data delivery to this specific destination. This helps users track delivery cadence and identify potential delays or interruptions.

[0131] Another type of information is last attempted delivery, which is a timestamp of the most recent delivery attempt, providing context for troubleshooting failed deliveries or pending statuses.

[0132] Another type of information is delivery status, a granular status indicator detailing the outcome of the last delivery attempt (e.g., "In process", "completed", "Failed"). This status is often accompanied by real-time progress updates, especially for large data transfers.Attorney Docket No. RAMP-00327-WO

[0133] Another type of information is error messages (if any). In case of delivery failures, the system provides specific error codes and detailed messages. These messages are not just generic alerts but are designed to provide actionable insights into the root cause of the failure. They might include authentication errors (authorization failures, API key issues), permission restrictions, data format incompatibilities, rate limiting issues, or network connectivity problems. This aligns with the system's focus on robust error handling and diagnostics

[0134] Another type of information is data volume delivered, which provides the precise amount of data (e.g., number of records, file size in bytes) successfully delivered. This metric is vital for auditing and ensuring that the intended volume of data reached the destination.

[0135] Another type of information is identifier types set, a comprehensive list of the specific identifier types transmitted to this destination. This list is dynamically generated based on the audience segments selected and the publisher's accepted identifiers, reflecting the system's intelligent identifier management capabilities.

[0136] Another type of information is identifier match rate for each identifier type, indicating the percentage of identifiers successfully matched by the destination platform. This is a useful metric for assessing data quality and the effectiveness of identifier mapping. Low match rates can indicate data quality issues or the need for refining identifier translation rules.

[0137] Another type of information is data governance policies applied, which is a summary of the specific data governance policies and restrictions applied to the data delivered to this destination. This includes details on data usage limitations, privacy regulations, and any commercial conditions (fees, usage limits) imposed by the publisher, walled garden, RMN owner, or CMN owner. This aligns with the emphasis on secure and compliant data sharing within a multi-party environment.

[0138] This segment-centric view as shown in Fig. 8 is useful for several reasons. First, it enhances the understanding of segment performance. It allows users to understand how specific audience segments are performing across different identifiers. For example, a brand owner might discover that a high-value customer segment performs exceptionally well on a premium publisher running on proprietary third-party IDs, but poorly on a social media platform that is running on online cookies. This insight allows for strategic budget allocation and creative optimization. In a particular use case, a retail brand owner may observe that its loyalty program members segment has a 10% conversion rate on proprietary third-party IDs but only a 2%Attorney Docket No. RAMP-00327-WOconversion rate on cookies. The brand owner can then investigate the cause, potentially adjusting creative or bidding strategies for publishers that support those proprietary IDs.

[0139] Another reason that the segment-centric view is useful is for troubleshooting segmentspecific issues. The view helps pinpoint issues related to particular segments. For instance, a segment with a high percentage of outdated email addresses might experience low match rates with publishers relying on hashed PH.

[0140] Another reason that the segment-centric view is useful is for optimizing targeting strategies. It enables users to refine their targeting strategies by analyzing the performance of different segments. A brand might find that a retargeting segment performs best with a specific creative message, while a prospecting segment requires a different approach. In a particular use case, an e-commerce brand owner may discover that its abandoned cart segment responds best to a discount offer, while its new visitors segment responds better to product information and testimonials. The brand owner can then tailor its campaigns accordingly.

[0141] Another reason that the segment-centric view is useful is for ensuring data governance and compliance. This view provides visibility into the data sources, segmentation rules, and usage restrictions for each segment. This is useful for ensuring compliance with data governance policies and privacy regulations, especially when dealing with sensitive data or segments derived from collaborative data sources. In a particular use case, a healthcare provider needs to ensure that its patients with condition X segment is only delivered to publishers with HIPAA-compliant data handling practices. The segment status view allows the healthcare provider to verify and audit these permissions.

[0142] By providing this detailed view of the data delivery group status from the audience segment perspective, the system empowers users to effectively manage and optimize their campaigns with a fine-grained understanding of how each segment is performing. This aligns with the emphasis on unified audience management, cross-publisher measurement, and iterative optimization.

[0143] Referring now to data delivery group audit log Ul 84 of Fig. 9, the system maintains rich logs and activity history 86 to provide a complete audit trail 88 and facilitate debugging for each data delivery group. This includes detailed event logs, which provide a comprehensive record of every event related to each data delivery group, including creation, modification, activation, data delivery, and error handling. This also includes user activity logs, which provide a record of user actions related to data delivery groups, including who made changes, when, and whatAttorney Docket No. RAMP-00327-WOchanges were made. This also includes system activity logs, which provide a record of systemgenerated events, such as automated data delivery attempts, error detection, and recovery actions. Importantly, Al agent activities are also logged on this page, providing transparency to brand owners about the level of automation and decisioning added by the provider's Al assistant.

[0144] This detailed status and logging system ensures transparency, accountability, and efficient troubleshooting, enabling users to effectively manage and optimize their crosspublisher campaigns, particularly within the dynamic and complex environments of RMNs and CMNs.

