The AI Advertising Podcast: S1

Episode 3

AI in CTV: Smarter Targeting and Seamless Buying

Cover of the AI Advertising Podcast featuring David Wells from Snowflake.

About This Podcast

Apr 9, 2025

As streaming platforms continue to dominate, connected TV (CTV) is becoming one of the most powerful advertising channels. But how can advertisers ensure better targeting, seamless media buying, and smarter ad placements in this evolving landscape? The answer lies in AI.

David Wells | Industry Principal, Snowflake

Greg Joseph | VP, Inventory Development, StackAdapt

Johnny Tan | Senior Director, Inventory, StackAdapt

00:00

Transcript

Diego Pineda (00:00:00)

Think about the last time you watched a show on a streaming platform. Did you see the same ad again… and again… and again? Or maybe you noticed something strange—your favourite streaming service has started serving you way more relevant ads. This is AI at work.

Connected TV or CTV is one of the fastest-growing advertising channels, and AI is transforming how brands buy media, target audiences, and measure performance in real time.

But how exactly is AI reshaping CTV advertising? And what’s coming next? To find out, we’re talking to Greg Joseph, Vice President of Inventory Development at StackAdapt, and Johnny Tan, Senior Director of Inventory at StackAdapt—two experts at the forefront of AI-powered CTV. We’ll also hear from David Wells, Industry Principal at Snowflake, who will break down how AI can manage audience segmentation and ad frequency like never before.

Podcast Intro (00:01:07)

Welcome to the AI Advertising Podcast, brought to you by StackAdapt. I’m your host, Diego Pineda. Get ready to dive into AI, Ads, and Aha moments.

Diego Pineda (00:01:21)

AI is already transforming digital advertising, but its impact on CTV has been a game changer. Traditionally, TV advertising was a static, bulk-buying process—brands would purchase ad slots in advance, hoping for the best. But Greg Joseph explains how AI is revolutionizing media buying on CTV, making ad placements smarter, faster, and more effective.

Greg Joseph (00:01:49)

Traditional TV is just a static buy. You go to a TV broadcaster and you say, I’m going to buy impression one through a million. And they’re going to be instance instances where performance are good. And they’re going to be instances where performance is bad because you’re buying supply in bulk.

I’ll use a bag of M&Ms. There are instances where I could buy a bag of M&Ms and there are some pieces that are not shaped the best. If I had the chance of picking which pieces I want and only getting the best pieces, I would.

The value of programmatic or programmatic and what we’re seeing with the evolution of AI and programmatic versus like the legacy way of buying TV is the legacy way of buying TV, you have to take what you got. And if the performance was bad, there’s really nothing you can do.

In the world of automation and AI and programmatic, you can adjust things in real time. You no longer have to say, I’m going to buy inventory from impression 1 through a million. I can actually say I’m going to buy Impression 1, Impression 10, and Impression 100 because they’re the most valuable to me and it’s going to allow performance to be consistent versus the like-see way of buying. You really have to take what you got and if performance wasn’t great, there was nothing you could do. Unlike what we see now.

Diego Pineda (00:02:58)

That’s a huge shift. Instead of wasting ad spend, AI-powered DSPs (like StackAdapt) analyze real-time data from premium publishers—think Disney, NBC, Paramount—to identify the most high-value audience segments. But AI isn’t just helping brands buy smarter—it’s improving how ads are measured.

Diego Pineda (00:03:24)

One of the biggest challenges in CTV advertising has always been measurement. Unlike digital ads, where you can track clicks, conversions, and attributions, CTV is more complex. But Johnny Tan says AI is closing that measurement gap.

