How Attention is Reshaping CTV – Viant, Mastercard & TVision on Attention Economics
CANNES, France – Impatience with legacy measurement is driving a broader rethink of how CTV advertising is planned, bought, and evaluated. Attention data, once a niche academic curiosity, is increasingly being positioned as the connective tissue between media and creative, between impressions delivered and outcomes actually achieved. And for Mastercard, it has already started to reshape how the brand allocates spend and briefs agencies.
The conversation took place at the Beet.TV Leadership Sessions at Cannes Lions, where three executives sat down to discuss what happens when you replace gut instinct with second-by-second human engagement data:
– Jay Altschuler, SVP, Global Media & Agency Relations, Mastercard
– Tim Vanderhook, CEO, Viant
– Yan Liu, CEO, TVision
They were hosted by Beet.TV correspondent Pooja Midha.
The problem with proxies
Liu founded TVision in 2015 while a graduate student at MIT, built around a deceptively simple premise: use a camera and AI to passively measure whether people are actually watching TV, rather than asking them to press a button. The company was acquired by Viant in a deal that closed in May 2026 for $40 million, combining TVision’s attention measurement with Viant’s AI-powered programmatic platform.
“On CTV, have you ever seen a 50% viewable ad?” Liu said. “It doesn’t really happen. So you have to get audience attention from the true human. That’s what we do.” The viewability standard, borrowed from digital display a decade ago, was always a proxy for attention rather than a measure of it. On a television screen filling a living room wall, the concept barely applies.
Altschuler said the gap between theoretical and actual reach is larger than most planners assume. “Once you start looking at attention metrics, you realize there is sometimes a delta of actual attentive reach to opportunity to reach of sometimes 30 to 40 basis points,” he said. “Those wonderful flowcharts that you get from the agency, all the modeling that we do, that you need to hit 70% reach over 12 weeks, X amount of frequency, is all theoretical until you start to layer on attention data. And then it becomes real.”
Creative is the variable nobody was measuring
One of the more striking findings to emerge from TVision’s dataset is how dramatically attention scores shift not just by placement, but by creative. The same show, the same slot, a different ad, and the numbers can move substantially. Vanderhook described seeing brands with five to seven pieces of creative in market simultaneously, where the heaviest-weighted execution was not the best-performing one by attention.
“Creative A has much better attention and human engagement, but it only received 5% of the impressions. Creative B had maybe a 15-point difference in attention, but it received 70% of the impressions,” Vanderhook said. “I think it’s replacing gut instinct on the creative side with actual hard data on consumer engagement.”
Liu said TVision’s data spans linear TV, YouTube, walled gardens, and the open web, allowing the company to model which programming environments best amplify specific creative messages. Travel advertising, he said, indexes unusually well against country music programming. Food advertising, perhaps counterintuitively, performs strongly adjacent to crime content.
“If you sell food, if that gets people excited and want to engage with your show, maybe it’s not too bad,” Liu said. With more than 700 contextual segments defined by the IAB, the practical value of data-driven context matching is considerable.
Attention as organizational glue
For Altschuler, the operational implication has been as significant as the measurement insight. Media and creative have functioned as separate disciplines with separate agencies and separate lead times for years. Attention data, he argued, is pulling them back together.
“Somehow attention is this connective tissue to bridge these two conversations that were originally together 30 years ago, and then unbundled itself and became two separate processes,” he said. “Now I think attention is bringing it back together.”
At Mastercard, that has meant integrated agency teams planning holistically, with attention scores used to pre-evaluate creative before it runs, and media placements selected to maximize contextual fit.
There was also an unexpected sustainability dividend. “When you start to optimize based on attention, you happen to be in the most brand-safe places. And when you’re in the most brand-safe places and optimizing for attention, you need fewer impressions. When you’re running fewer impressions, less carbon,” Altschuler said.
That observation prompted agreement from Vanderhook: “When you optimize to attention, you’re really moving into the quality side of the inventory.”
You’re watching The Beet.TV Leadership Sessions at Cannes Lions 2026, presented by Viant. For more videos from this series, please visit this page. You can find all of our coverage from Cannes Lions 2026 here.