[0145] In addition to data distribution capabilities, the system according to certain embodiments of the invention provides a robust analytics and measurement framework. The data delivery group insights and measurement Ul 90, shown in Fig. 10, serves as a comprehensive analytics and optimization engine, providing deep, actionable insights into campaign performance across multiple publishers and diverse identifier ecosystems. This component leverages the aggregated, deduplicated data from the data collaboration crosspublisher clean room 18 to offer a holistic view of campaign effectiveness, empowering users to make data-driven decisions, fine-tune their strategies, and maximize ROL

[0146] The unified performance dashboard is accessible via the measurement tab 92 within the data delivery group detail asset section. This dashboard acts as a single source of truth, presenting both aggregated cross-publisher metrics and granular publisher-specific performance data.

[0147] Aggregated, deduplicated metrics 94 at the top section of the dashboard provides a high-level overview of campaign performance across all participating publishers. Key metrics such as total exposure, total audience (segmented by demographics like age), and total conversions are presented. These metrics are deduplicated, ensuring an accurate and unbiased representation of campaign reach and impact across all the publishers.

[0148] Recognizing the diverse analytical needs of users, the dashboard offers a highly configurable environment. Users can add additional visualization widgets to track a wide range of performance metrics. One such metric is lift and attribution analysis, which is useful for quantifying the incremental impact of the campaign, attributing conversions to specific touchpoints and channels. Another such metric is deduplicated reach and frequency metrics, providing an accurate count of unique users reached and the frequency of ad exposures,Attorney Docket No. RAMP-00327-WOeliminating double-counting across publishers. Another such metric is projected scaled reach frequency estimates, which forecasts the potential reach and frequency of the campaign at a larger scale, aiding in budget allocation and strategic planning. Another such metric is incremental reach calculations, which measure the additional reach achieved through specific campaign optimizations or targeting strategies. Another such metric is return on ad spend (ROAS) calculations, which assess the overall financial effectiveness of the campaign, measuring revenue generated per dollar spent on messaging.

[0149] Publisher-specific performance data 96, in addition to aggregated metrics, presents granular performance data specific to each publisher or walled garden. This allows users to compare performance across different platforms, identify high-performing channels, and understand the nuances of audience engagement on each site. For example, users can view exposure data specific to several publishers, providing valuable insights into platform-specific performance variations.

[0150] To facilitate data interpretation and trend analysis, data delivery group insights and measurement Ul 90 employs a variety of interactive charts, graphs, and data visualization techniques. These visual aids allow users to quickly grasp complex data sets, identify patterns, and make informed decisions. One such component is customizable dashboards, which allows users to create custom dashboards tailored to their specific analytical needs, selecting the metrics and visualizations that are most relevant to their campaign objectives. Another such component is data filtering and segmentation, allowing users to filter and segment data based on various criteria, such as publisher, audience segment, time period, or identifier type. This enables granular analysis and the identification of specific performance drivers. Another such component is data export, whereby users can export reports and data visualizations in various formats (e.g., CSV, PDF, PNG) for further analysis or sharing with stakeholders.

[0151] Leveraging the power of AI / ML algorithms, this component goes beyond basic reporting to provide Al-driven recommendations and optimization insights. This includes anomaly Detection, whereby the system can detect anomalies in campaign performance, such as sudden drops in conversions or unexpected spikes in frequency, alerting users to potential issues. Another feature is audience optimization suggestions, whereby, based on the observed data, the system can suggest optimizations to audience targeting, creative messaging, or bidding strategies. Another feature is predictive analytics, whereby the system can provide predictiveAttorney Docket No. RAMP-00327-WOanalytics, forecasting future campaign performance and identifying potential opportunities or risks.

[0152] This component provides for real-time data updates. Performance data is continuously collected and updated in real-time, providing users with up-to-date insights into campaign effectiveness. This component also provides for closed-loop feedback. The insights derived from the component directly inform subsequent campaign iterations, allowing users to make data-driven adjustments and improve performance.

[0153] The data activation permissions Ul 98a component, shown in Fig. 11 acts as the orchestrator of trust and control in certain embodiments of the invention. This component empowers brands, publishers, RMN / CMN owners, and data sellers to maintain meticulous control over how their data is used and activated, ensuring transparency and accountability at every step. In particular, this component allows users to manage highly granular access permissions for each audience segment within a data collaboration, whether between a brand owner, publisher, CMN owner, or data seller. These permissions dictate precisely which publishers can receive and utilize specific audience segments. The Ul provides a comprehensive, easily navigable overview of existing permissions, making it simple to audit, modify, or revoke access controls as needed. This level of detail is useful in a multi-party environment where data sensitivity and compliance are paramount, as seen in the diverse use cases outlined throughout the document.