Johnny Tan (00:03:41)

AI is really just a tool that depends on like what what is there underlying. So will AI help with measurement and attribution in a vacuum? Probably not. How would AI help with measurement and attribution? It’s really like having strong signals, having a strong product behind the scenes. So the work that we’re doing for martech initiatives, where we have a lot more oversight into the entire process from the delivery of the campaign into like how users specifically action off of different tactics within a campaign, having that data is what will power AI better, but it’s an intererative iterative process, right? Like if anybody had a blueprint to do it, then everybody would be able to do it really, really well. AI is just simply something that we need to continue to watch, manually oversee and optimize, just as the experts of the thing. And over time, I think it definitely will play a role in reducing the manual inputs so we can get performance faster. But I think it’s going to be a longer standing like process.

Diego Pineda (00:04:41)

By analyzing millions of ad impressions, AI can detect patterns in audience behaviour and optimize performance at scale—without requiring advertisers to be hands-on every step of the way. This is critical because marketers today don’t just use CTV—they run campaigns across multiple channels. Which brings us to AI’s Role in cross-channel orchestration.

Diego Pineda (00:05:11)

AI isn’t just improving CTV campaigns—it’s helping advertisers manage campaigns across multiple channels, including digital out-of-home, audio, display, and native ads. Johnny Tan explains how AI is becoming an orchestrator that connects the dots between different ad formats.

Johnny Tan (00:05:33)

The best way to put it is, like CTV is great. It’s easy to imagine where your ads are going to be played. Everybody’s seen a television commercial, but it is one part of the overall marketer’s toolbox. Performance is really driven through the whole like the whole funnel. It’s a mix of all of the different campaigns. CTV is a part of like you know, digital out of home, audio, video, native, the email marketing and stuff. So I think once you start adding so many different tactics and channels to reach your audience, the role that AI will play is to orchestrate all of them together in a more effective way. So you’re not just like going in and trying to optimize CTV alone, going in optimizing like audio alone, because that’s where you lose a lot of time. So being able to do it holistically and cross channel will be I think for me at least the most effective role of AI so that we marketers can really focus on like top line items instead of like going into the weeds and having to pick and choose like where your CTV ad is playing or which audience is to use like when you’re able to dump a lot of information and then just call out top line performance objectives with a polished and well-trained AI it should be able to allocate the most effective performance tactics across all channels. So CTV is just one part of it all but it really depends on like the reach of the different channels like the um integrity of the data for the users that are being like targeted and so on and so forth.

Diego Pineda (00:07:10)

This means less time manually tweaking campaigns and more time focusing on strategy and creative. But AI isn’t just streamlining media buying—it’s also changing how audiences are segmented and targeted.

Diego Pineda (00:07:27)

One of the biggest advantages AI brings to CTV advertising is smarter audience segmentation. In the past, targeting was broad—brands would rely on demographic data, hoping to reach the right people. But as David Wells from Snowflake explains, AI is creating hyper-specific “super segments” that improve targeting and ad relevance.

David Wells (00:07:50)

I think that AI will one, it’ll happen upstream of the activation platforms, right? it will It will help with some of the segmentation. It will help provide some of the connections of the different identity linkages, right? You know, creating a super segment or a federation of identity that underpins the platform, like we have a partner today that actually will normalize different identity linkages into a super segment. And then you can go in and you can actually create your own subset that is unique to you and you can use it to power your ad platform. That’s a AI use case. And then you have that managing greater frequency or more, I would say, efficient frequency on inventory types like CTV.

Diego Pineda (00:08:38)

AI-powered super segments aren’t just making targeting more precise—they’re also solving one of CTV’s biggest problems: ad fatigue.

David Wells (00:08:48)

So I think the thing about CTV today is that oftentimes it’s a very powerful interactive ad unit and and channel in which to target. However, some people think that they’re overwhelmed with the same ad over and over again. And then you as a marketer, you’re not accomplishing what you want. Yes, there’s recall, although maybe there’s some negative sentiment with that recall. If you’re using AI and to create this super segment and manage your frequency, then you know we need to hit David once ah during the week and then once on the weekend, and that’s going to spur him to have that powerful association with that brand and you know actually make a purchase.

