TiVo Ads’ Fariba Zamaniyan: AI Won’t Fix Bad Data

RANCHO PALOS VERDES, CALIF. — The TV industry has more viewership data than it has ever had. But the perennial “big data problem” persists. Namely, most TV measurement lacks a way to join those data sets into a single picture of how people actually watch.

Fariba Zamaniyan, SVP of Global Data Licensing at TiVo Ads, argues that this seeming inability to connect data with actions matters more as AI takes on a larger role in advertising and measurement. The tools summarize, pull together, and predict, but they can only work with what they are given.

“AI is amazing. It’s an incredible component to my day,” Zamaniyan told Beet.TV contributor David Kaplan at Beet Retreat LA. “But AI does not fix data. It is only going to be as good as what goes in.”

Joining the signals

Viewership data has long been analyzed platform by platform, particularly in CTV. Zamaniyan sees that siloed view as the obstacle to understanding audiences. TiVo Ads recently combined set top box and ACR data into a single US TV viewership feed, which she described as the company’s answer.

As an independent provider, she said, TiVo Ads is not tied to a media buy and does not sit behind a walled garden. She framed the combined feed as a way to democratize data so that innovation reaches more than a few select platforms.

“When we think about what we want to watch, we’re not thinking, ‘Oh, I’m going to watch it on this platform or I’m going to watch it on that platform.’ We just want to watch a particular piece of content,” Zamaniyan said. “Then at that point in time, we decide where and how we’re going to watch it.”

Quality and transparency

The industry has been obsessed with scale, Zamaniyan said, but size is no longer the point. She wants buyers to evaluate data on its quality and representativeness, and to ask what a data set doesn’t provide as well as what it does.

“We as an industry need to do a better job when we’re individually evaluating assets that we’re asking the questions of what does that data provide, but also equally important, what doesn’t it provide, so that the limitations are clear as well as the advantages,” Zamaniyan said.

Beyond the living room

Zamaniyan expects audience intelligence to follow consumers out of the home. Cars now offer navigation, audio, and video on their screens, which shifts the conversation from platforms to environments.

“We are now advancing from not just thinking about platforms, it’s about environments,” Zamaniyan said. “It’s not just what happens on the glass screen. It’s not just their mobile device, it’s also about what’s happening in their car.”

A return to channels

Looking toward 2027, Zamaniyan expects a sharp rise in automation across planning, targeting, activation, and measurement. On the publisher side, she anticipates continued consolidation, and she sees a possible full-circle effect from it. Viewers who once chose among five channels may find themselves navigating a more channel-based landscape again, with AI doing more of the recommending.

“Fragmentation is going to get a little bit more solidified. AI is going to provide a better set of recommendations, making it easier to find, watch, and enjoy entertainment,” Zamaniyan said. “I’m excited that we are going to be able to harness that potential and move the industry forward on a global scale.”

RANCHO PALOS VERDES, CALIF. — The TV industry has more viewership data than it has ever had. But the perennial “big data problem” persists. Namely, most TV measurement lacks a way to join those data sets into a single picture of how people actually watch.

Fariba Zamaniyan, SVP of Global Data Licensing at TiVo Ads, argues that this seeming inability to connect data with actions matters more as AI takes on a larger role in advertising and measurement. The tools summarize, pull together, and predict, but they can only work with what they are given.

“AI is amazing. It’s an incredible component to my day,” Zamaniyan told Beet.TV contributor David Kaplan at Beet Retreat LA. “But AI does not fix data. It is only going to be as good as what goes in.”

Joining the signals

Viewership data has long been analyzed platform by platform, particularly in CTV. Zamaniyan sees that siloed view as the obstacle to understanding audiences. TiVo Ads recently combined set top box and ACR data into a single US TV viewership feed, which she described as the company’s answer.

As an independent provider, she said, TiVo Ads is not tied to a media buy and does not sit behind a walled garden. She framed the combined feed as a way to democratize data so that innovation reaches more than a few select platforms.

“When we think about what we want to watch, we’re not thinking, ‘Oh, I’m going to watch it on this platform or I’m going to watch it on that platform.’ We just want to watch a particular piece of content,” Zamaniyan said. “Then at that point in time, we decide where and how we’re going to watch it.”

Quality and transparency

The industry has been obsessed with scale, Zamaniyan said, but size is no longer the point. She wants buyers to evaluate data on its quality and representativeness, and to ask what a data set doesn’t provide as well as what it does.

“We as an industry need to do a better job when we’re individually evaluating assets that we’re asking the questions of what does that data provide, but also equally important, what doesn’t it provide, so that the limitations are clear as well as the advantages,” Zamaniyan said.

Beyond the living room

Zamaniyan expects audience intelligence to follow consumers out of the home. Cars now offer navigation, audio, and video on their screens, which shifts the conversation from platforms to environments.

“We are now advancing from not just thinking about platforms, it’s about environments,” Zamaniyan said. “It’s not just what happens on the glass screen. It’s not just their mobile device, it’s also about what’s happening in their car.”

A return to channels

Looking toward 2027, Zamaniyan expects a sharp rise in automation across planning, targeting, activation, and measurement. On the publisher side, she anticipates continued consolidation, and she sees a possible full-circle effect from it. Viewers who once chose among five channels may find themselves navigating a more channel-based landscape again, with AI doing more of the recommending.

“Fragmentation is going to get a little bit more solidified. AI is going to provide a better set of recommendations, making it easier to find, watch, and enjoy entertainment,” Zamaniyan said. “I’m excited that we are going to be able to harness that potential and move the industry forward on a global scale.”