[0154] Beyond simply granting or denying access, users can set sophisticated time-based restrictions and data usage limitations. For example, a CMN owner might grant a brand access to a collaborative audience for a specific campaign period, after which the access automatically expires. Similarly, data usage limitations can be imposed, such as restricting the number of times an audience can be used or the specific purposes for which it can be employed. These restrictions ensure that data is used only within the agreed-upon terms, mitigating risks and fostering trust among collaborating parties. This is particularly relevant in scenarios where temporary access to specific audience segments might be granted for promotional campaigns.

[0155] Maintaining a comprehensive audit trail is crucial for compliance and accountability. For that reason, this component meticulously logs all permission changes, providing a detailed record of who granted which permissions to whom, when, and under what conditions. This audit trail is accessible to authorized, administrator users, allowing them to review past actionsAttorney Docket No. RAMP-00327-WOand identify any anomalies. This is useful for addressing potential disputes and demonstrating compliance with data privacy regulations.

[0156] The system implements role-based access control (RBAC) to ensure that only authorized personnel can manage data activation permissions. Different levels of users (e.g., campaign managers, data analysts, and legal teams) are assigned specific roles with corresponding permissions. This prevents unauthorized users from modifying critical permissions and ensures data access aligns with organizational policies. For instance, a campaign manager might be able to grant access to specific publishers, while a legal team member might be required to approve permissions involving sensitive Pll data.

[0157] Referring now more particularly to Fig. 11, data activation permissions Ul 98a features a comprehensive wizard that guides users through setting activation rules. At activation rules step 100, a CMN owner is given control over where a collaborative audience can be activated. The CMN owner can choose to enable activation to any destination (i.e., any publisher or platform) at allow activation box 102. The CMN owner can also choose specific destinations (i.e., a specific set of publishers) at specific destinations box 104, and can choose from the destinations list 106.

[0158] Alternatively, as shown in data activation permissions Ul 98b component of Fig. 12, the CMN owner can choose its own destination connection at my destination button 108, and then drill down to a particular destination connection at my destination connection 110. This option is particularly powerful, allowing the CMN owner to mandate that the brand owner uses a destination connection configured and controlled by the CMN owner. This is valuable in scenarios where the CMN owner wants to coordinate the campaign on behalf of the brand owner, as will be described in a use case below.

[0159] Given the complexity of managing multiple data sources, collaborations, permissions, commercial fees, and identifier types, the Uls of certain embodiments of the invention provide clear visibility and tracking. Users can easily see which audiences are being shared with which publishers, under what conditions, and with which identifiers. This transparency is crucial for managing complex campaigns and ensuring everything operates as intended. The Uls also provide status updates on data activation, showing whether data has been successfully delivered to each publisher and alerting users to any issues.

[0160] The data activation permissions component Ul 98a, 98b provides granular control, robust security, and comprehensive tracking, ensuring that data is used responsibly,Attorney Docket No. RAMP-00327-WOcompliantly, and effectively. Integrating with other components and providing detailed logging and RBAC empowers users to manage complex data collaborations confidently.

[0161] There are a number of use cases that may be understood from the foregoing description of certain embodiments of the present invention. These include extensions of existing use cases for programmatic networks. One example is the expansion of retail media networks (RMNs). Existing RMN scenarios involve, for example, a retailer A allowing brand owners B, C, and D to deliver messages on their website or app. Only limited data sharing is possible in this scenario. Using embodiments of the present invention, the brand owners B, C, D are able to bring their own first-party data into the collaborative clean room. Retailer A shares anonymized purchase data, unified audience segments are created (e.g., "customers who bought product X and are also interested in category Y"). Campaigns are launched across Partner A's properties and Partner C (which could be, for example, a major online publication) and Partner D (a social media platform). Cross-publisher measurement reveals that the campaign performed best on Partner D, so Partner B reallocates budget to Partner D and creates a new, more specific segment of "Customers who bought product X, interested in Y, and live in Z region" within the clean room and continues the retargeting.

[0162] A related use case is the expansion of commerce media networks (CMNs) Current CMN scenarios involve, for example, an e-commerce platform (Partner A) that lets suppliers (Partners B, C, D) deliver messages concerning their products on the platform. Using embodiments of the invention, partners B, C, and D may bring customer data from their own systems into the clean room. Partner A shares data on browsing behavior and purchase history. Audiences are created (e.g., "high-value customers who frequently purchase items from multiple suppliers").Campaigns are launched on Partner A's site and Partner E (a related review site), Partner F (an influencer network), and Partner G (a direct mail vendor) for multi-channel execution. The system measures which channels and publishers perform the best. Partner C then optimizes creative messaging for best-performing segments and channels, and the cycle repeats.

[0163] A new use case using embodiments of the present invention is, when a brand owner is launching a new product, it can identify the most engaged existing customers within the clean room. The brand owner can then create campaigns to reach those customers and similar profiles on other publisher platforms to drive trial and awareness. Measurement allows the brand owner to understand the effectiveness of the campaign in driving initial sales and longterm customer engagement across different channels. The data collected in this model mayAttorney Docket No. RAMP-00327-WOinclude tables of customer IDs, engagement scores, campaign IDs, publisher IDs, impression counts, and product purchase events.

