Uber launches platform-specific attention metric with Adelaide and Kantar

Uber Advertising announced on October 31, 2025, a partnership with Adelaide and Kantar to create a custom attention measurement system for its advertising platform. The collaboration produces what the companies describe as the first platform-specific, performance-based custom model for Adelaide's AU metric and the first integration of Kantar brand lift data into this type of attention measurement framework.

Custom AU for Uber Advertising fuses Kantar's brand lift results with Adelaide's predictive attention methodology. Adelaide's system analyzes signals including ad size, time in view, clutter, and position to predict a placement's likelihood of capturing attention and delivering business results. The integration with Kantar's brand lift data creates what Uber characterizes as a more precise measure of media quality compared to traditional viewability or engagement metrics.

The measurement approach addresses a persistent challenge in digital advertising: determining whether advertisements not only reach consumers but actually drive meaningful business outcomes. Edwin Wong, Global Head of Measurement Science at Uber Advertising, stated in the announcement that advertisers want to know more than whether an ad is viewable. They want to know whether it drives attention and engagement, he explained.

Testing the custom model across Uber Advertising campaigns revealed performance benchmarks that exceed industry standards. JourneyTV, Uber's in-ride video format displayed on tablets, scored 11% higher than average tablet video benchmarks. Journey Video Ads on mobile devices achieved 41% higher scores than mobile video benchmarks. Journey Display Ads shown on in-ride screens registered 39% higher than mobile display benchmarks.

Post Check Out placements on Uber Eats demonstrated particularly strong results. Display ads in this format exceeded mobile display benchmarks by 40%, while video ads in the same placement outperformed mobile video benchmarks by 43%. These findings validate what Uber describes as the high-attention value of its real-world advertising environments.

The methodology represents a departure from standard attention measurement approaches. Traditional metrics focus primarily on whether users see advertisements, measured through viewability thresholds. Engagement metrics track interactions like clicks or video completion rates. The custom AU model attempts to quantify something more complex: the relationship between observable attention signals and actual brand impact.

Marc Guldimann, CEO and Co-founder of Adelaide, characterized the collaboration as an evolution in AU model training. Uber Advertising's willingness to share comprehensive brand lift data reflects the kind of transparency all publishers and platforms should strive for, he noted. As the AU Ecosystem expands, Adelaide hopes to see more partners embrace this approach.

The integration process combined multiple data sources. Adelaide's methodology incorporates detailed information about Uber Advertising placements and ad delivery into its predictive model. Kantar's brand lift results provide outcome data that enables the system to correlate attention patterns with actual changes in brand metrics. This combination allows the model to predict which placements will drive business results rather than simply measuring whether ads receive attention.

Nicole Jones, Chief Media Commercial Officer at Kantar, explained that attention has become a critical metric for advertisers. This initiative shows how brand lift data can be used to create smarter, more predictive models, she stated in the announcement.

For the advertising technology sector, attention measurement represents a growing priority. Multiple measurement providers expanded attention capabilities throughout 2024 and 2025, reflecting advertiser demand for insights beyond traditional performance indicators. Nielsen integrated Adelaide's AU metric with reach data in October 2025, creating efficiency metrics that combine audience reach with attention quality.

The technical implementation relies on Adelaide's AU Ecosystem, a marketplace where attention measurement and activation tools connect advertisers with media quality data. The metric enables media quality scores calibrated with a platform's detailed exposure, attention, placement, and outcome data. Uber Advertising becomes the first platform to introduce this solution through the AU Ecosystem program.

Brand lift measurement has become standard practice for evaluating upper-funnel advertising effectiveness. Platforms including YouTube, TikTok, and Amazon offer brand lift studies that measure changes in awareness, consideration, favorability, and purchase intent among consumers exposed to advertising compared to control groups. These studies typically survey consumers after ad exposure to quantify campaign impact on brand perceptions.

The distinction between Uber's approach and existing brand lift studies lies in the integration methodology. Rather than conducting separate attention measurement and brand lift studies, the Custom AU model combines both data types during model training. This integration enables the system to predict brand impact based on attention signals rather than measuring impact after campaigns conclude.

For marketers operating within Uber's platform, the measurement system provides insights into how different ad placements perform relative to both Uber's internal benchmarks and broader industry standards. The 43% performance advantage for Post Check Out video ads on Uber Eats, for example, suggests these placements capture substantially more effective attention than typical mobile video advertising environments.

The emphasis on real-world contexts distinguishes Uber's advertising inventory from purely digital environments. Riders viewing ads during transportation or customers seeing ads after completing food orders exist in physical situations that may influence attention patterns differently than users scrolling social media feeds or browsing websites. The custom measurement model accounts for these contextual factors when predicting attention quality.

Research into attention's relationship with advertising effectiveness has produced nuanced findings. Analysis of 873 multimedia campaigns representing over $3.2 billion in media spend revealed that overall channel attention levels do not directly correlate with channel-level brand contributions or cost-effectiveness. Some advertisers may be overinvesting in high-attention media placements while underinvesting in lower-attention placements that deliver stronger business results.

This complexity underscores the value proposition of outcome-based attention models. Rather than optimizing purely for attention duration or intensity, systems that integrate attention signals with actual brand impact data can identify which types of attention correlate with business outcomes. A placement that generates moderate attention but consistently drives brand lift may deliver better value than high-attention placements with weak outcome correlation.