CANNES, France – Impatience with legacy measurement is driving a broader rethink of how CTV advertising is planned, bought, and evaluated. Attention data, once a niche academic curiosity, is increasingly being positioned as the connective tissue between media and creative, between impressions delivered and outcomes actually achieved. And for Mastercard, it has already started to reshape how the brand allocates spend and briefs agencies.
The conversation took place at the Beet.TV Leadership Sessions at Cannes Lions, where three executives sat down to discuss what happens when you replace gut instinct with second-by-second human engagement data:
– Jay Altschuler, SVP, Global Media & Agency Relations, Mastercard
– Tim Vanderhook, CEO, Viant
– Yan Liu, CEO, TVision
They were hosted by Beet.TV correspondent Pooja Midha.
The problem with proxies
Liu founded TVision in 2015 while a graduate student at MIT, built around a deceptively simple premise: use a camera and AI to passively measure whether people are actually watching TV, rather than asking them to press a button. The company was acquired by Viant in a deal that closed in May 2026 for $40 million, combining TVision’s attention measurement with Viant’s AI-powered programmatic platform.
“On CTV, have you ever seen a 50% viewable ad?” Liu said. “It doesn’t really happen. So you have to get audience attention from the true human. That’s what we do.” The viewability standard, borrowed from digital display a decade ago, was always a proxy for attention rather than a measure of it. On a television screen filling a living room wall, the concept barely applies.
Altschuler said the gap between theoretical and actual reach is larger than most planners assume. “Once you start looking at attention metrics, you realize there is sometimes a delta of actual attentive reach to opportunity to reach of sometimes 30 to 40 basis points,” he said. “Those wonderful flowcharts that you get from the agency, all the modeling that we do, that you need to hit 70% reach over 12 weeks, X amount of frequency, is all theoretical until you start to layer on attention data. And then it becomes real.”
Creative is the variable nobody was measuring
One of the more striking findings to emerge from TVision’s dataset is how dramatically attention scores shift not just by placement, but by creative. The same show, the same slot, a different ad, and the numbers can move substantially. Vanderhook described seeing brands with five to seven pieces of creative in market simultaneously, where the heaviest-weighted execution was not the best-performing one by attention.
“Creative A has much better attention and human engagement, but it only received 5% of the impressions. Creative B had maybe a 15-point difference in attention, but it received 70% of the impressions,” Vanderhook said. “I think it’s replacing gut instinct on the creative side with actual hard data on consumer engagement.”
Liu said TVision’s data spans linear TV, YouTube, walled gardens, and the open web, allowing the company to model which programming environments best amplify specific creative messages. Travel advertising, he said, indexes unusually well against country music programming. Food advertising, perhaps counterintuitively, performs strongly adjacent to crime content.
“If you sell food, if that gets people excited and want to engage with your show, maybe it’s not too bad,” Liu said. With more than 700 contextual segments defined by the IAB, the practical value of data-driven context matching is considerable.
Attention as organizational glue
For Altschuler, the operational implication has been as significant as the measurement insight. Media and creative have functioned as separate disciplines with separate agencies and separate lead times for years. Attention data, he argued, is pulling them back together.
“Somehow attention is this connective tissue to bridge these two conversations that were originally together 30 years ago, and then unbundled itself and became two separate processes,” he said. “Now I think attention is bringing it back together.”
At Mastercard, that has meant integrated agency teams planning holistically, with attention scores used to pre-evaluate creative before it runs, and media placements selected to maximize contextual fit.
There was also an unexpected sustainability dividend. “When you start to optimize based on attention, you happen to be in the most brand-safe places. And when you’re in the most brand-safe places and optimizing for attention, you need fewer impressions. When you’re running fewer impressions, less carbon,” Altschuler said.
That observation prompted agreement from Vanderhook: “When you optimize to attention, you’re really moving into the quality side of the inventory.”
You’re watching The Beet.TV Leadership Sessions at Cannes Lions 2026, presented by Viant. For more videos from this series, please visit this page. You can find all of our coverage from Cannes Lions 2026 here.