[0164] Another new use case is that non-competing brands can collaborate within the clean room to reach shared audiences. For example, a beverage brand and a snack brand partner to target people interested in "outdoor activities." This allows for cost-sharing and greater reach than either brand could achieve alone, with measurement of shared and individual campaign success. Data collected in this model may include a table with campaign IDs, brand IDs, shared audience segment IDs, and individual campaign performance metrics.

[0165] Another new use case is that the system according to certain embodiments enables precise control and management of holdout / control groups across all publishers. This may be useful for accurate measurement of campaign effectiveness. For example, a 10% holdout can be consistently maintained across five publishers, ensuring accurate A / B testing. In this use case, data collected according to the model may include a table with holdout segment flags, user IDs, publisher IDs, and campaign IDs.

[0166] Another new use case is cross-publisher frequency capping. Frequency capping is intended to prevent over-exposure of messages to the same individuals across multiple publishers, since effectiveness is known to decline past a certain point of exposure. The system in certain embodiments tracks impression frequency within the clean room's user level and enforces frequency caps, providing the brand owner with control and oversight. In this example, data collected according to the model may include a table with user IDs, campaign IDs, publisher IDs, and timestamps of impressions.

[0167] Another new use case is attribution modeling with enhanced accuracy. More sophisticated attribution models (e.g., multi-touch attribution) can be developed and applied with unified cross-publisher data in certain embodiments. This provides a more accurate view of which touchpoints are driving conversions.

[0168] Another new use case is artificial intelligence / machine learning (AI / ML)-Powered Audience Discovery. AI / ML algorithms within the clean room can analyze the combined data to uncover new high-value audience segments that may not have been apparent before. For example, suppose that an Al model discovers a segment of consumers who read cooking blogs, buy a specific type of cooking appliance from one retailer, and buy their groceries at an entirely different retailer to find insights and activate new campaigns. In this example, data collectedAttorney Docket No. RAMP-00327-WOaccording to the model may include a model output table with discovered audience segment definitions, associated attributes, and predicted value.

[0169] It will be understood from the foregoing description of certain embodiments of the invention that those embodiments lead to significant computing environment efficiencies and improvements. More specifically, the multi-party data activation system significantly reduces compute impact through several key mechanisms. First, it implements centralized processing and orchestration. Instead of each publisher or retailer needing to run separate campaign management and optimization processes, the system consolidates key computational tasks within the secure clean room environment. This drastically reduces redundant computations. For instance, calculating campaign performance metrics, audience overlap, or optimization algorithms must only run once for all participating parties. This centralized approach eliminates the need for each publisher to maintain its matching functions or other compute-intensive processes.

[0170] Second, the system optimizes data flows and reduces data movement. The need to transfer large datasets between different systems and partners is minimized by processing data within the clean room. Data transfer and identifier matching / projection are very computeintensive tasks, and reducing them lowers network traffic and the overall processing power required.

[0171] Third, the system enables efficient algorithm execution. Having all the relevant data in one place within the clean room allows for developing and deploying more efficient algorithms. These algorithms can be optimized for batch processing, parallel processing, and other techniques that leverage the combined datasets, rather than processing data in small, isolated chunks.

[0172] Fourth, matching or segmentation algorithms are computed centrally and only once instead of each publisher handling them separately.

[0173] Finally, the system reduces application programming interface (API) calls and integrations. Instead of numerous point-to-point integrations between each party sending messages and each publisher, and then more for measurement, there is a streamlined and standardized set of interactions with the core platform. This significantly reduces the overhead associated with making and managing countless API calls.

[0174] Regarding storage, the system reduces data storage impact through consolidated data storage. While the clean room stores data, it's far more efficient to have a consolidated andAttorney Docket No. RAMP-00327-WOoptimized data storage system compared to each publisher and messaging party maintaining redundant copies of the same or similar data. Data deduplication and compression techniques can be implemented centrally, saving considerable storage space. With a centralized system handling the heavy lifting and matching, there is no need for each publisher or brand owner to keep large amounts of data for such purposes.

[0175] Optimized data structures and indexing strategies can be designed and implemented, tailored to the specific needs of cross-publisher campaign management, in these embodiments of the present invention. This improves data retrieval efficiency and reduces the need to store redundant or unnecessary data. For example, such systems may use columnar databases for analytical processing and data cubes to reduce the compute load of specific calculations.Implementing effective data lifecycle management policies (data retention, archiving, and deletion) becomes much simpler with a centralized system. Older data no longer needed can be automatically removed or archived, freeing up storage space, which is difficult to achieve with distributed systems. Data is often replicated across multiple publisher systems, brand owner systems, and various third-party vendors in a fragmented ecosystem. This approach minimizes data replication by keeping the core data within the secure clean room environment. Data virtualization and federated queries can also be used to reduce the need for data replication further.

[0176] Overall, it may be seen that a multi-party data activation system according to certain embodiments of the present invention significantly reduces compute and storage impact by centralizing processing and avoiding redundant calculations; optimizing data flows and minimizing data movement; consolidating data storage and implementing data lifecycle management; and reducing API calls and system integrations. By considering these architectural considerations, one may build a highly efficient and scalable platform that minimizes resource consumption and maximizes performance.

