Attention Score methodology discovers higher performance potential with YouTube video on CTV
2025-04-10 | Case study | Insights
Digitl's proposal for measuring performance through the custom attention score has provided us with insights that go beyond traditional KPIs. This framework is especially beneficial where there are constraints with access to sales data of clients or the resources for applying advanced methodologies like media mix modeling are limited. Unlike brand lift studies, this approach offers a cost-efficient way to gauge campaign success.
Anna Gneist | Head of Programmatic at SlopeLift
higher attention scores on CTV compared to the overall campaign average.
automation of data extraction and transformation using BigQuery, DV360 API, Dagster, and DBT.
distinct metrics combined into a unified Attention Score for comprehensive engagement measurement.
Attention Is the Currency for Video Advertising
Attention Is the Currency for Video Advertising
For a campaign to be truly effective, it must do more than just reach audiences – it must capture and sustain real user attention. However, standard video metrics, such as viewable impressions and view-through rates, often provide an incomplete picture. They measure exposure but fail to account for full engagement indication.
To address this, Digitl developed a formula, designed to measure how well video ads captured user attention in the various targeting segments. By using 10 metrics combining different aspects of viewability, audibility, and completion rates, this methodology provided a holistic view of engagement, allowing for optimizations and a smarter budget allocation
Through this innovative measurement framework, Digitl helped ensure that campaign performance was guided by actual user engagement, rather than surface-level metrics.
A Research-Backed Approach to Measuring Attention
A Research-Backed Approach to Measuring Attention
The team developed a structured, research-backed Attention Score methodology to measure YouTube video campaign performance, categorizing user engagement into three key stages: initial, ongoing, and completion. Each stage captured different aspects of user behavior across CTV, desktop, tablet, and mobile, ensuring a multi-dimensional assessment.
The process began with an analysis of past campaign data, identifying Active View and general campaign metrics as reliable indicators of attention. Using academic research on ad visibility, audibility, and attention retention, Digitl mapped key engagement drivers – such as ad audibility, visibility at various video stages, and user retention patterns – into a unified scoring model.
Digitl then automated data extraction and transformation, integrating BigQuery, DV360 API, Dagster, and DBT for seamless processing and real-time insights. These findings were visualized in a multi-dimensional Looker Studio dashboard, offering results at different levels and detailed breakdowns by device. This automation enabled continuous insights, allowing for adjustments to targeting, budget allocation, and creative strategies based on actual engagement trends.
Attention Score Drives CTV Success and Campaign Insights
Attention Score Drives CTV Success and Campaign Insights
The Looker Studio dashboard visualized various campaign breakdowns, including performance by device, days of the week, ad positions, environments, inventory types, regions, and line items. Key findings included the superior performance of CTV, particularly on weekends, and the potential for improvement by reducing outstream ads and focusing on YouTube inventory.
The analysis revealed that Connected TV placements delivered significantly higher attention scores (0.71) compared to the overall campaign average (0.57). This suggests that a campaign focused solely on CTV could have potentially yielded a 25% uplift in attention, leading to increased engagement and better outcomes.
The insights revealed aspects that the team didn't know about in this detail and helped them further improve their understanding of the activities. In the future, SlopeLift will especially use the Attention Score approach for upper-funnel campaigns. Among the future activities they identified as relevant to be evaluated against the attention score was an upcoming Masthead ads campaign.
The Goals
- Evaluate YouTube campaign performance.
- Achieve a minimum 15% increase in attention paid to impressions by optimizing targeting and budget allocation.
- Maximize audience engagement across CTV, desktop, tablet, and smartphone by focusing on high-impact exposure metrics.
The Approach
- Develop a custom Attention Score methodology to evaluate campaign performance.
- Group user engagement into stages – initial, ongoing, and completion – for deeper analysis of attention trends.
- Use advanced tools to automate data extraction, transformation, and analysis.
- Analyze past YouTube campaigns using the Attention Score to identify patterns in user engagement across different devices and further targeting segments.
The Results
- The Attention Score provided a clear, data-driven view of engagement across CTV, desktop, tablet, and smartphone.
- CTV emerged as the top-performing device, with 25% higher Attention Scores than the campaign average.
- The methodology helped identify high-performing audience segments, enabling smarter budget reallocation.
- In the absence of brand lift study relevant budgets or with insufficient data to conduct marketing mix modeling projects, the Attention Score methodology provides a cost-effective media efficiency framework.