The data requirements for creating custom attention models present barriers for smaller platforms. Building predictive models requires substantial volumes of brand lift data across diverse campaigns, advertisers, and creative executions. Uber's scale as both a ride-sharing platform and food delivery service provides the data infrastructure necessary to support model training.

Custom AU for Uber Advertising is available to premium Uber Advertising partners. The announcement did not specify whether the measurement capabilities will expand to all advertisers using the platform or remain limited to select partners.

For the measurement industry, the announcement signals continued momentum toward platform-specific attention metrics rather than universal attention standards. Industry discussions about attention measurement standardization have revealed challenges in creating unified frameworks that work across diverse advertising environments. Different platforms feature distinct user behaviors, content types, and contextual factors that influence attention quality.

DoubleVerify's launch of social attention measurement with Snapchat in June 2025 demonstrated similar platform-specific approaches. That solution combined DoubleVerify's scalable ad exposure data with eye-tracking insights from Lumen Research to create attention metrics tailored to Snapchat's immersive advertising formats.

The proliferation of custom attention metrics creates both opportunities and complications for advertisers managing campaigns across multiple platforms. Access to more granular attention data enables better optimization within individual platforms. However, comparing performance across platforms becomes more complex when each environment uses different measurement methodologies and benchmarks.

Adelaide's AU Ecosystem addresses this challenge through standardized attention metrics that platforms can customize while maintaining comparability. The AU metric provides a common framework that platforms augment with their specific brand lift and performance data. This approach attempts to balance the need for platform-specific optimization with the value of cross-platform measurement consistency.

The performance benchmarks Uber reported span video and display formats across both ride-sharing and food delivery contexts. This breadth suggests the custom model accounts for format-specific attention patterns rather than applying uniform predictions across all Uber advertising placements. A video ad displayed on a tablet in a vehicle likely generates different attention patterns than a display ad shown after a food order, requiring format-specific modeling.

The 11% performance advantage for tablet video compared to industry benchmarks represents a more modest improvement than the 40-43% advantages observed in mobile formats. This variance may reflect differences in how tablet versus mobile advertising performs generally, or it may indicate that Uber's in-ride tablet environment offers less differentiation compared to standard tablet video placements than its mobile environments offer relative to typical mobile advertising.

For Adelaide, the partnership validates the AU Ecosystem's ability to accommodate platform-specific customization while maintaining measurement rigor. The company's focus on transparency and comprehensive data sharing as prerequisites for effective attention modeling reflects broader industry debates about measurement standards and data access.

Kantar's participation extends its presence in attention measurement. The company earned recognition as a TikTok Measurement Partner for Brand Lift studies and has conducted extensive research into attention's role in advertising effectiveness. Integration with Adelaide's attention modeling platform expands Kantar's brand lift data applications beyond traditional study formats.

The announcement positions attention measurement as a differentiator for advertising platforms competing for marketer budgets. Platforms that can demonstrate superior attention quality relative to industry benchmarks gain an argument for premium pricing or increased budget allocation. The quantification of attention advantages through standardized metrics makes these arguments more concrete than qualitative claims about advertising environment quality.

Timeline

• July 2024: Kantar publishes attention research analyzing 873 campaigns and $3.2 billion in media spend
• July 2024: Integral Ad Science expands Quality Attention measurement to mobile in-app environments
• December 2024: Nielsen integrates attention metrics into campaign effectiveness measurement through Realeyes partnership
• December 2024: IAS launches attention optimization tool showing 130% conversion lift
• June 2025: DoubleVerify debuts social attention measurement with Snap partnership
• July 2025: IAB Europe panel discusses attention metrics influence on programmatic decisions
• July 2025: Nielsen launches Outcomes Marketplace with Realeyes as first measurement provider
• October 2025: Nielsen and Adelaide integrate attention metrics with reach data
• October 31, 2025: Uber Advertising announces Custom AU metric with Adelaide and Kantar

Summary

Who: Uber Advertising partnered with Adelaide, an attention measurement company, and Kantar, a brand research firm, to develop the Custom AU metric. Marc Guldimann serves as CEO and Co-founder of Adelaide. Edwin Wong holds the position of Global Head of Measurement Science at Uber Advertising. Nicole Jones works as Chief Media Commercial Officer at Kantar.

What: The partnership created Custom AU for Uber Advertising, described as the first platform-specific performance-based custom model for Adelaide's AU metric. The system combines Kantar's brand lift data with Adelaide's predictive attention methodology, which analyzes signals like ad size, time in view, clutter, and position. Testing showed Uber's ad formats outperformed industry benchmarks by 11% to 43% across different placements.

When: Uber Advertising announced the partnership and Custom AU metric on October 31, 2025. The development represents the first integration of Kantar brand lift data into Adelaide's AU model type.

Where: The measurement system operates across Uber Advertising placements including JourneyTV in-ride video on tablets, Journey Video Ads on mobile, Journey Display Ads on in-ride screens, and Post Check Out placements on Uber Eats for both display and video formats. Custom AU for Uber Advertising is part of Adelaide's AU Ecosystem marketplace.

Why: The collaboration addresses advertiser demand for metrics that measure not just ad visibility but actual attention capture and business impact. Traditional viewability and engagement metrics do not quantify whether advertisements drive meaningful brand outcomes. By combining attention signals with brand lift data, the custom model enables prediction of which placements will generate business results, allowing advertisers to optimize spend across Uber's advertising surfaces for stronger return on ad spend.