[0177] Al Agents, retrieval-augmented generation (RAG), and large language models (LLMs) open up new possibilities for use cases with multiple partners and data activation. The system's strength lies in managing diverse identifiers (such as proprietary third-party IDs, MAIDs, offline data, email addresses, telephone numbers, browser cookies) across multiple publishers through a single, centralized Ul, a robust foundation for advanced Al applications.Attorney Docket No. RAMP-00327-WO

[0178] In an example, one may consider an Al agent tasked with optimizing a campaign's reach and conversion. This agent can leverage the system's unified identifier management to analyze user journeys across various touchpoints. For instance, it might identify that a user initially engaged via email (offline identifier), then interacted on a mobile app (MAID), and finally converted through a website (cookies). With this comprehensive view, the Al agent can dynamically adjust bidding strategies and creative delivery based on each user's most effective identifier and platform, ensuring optimal engagement. This is made possible by the system's ability to activate and manage these different identifiers across multiple publishers seamlessly.

[0179] Furthermore, collaborative RAG can enhance the Al agent's decision-making. The RAG system provides the Al agent with enriched context by accessing a shared knowledge base that includes brand guidelines, historical campaign data, and real-time performance metrics across all publishers. For example, when the Al agent detects a drop in performance on a specific publisher, RAG can retrieve past successful campaigns on that platform, along with the audience segments and creative assets that resonated best. This enables the Al agent to quickly adapt and optimize the campaign, drawing from a collective intelligence pool. The centralized Ul is the access point to this knowledge base, providing a comprehensive view of all relevant data.

[0180] Collaborative LLMs can take this a step further. The system creates a dynamic repository of insights and best practices, allowing brands and publishers to contribute to a shared LLM. This LLM, powered by the multi-party data within the clean room, can generate highly personalized creative content tailored to specific audience segments and identifiers. For example, it can create unique message variants for consumers identified by MAIDs on a mobile app versus consumers identified by cookies on a desktop website. The LLM can also provide strategic recommendations for campaign optimization, suggesting new audience segments or touchpoint combinations based on its unified data analysis. The "single pane of glass" Ul provides access to these LLM-generated insights, making them actionable for campaign managers.

[0181] In summary, with its ability to manage diverse identifiers across multiple publishers and provide a centralized Ul, the invention in certain embodiments creates an ideal environment for Al agents, RAG, and collaborative LLMs. These Al technologies can fully leverage the system's core innovations to drive unparalleled efficiency, personalization, and effectiveness in crosspublisher data activation. Combining unified data management and advanced Al capabilities ensures that campaigns are optimized at every touchpoint, leading to superior results and ROI.Attorney Docket No. RAMP-00327-WO

[0182] The methods described herein may in various embodiments be implemented by any combination of hardware and software. For example, in one embodiment, the methods may be implemented by a computer system or a collection of computer systems, each of which includes one or more hardware processors executing program instructions stored on a computer- readable physical storage medium coupled to the hardware processors. The program instructions may implement the functionality described herein (e.g., the functionality of various hardware servers and other components that implement the network-based cloud and noncloud computing resources described herein). The various methods as illustrated in the figures and described herein represent example implementations. The order of any method may be changed, and various elements may be added, modified, or omitted.

[0183] Fig. 13 is a block diagram illustrating an example computer hardware system, according to various embodiments. Computer system 140 may implement a hardware portion of a cloud computing system as forming parts of the various implementations of the present invention. Computer system 140 may be any of various types of hardware devices, including, but not limited to, a commodity server, personal computer system, desktop computer, laptop or notebook computer, mainframe computer system, handheld computer, workstation, network computer, a consumer device, application server, physical storage device, telephone, mobile telephone, or in general any type of computing node, compute node, compute device, and / or hardware computing device.

[0184] Computer system 140 includes one or more hardware processors 140a, 141b... Mln (any of which may include multiple processing cores, which may be single or multi-threaded) coupled to a physical system memory 142 via an input / output (I / O) interface 144. Computer system 140 further may include a network interface 146 coupled to I / O interface 144. In various embodiments, computer system 140 may be a single processor system including one hardware processor 141a, or a multiprocessor system including multiple hardware processors 141a, 141b... Mln. Processors 141a, etc. may be any suitable processors capable of executing computing instructions. For example, in various embodiments, processors 141a, etc. may be general-purpose or embedded processors implementing any of a variety of instruction set architectures.

[0185] In multiprocessor systems, each of processors 141a, etc. may commonly, but not necessarily, implement the same instruction set. The computer system 140 also includes one or more hardware network communication devices (e.g., network interface 146) forAttorney Docket No. RAMP-00327-WOcommunicating with other systems and / or components over a communications network, such as a local area network, wide area network, or the Internet. For example, a client application executing on system 140 may use network interface 146 to communicate with a server application executing on a single hardware server or on a cluster of hardware servers that implement one or more of the components of the systems described herein in a cloud computing environment as implemented in various sub-systems. In another example, an instance of a server application executing on computer system 140 may use network interface 146 to communicate with other instances of an application that may be implemented on other computer systems.