Diego Pineda (00:09:30)

So where is all of this heading? Greg Joseph believes AI will play a huge role in dynamic ad placement—where AI-generated creatives seamlessly integrate into live content.

Greg Joseph (00:09:43)

When you think about live sports and the dynamics of live sports and some of the things that are possible. So one of the things that we’ve seen already in some of the companies that we’re talking to is the ability to have in-game advertising. So if you watch NBA playoffs or NBA basketball, one of the things you’ll know is when watching the game and looking on the center court, you’ll see a board that sometimes manually changes ads. We’ve started to talk to companies that actually have the ability to overlay ads on top of those boards via the stream, the live stream. And so they’re using AI technology to understand what’s in the stadium and then say, rather than going to commercial break, why not like just layer ads in the stream of the content to create better experiences?

We’re also seeing this for other partners we’re talking to when we think about our partnerships with TripleLift and ShareThrough as well as others, Cargos, a great partner of ours as well, is how do we leverage what we’re seeing in the stream of content outside of live sports to potentially add in advertising for that. And you’ve had companies like TripleLift doing very, very well doing that. You’ve had others excelling. And all of that is utilizing AI. It’s scanning the content, understanding what the content’s about, and then figuring out ad placement for that content. So if a family’s watching TV on a stream of content, and I’ll use like Modern Family, is there an ability to kind of have an ad or have like a consumer product on the table where the family is eating. That’s possible today. So I think a lot of that is coming in the future. Right now, we’re having a lot of these preliminary discussions to understand what is actually possible. We’re hoping with some of our partnerships with the likes of Curve and others, we’re gonna see a significant ramp up in how AI is utilized, not only for live sports, which I think is probably the next best opportunity but for all streaming content, how do you get there the right ads in the right place? And maybe that is not a commercial break. Maybe that’s within the content itself.

Diego Pineda (00:11:41)

This blending of AI, CTV, and creative automation will redefine how brands advertise on streaming platforms. Meanwhile, Johnny Tan sees AI continuing to reduce manual work for advertisers—allowing them to focus on high-level strategy.

Johnny Tan (00:11:59)

Whenever I work on stuff, I call it the North Star or the blank check, right? Like what do we need to get to in a future state that would essentially work me out of a job? I think the easiest way to imagine that is an advertiser comes in and they say like, I want maximum site visits for my specific brand. Create me a plan that will do that and tell me what sort of budget is required. I think there is a future that’s possible where Gen AI is able to like look at the historical data sets, look at the information that we’ve been able to digest and create like some sort of structured plan. It’s never going to be a one-size-fits-all all. There’s likely diminishing returns as you go through different buckets. Like for example if there’s 10 people in one audience, three of those are probably higher more likely to convert than the next two people. So like you as an advertiser are you trying to be effective with your budget and you only want to target the three that are what’s called the lowest hanging fruit or are you willing to spend a little bit more so you can, you know, increase your conversion from that three easy converters to the next five, which might be a little bit more um more difficult to convert or even extend it further. The future that I imagine is you put in a couple of prompts and then we could provide kind of different plans based on your willingness to ah invest a little bit more, so that way you can like really choose the options that are right for you.

Diego Pineda (00:13:27)

AI is already reshaping how brands advertise on CTV. But this is just the beginning.

Diego Pineda (00:13:36)

So what does this mean for advertisers today? Here are three key takeaways:

One, AI is making CTV more precise – Smarter targeting, better segmentation, and improved measurement.

Two, AI is optimizing across multiple channels – Helping advertisers orchestrate campaigns holistically.

And three, the future of AI in CTV is immersive – AI-generated creatives and real-time ad placements are on the horizon.

If you’re in CTV advertising, now is the time to embrace AI.

Podcast Outro (00:14:13)

Thanks for listening to this episode of The AI Advertising Podcast. This podcast is produced by StackAdapt. Visit us at stackadpat.com for more information about using AI in your advertising campaigns. If you liked what you heard, remember to subscribe, and we’ll see you next time.