Interviews, insight & analysis on digital media & marketing

Madfest Reports: Auditing attention, Dell on turning a buzzword into a benchmark

As the advertising industry continues to chase new ways to measure performance, attention metrics are one of the most talked-about topics in the industry. The measurement of attention and how to correlate it with business outcomes is a continuing challenge.

At one of Madfest’s most illuminating sessions, Marc Guldimann, CEO and Founder, Adelaide, Adam Edelshain, Business Director, MediaSense, and Vicky French, Media and Planning Consultant, Dell Technologies discussed how how attention can be measured most effectively.

Making the case for attention

Vicky French opened the discussion by explaining why Dell has embraced attention metrics within its media strategy. As she explained, Dell operates as both a brand and a retailer, with French focused on performance across the mid to lower funnel through dell.com in Europe.

“Our challenge,” she said, “is not just to raise awareness, but to actually drive customers to consider and purchase. That means every impression has to work harder.”

Traditional reach and frequency models fall short in such a fragmented ecosystem, she argued, while attention metrics “can help assess the relative quality of media and give us more confidence in the plan.”

French also highlighted that attention metrics are particularly useful for evaluating mid-funnel activity, which is “typically much trickier to KPI.” By bridging the gap between exposure and action, she said, attention metrics allow Dell to better understand what’s really driving movement through the customer journey.

Perhaps most importantly, she suggested that by incentivising media owners to prioritise quality over quantity, attention metrics can help shape a better media landscape. “If we’re telling media owners that we care about the quality of placements, they’ll produce better media products,” she said.

The AU approach explained

Marc Guldimann then took the stage to explain how Adelaide developed AU, a metric that scores media placements from 0 to 100 based on their likelihood to capture attention. “Think of it like a credit rating for media,” he said. “We wanted to create something that’s connected to business outcomes, easy to use, and, critically, auditable.”

Guldimann described Adelaide’s four-step process for generating AU scores, using three core data types, eye-tracking or lab-based attention research, detailed placement data, and outcome data from campaigns. Each channel, whether CTV, web, social, or audio, has its own attention drivers, and Adelaide builds bespoke algorithms for each.

After gathering input data via tagging, licensing, and scraping, Adelaide applies a machine-learning model to weigh various metrics.

“It’s a combination of research and outcome-based learning,” he said. “The goal is a predictive, practical metric that works across media and is embedded into DSPs, SSPs, and MMM tools.”

To discover how AU actually predict performance, it ws audited by MediaSense.

“Transparency was key,” Edelshain said. “We didn’t want a black box, and Adelaide welcomed that. We reviewed the data inputs and model operations and found the methodology to be outcome-driven and based on solid data science principles.”

Using randomly sampled data, MediaSense recreated the AU curves and tested them against four KPIs, aided awareness, recall, intent, and familiarity. The results showed a clear and statistically significant positive correlation between AU scores and all four outcomes.

While Edelshain acknowledged the need for further work, especially more data at the high and low ends of the AU spectrum, he was enthusiastic about the implications.

“Fifteen years ago, linking attention to outcomes was just a pipe dream,” he said. “This project shows that dream is now a reality.”

Where next for attention metrics?

The panel ended by looking to the future. Edelshain stressed the need for more outcome categories, more live campaign data, and more extreme-range data to deepen understanding. “It takes effort and coordination, but bringing real-world case studies to market would be hugely valuable,” he said.

Guldimann agreed, noting that the goal is to shift the market toward better metrics for trading. “We’re 15 years into this journey,” he said, “but the next 15 could be about actually changing the way we value media.”

French echoed the sentiment. “For advertisers like us, it’s about confidence,” she said. “We want to know that the media we’re buying is genuinely moving the needle. And attention metrics, when done right, help give us that confidence.”