[0186] In the illustrated embodiment, computer system 140 also includes one or more physical persistent storage devices 148 and / or one or more I / O devices 150. In various embodiments, persistent storage devices 148 may correspond to disk drives, tape drives, solid-state memory or drives, other mass storage devices, or any other persistent storage devices. Computer system 140 (or a distributed application or operating system operating thereon) may store instructions and / or data in persistent storage devices 148, as desired, and may retrieve the stored instructions and / or data as needed. For example, in some embodiments, computer system 140 may implement one or more nodes of a control plane or control system, and persistent storage 148 may include the solid-state drives (SSDs) attached to that server node. Multiple computer systems 140 may share the same persistent storage devices 148 or may share a pool of persistent storage devices, with the devices in the pool representing the same or different storage technologies, including such technologies as described above.

[0187] Computer system 140 includes one or more physical system memories 142 that may store code / instructions 143 and data 145 accessible by processor(s) 141a, etc. The system memories 142 may include multiple levels of memory and memory caches in a system designed to swap information in memories based on access speed, for example. The interleaving and swapping may extend to persistent storage devices 148 in a virtual memory implementation, where memory space is mapped onto the persistent storage devices 148. The technologies used to implement the system memories 142 may include, by way of example, static randomaccess memory (RAM), dynamic RAM, read-only memory (ROM), non-volatile memory, solid- state memory, or flash-type memory. As with persistent storage devices 148, multiple computer systems 140 may share the same system memory systems 142 or may share a pool of system memories 142. System memory or memory systems 142 may contain programAttorney Docket No. RAMP-00327-WOinstructions 143 that are executable by processor(s) 141a, etc. to implement the routines described herein.

[0188] In various embodiments, program instructions 143 may be encoded in binary, Assembly language, any interpreted language such as Java, compiled languages such as C / C++, or in any combination thereof; the particular languages given here are only examples. In some embodiments, program instructions 143 may implement multiple separate clients, server nodes, and / or other components.

[0189] In some implementations, program instructions 143 may include instructions executable to implement an operating system (not shown), which may be any of various operating systems, such as UNIX, LINUX, Solaris™, MacOS™, or Microsoft Windows™. Any or all of program instructions 143 may be provided as a computer program product, or software, that may include a non-transitory computer-readable storage medium having stored thereon instructions, which may be used to program a computer system (or other electronic devices) to perform a process according to various implementations.

[0190] A non-transitory computer-readable storage medium may include any mechanism for storing information in a form (e.g., software or processing application) readable by a machine (e.g., a physical computer). Generally speaking, a non-transitory computer-accessible medium may include computer-readable storage media or memory media such as magnetic or optical media, e.g., disk or DVD / CD-ROM, coupled to or in communication with computer system 140 via I / O interface 144. A non-transitory computer-readable storage medium may also include any volatile or non-volatile media such as RAM or ROM that may be included in some embodiments of computer system 140 as system memory 142 or another type of memory. In other implementations, program instructions may be communicated using optical, acoustical or other form of propagated signals (e.g., carrier waves, infrared signals, digital signals, etc.) conveyed via a communication medium such as a network and / or a wired or wireless link, such as may be implemented via network interface 606. Network interface 146 may be used to interface with other devices 142, which may include other computer systems or any type of external electronic device.

[0191] In some embodiments, system memory 142 may include data store 145, as described herein. In general, system memory 142 and persistent storage 148 may be accessible on other devices 142 through a network and may store data blocks, replicas of data blocks, metadataAttorney Docket No. RAMP-00327-WOassociated with data blocks, and / or their state, database configuration information, and / or any other information usable in implementing the routines described herein.

[0192] In one embodiment, I / O interface 144 may coordinate I / O traffic between processors 141a, etc., system memory 142, and any peripheral devices in the system, including through network interface 146 or other peripheral interfaces. In some embodiments, I / O interface 144 may perform any necessary protocol, timing or other data transformations to convert data signals from one component (e.g., system memory 142) into a format suitable for use by another component (e.g., processors 141a, etc.). In some embodiments, I / O interface 144 may include support for devices attached through various types of peripheral buses, such as a variant of the Peripheral Component Interconnect (PCI) bus standard or the Universal Serial Bus (USB) standard, as examples. Also, in some embodiments, some or all of the functionality of I / O interface 144, such as an interface to system memory 142, may be incorporated directly into processor(s) 141a, etc.

[0193] Network interface 146 may allow data to be exchanged between computer system 140 and other devices attached to a network, such as other computer systems (which may implement one or more storage system server nodes, primary nodes, read-only node nodes, and / or clients of the database systems described herein), for example. In addition, I / O interface 144 may allow communication between computer system 140 and various I / O devices 150 and / or remote storage 148. Input / output devices 150 may, in some embodiments, include one or more display terminals, keyboards, keypads, touchpads, scanning devices, voice or optical recognition devices, or any other devices suitable for entering or retrieving data by one or more computer systems 140. These may connect directly to a particular computer system 140 or generally connect to multiple computer systems 140 in a cloud computing environment, grid computing environment, or other system involving multiple computer systems 140.

[0194] Multiple input / output devices 150 may be present in communication with computer system 140 or may be distributed on various nodes of a distributed system that includes computer system 140. In some embodiments, similar input / output devices may be separate from computer system 140 and may interact with one or more nodes of a distributed system that includes computer system 140 through a wired or wireless connection, such as over network interface 146.

[0195] Network interface 146 may commonly support one or more wireless networking protocols (e.g., Wi-Fi / I EEE 802.11, or another wireless networking standard). Network interfaceAttorney Docket No. RAMP-00327-WO146 may support communication via any suitable wired or wireless general data networks, such as other types of Ethernet networks, for example. Additionally, network interface 146 may support communication via telecommunications / telephony networks such as analog voice networks or digital fiber communications networks, via storage area networks such as Fibre Channel SANs, or via any other suitable type of network and / or protocol. In various embodiments, computer system 140 may include more, fewer, or different components (e.g., displays, video cards, audio cards, peripheral devices, or an Ethernet interface).

[0196] Any of the distributed system embodiments described herein, or any of their components, may be implemented as one or more network-based services in the cloud computing environment. For example, a read-write node and / or read-only nodes within the database tier of a hardware database system may present database services and / or other types of physical data storage services that employ the distributed storage systems described herein to clients as network-based services.

[0197] In some embodiments, a network-based service may be implemented by a software and / or hardware system designed to support interoperable machine-to-machine interaction over a network. A web service may have an interface described in a machine-processable format. Other systems may interact with the network-based service in a manner prescribed by the description of the network-based service's interface. For example, the network-based service may define various operations that other systems may invoke, and may define a particular application programming interface (API) to which other systems may be expected to conform when requesting the various operations.

[0198] In various embodiments, a network-based service may be requested or invoked through the use of a message that includes parameters and / or data associated with the network-based services request. Such a message may be formatted according to a particular markup language such as Extensible Markup Language (XML), and / or may be encapsulated using a protocol. To perform a network-based services request, a network-based services client may assemble a message including the request and convey the message to an addressable endpoint (e.g., a Uniform Resource Locator (URL)) corresponding to the web service, using an Internet-based application layer transfer protocol such as Hypertext Transfer Protocol (HTTP).

[0199] Unless otherwise stated, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.Attorney Docket No. RAMP-00327-WO

[0200] Although any methods and materials similar or equivalent to those described herein can also be used in the practice or testing of the present invention, a limited number of the exemplary methods and materials are described herein. It will be apparent to those skilled in the art that many more modifications are possible without departing from the inventive concepts herein.

[0201] All terms used herein should be interpreted in the broadest possible manner consistent with the context.

[0202] When a grouping is used herein, all individual members of the group and all combinations and sub-combinations possible of the group are intended to be individually included.

[0203] When a range is stated herein, the range is intended to include all sub-ranges within the range, as well as all individual points within the range.

[0204] When "about," "approximately," or like terms are used herein, they are intended to include amounts, measurements, or the like that do not depart significantly from the expressly stated amount, measurement, or the like, such that the stated purpose of the apparatus or process is not lost.

[0205] All references cited herein are hereby incorporated by reference to the extent that there is no inconsistency with the disclosure of this specification.

[0206] The present invention has been described with reference to certain preferred and alternative embodiments that are intended to be exemplary only and not limiting to the full scope of the present invention, as set forth in the appended claims.

Claims

Attorney Docket No. RAMP-00327-WOClaims1. A secure multi-party collaborative data activation system comprising:a processor configured to execute instructions stored in memory;a multi-tiered clean room architecture comprising a centralized cross-publisher and cross-walled garden clean room synchronized with individual one-to-one clean rooms for each brand-publisher relationship;a data delivery group management module that creates and manages data delivery groups containing audience segments from multiple data sources, wherein each data delivery group includes identifier mapping rules for distributing audience data across multiple publisher platforms using different identifier types;a cross-publisher measurement module that aggregates campaign performance data from the multiple publisher platforms within the centralized cross-publisher clean room, wherein the aggregated data is deduplicated to provide unified performance metrics across all participating publishers;a closed-loop optimization module that analyzes the unified performance metrics in the centralized clean room to identify high-performing audience segments, initiates targeted data collaboration sessions in individual one-to-one clean rooms to create refined audience segments, and automatically generates new data delivery groups for subsequent campaign iterations; and an identifier management module that intelligently selects and distributes different identifier types including at least one of mobile advertising identifiers (MAIDs), cookies, proprietary third- party identifiers, and hashed personally identifiable information to each publisher platform based on optimal match rates and platform-specific requirements.

2. The system of claim 1, wherein the multi-tiered clean room architecture maintains data isolation by keeping raw publisher data within individual one-to-one clean rooms while sharing only aggregated and anonymized performance metrics with the centralized cross-publisher clean room.

3. The system of claim 1, wherein the closed-loop optimization module creates new data delivery groups containing the refined audience segments and automatically activates campaigns across the same multiple publisher platforms to enable iterative performance improvement.Attorney Docket No. RAMP-00327-WO4. The system of claim 1, wherein the cross-publisher measurement module calculates deduplicated reach and frequency metrics by identifying user overlap across the multiple publisher platforms within the centralized cross-publisher clean room.

5. The system of claim 1, further comprising a publisher inventory access control module that enables retailer-specific retargeting by restricting audience segment activation to publishers that have authorized access to specific retailer inventory.

6. The system of claim 1, wherein the data delivery group management module supports integration of first-party audience data, collaborative audience data from commerce media network owners, and third-party audience data with distinct usage restrictions for each data source type.

7. The system of claim 1, wherein the cross-publisher and cross-walled garden measurement module provides performance analytics including lift and attribution analysis, projected scaled reach frequency estimates, incremental reach calculations, and return on ad spend assessments.

8. The system of claim 1, further comprising a permissions management module that enforces granular access controls for audience segment distribution, including time-based restrictions and data usage limitations for multi-party data collaborations.

9. The system of claim 1, wherein the closed-loop optimization module initiates targeted data collaboration sessions with individual publishers within dedicated clean room environments to create publisher-specific refined audience segments.

10. The system of claim 1, wherein the system maintains comprehensive audit trails of all data delivery group creation, modification, activation, and performance measurement activities.

11. A computer-implemented method for secure multi-party collaborative data activation comprising:establishing a multi-tiered clean room architecture comprising a centralized cross-publisher and cross-walled garden clean room synchronized with individual one-to-one clean rooms for each brand-publisher relationship;creating data delivery groups that encapsulate audience segments from multiple data sources with identifier mapping rules for distribution across multiple publisher platforms;executing campaigns simultaneously across the multiple publisher platforms using different identifier types selected based on platform-specific requirements;collecting campaign performance data from each of the multiple publisher platforms;Attorney Docket No. RAMP-00327-WOaggregating and deduplicating the campaign performance data within the centralized crosspublisher clean room to generate unified performance metrics;analyzing the unified performance metrics in the centralized clean room to identify high- performing audience segments;initiating targeted data collaboration sessions in individual one-to-one clean rooms to create refined audience segments based on the performance analysis; andcreating new data delivery groups containing the refined audience segments for subsequent campaign iterations to enable closed-loop optimization across the multi-tiered clean room architecture.

12. The method of claim 11, further comprising intelligently selecting identifier types including mobile advertising identifiers (MAIDs), cookies, proprietary third-party identifiers, and hashed personally identifiable information for each publisher platform based on optimal match rates.

13. The method of claim 11, wherein the multi-tiered clean room architecture maintains data isolation by keeping raw publisher data within individual one-to-one clean rooms while sharing only aggregated and anonymized performance metrics with the centralized cross-publisher clean room.

14. The method of claim 11, wherein aggregating the campaign performance data includes calculating deduplicated reach and frequency metrics by identifying user overlap across the multiple publisher and walled-garden platforms within the centralized cross-publisher and cross-walled garden clean room.

15. The method of claim 11, further comprising conducting targeted data collaboration sessions with individual publishers to create publisher-specific refined audience segments using advanced analytics techniques.

16. The method of claim 11, wherein the unified performance metrics include lift and attribution analysis, projected scaled reach frequency estimates, incremental reach calculations, and return on ad spend assessments.

17. The method of claim 11, further comprising enforcing granular access controls for audience segment distribution including time-based restrictions and data usage limitations for multi-partycollaborations.Attorney Docket No. RAMP-00327-WO18. The method of claim 11, further comprising maintaining cross-publisher frequency capping by tracking impression frequency within the clean room environment and enforcing frequency limits across all participating publisher platforms.

19. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform operations comprising:establishing a multi-tiered clean room architecture comprising a centralized cross-publisher clean room synchronized with individual one-to-one clean rooms for each brand-publisher relationship;managing data delivery groups containing audience segments with identifier mapping configurations for multiple publisher platforms;orchestrating cross-publisher campaign execution using intelligently selected identifier types for each publisher platform;aggregating campaign performance data from the multiple publisher platforms within the centralized cross-publisher clean room with deduplication to provide unified performance metrics;implementing closed-loop optimization by analyzing the unified performance metrics in the centralized clean room to identify high-performing audience segments, initiating targeted data collaboration in individual one-to-one clean rooms to create refined audience segments, and generating new data delivery groups for subsequent campaign iterations; andenabling retailer-specific retargeting through publisher inventory access controls that restrict audience activation based on authorized publisher-retailer relationships within the multi-tiered clean room structure.

20. The non-transitory computer-readable storage medium of claim 19, wherein the operations further comprise dynamically switching between different identifier types for individual consumers based on real-time match rate assessments and maintaining comprehensive audit trails of all data activation and optimization activities.

Images