The AI Advertising Podcast: S1

Episode 2

The Cookie Crumbles: How AI is Changing Ad Targeting

Cover of the AI Advertising Podcast, featuring Randy Newman.

About this Podcast

Mar 26, 2025

With 3rd-party cookies disappearing, the digital ad industry is at a turning point. But while cookies crumble, AI is stepping in to fill the gap, offering smarter, privacy-friendly ways to target audiences.

AI isn’t just replacing cookies—it’s reinventing digital advertising for the privacy-first era.

David Wells | Industry Principal, Snowflake

Randy Newman | CEO, Colour

Denis Laboda | Senior Director, Data, StackAdapt

00:00

Transcript

Diego Pineda (00:00:00)

For years, Google played a game of will-they-won’t-they with third-party cookies.
Marketers braced for impact… only to have the deadline pushed back. Again. And again.

But in 2024, Google changed the script. Instead of a hard cutoff, they’re giving users a choice: keep cookies, or opt out. And let’s be real—most people will opt out.

This means the future of ad targeting isn’t cookie-based at all. So, what’s next? AI.
Artificial intelligence is already stepping in—with contextual signals that are privacy-friendly, with machine learning, and predictive analytics to make ads smarter, not creepier.

Today, we’re pulling back the curtain on how AI is stepping in to replace cookies, and the impact it is already having on how advertisers target audiences.
And we’ve got the experts to help us figure it out:

Denis Laboda, Senior Director of Data at StackAdapt, who’s been preparing for this shift for years.
Randy Newman, CEO of the digital media agency, Colour, who’s seen first-hand how AI is revolutionizing micro-targeting,
And David Wells, Industry Principal of Media, Entertainment & Advertising at Snowflake, who will break down how data strategy is evolving in the AI era.

So, if you’re wondering how AI will shape the future of ad targeting—and what you must do to stay ahead—stick around.

Podcast Intro (00:01:29)

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 (00:01:43)

For years, Google’s back-and-forth on cookies had advertisers feeling like Lucy pulling the football away from Charlie Brown. But Denis Laboda, says smart advertisers saw the writing on the wall and prepared anyway.

Denis Laboda (00:01:57)

It’s been many years, right, in terms of Google’s announcements, cookies going away, cookies staying, lots of delays. Every time, of course, there’s been a new announcement, the feathers have been ruffled, there’s panic, things get delayed. People kind of forget about it. So I think overall, like the industry has been slowly moving towards you know having more, I’d say, sophisticated solutions and just cookie-less solutions in general.

But I think just the delays from Google have prolonged all of this. Right. So if like the first time Google said, ‘hey, cookies are dead’ and they actually stuck with it, I think things would have moved a lot faster. But I would say at the same time, you know Safari and Firefox haven’t had cookies for a while, right? Even with Chrome you know some people using incognito, they clear their cookies etc. So I think at this point, like if platforms and vendors don’t have a cookie-less strategy, I think they’re way far behind at this point.

Diego Pineda (00:02:52)

So, what’s replacing cookies? The industry has turned to three main solutions:
First-party data and universal IDs – Think email-based logins.

AI-powered contextual targeting – Where ads are placed based on content, not tracking.
Probabilistic and cohort-based modelling – This means using AI to predict user behaviour.

The biggest player in all of this? Artificial Intelligence.

Diego Pineda (00:03:25)

Without cookies, advertisers need a new way to deliver relevant, personalized ads. Enter AI. Instead of tracking users across the web, AI analyzes contextual signals and first-party data to understand intent. It’s like reading the room instead of stalking someone. Randy Newman has seen firsthand how AI-driven contextual targeting has made old-school keyword-based ads obsolete.

Randy Newman (00:03:52)

For a long time, you know, in this space, we have used contextual targeting, but contextual targeting was largely keyword-based, even though it was called contextual targeting, right? And what AI has done is it shifted it from that keyword focus to the actual context of the keyword. So it’s a powerful shift that’s happened because you know it’s allowed us to focus on content that really speaks to very specific audiences, you know, so for example, if if we want to talk to an endocrinologist about diabetes, but we don’t just want to you know look at that diabetes keyword, we want to talk about a specific subset of, you know, the illness or a specific treatment program. We can, you know, input the the necessary criteria to to attach that additional context. And that gives us a much more focused delivery method in terms of where these ads could show up.

Diego Pineda (00:04:57)

That’s a game-changer for brands—but also for publishers, because ads placed in more relevant environments perform better. But AI’s biggest impact? Micro-targeting.

Diego Pineda (00:05:12)

AI isn’t just making ad placement smarter—it’s reshaping audience targeting altogether. Randy explains how AI allows advertisers to hyper-focus on niche audiences, especially in industries like healthcare, where first-party data is limited.

Randy Newman (00:05:27)

Yeah, micro-targeting I’d say is is probably, you know for us working in healthcare, the biggest change, the biggest innovation and step forward. You know when we look at just across the board, all of our campaigns, if we’re targeting a broader audience, and we’re using AI to reach that broader audience, the performance that we see is closer to what we would expect the performance to be. So it’s close to a benchmark. On a micro-targeted audience, the AI tactics far surpass those benchmarks, It just speaks to the strength of AI and really focusing on you know those smaller segments.

We’ve done that for a number of different specialties where, you know, we don’t have first-party data on that specialty. So we look to you know contextual AI segments. We build off of our keyword lists. We comb through clinical data trials and studies and the content on our clients’ websites to really look at like what’s the core subject matter here that we want to speak about.

Especially when you’re talking healthcare care and medical, it’s you you’re dealing with a subject matter that’s very, very specific and very, very technical at times. Putting that kind of specific content in, you know, an AI-based targeting approach, the outcome will be a very specific and relevant audience.

Diego Pineda (00:06:55)

This isn’t just about healthcare—it applies across industries. AI can find small but valuable audience segments that traditional targeting methods would have missed.

Diego Pineda (00:07:07)

So, AI can optimize targeting. But where does the data come from? How do advertisers ensure they’re feeding AI the right signals? That’s where data strategy comes in. David Wells from Snowflake explains how companies need to rethink how they collect and use data.

David Wells (00:07:28)

One of the things that we like to look at when we talk to large marketers is, you know, let’s start with some of the metadata associated with the tech platforms that you leverage today, right? So if I’m a very large marketer and I want to drive more efficient activation to like acquire new customers, am I getting the metadata associated with those past impressions that I can query them in of themselves? So it’s not only about training a model, but am I at least providing some cursory analysis of that metadata so that I understand when I ran my last campaign, what are the different nuances or criteria that led that campaign to perform so that I can optimize on my next campaign activation.

And hopefully I can do that in such a way that it’s accessible to a non-technical persona. You can also use an AI generally to ask questions of that metadata to say, okay, my last campaign was, you know, six to eight weeks long, provide me with some level of insight around the DMAs, the devices, the inventory types that were most performant. And when they asked that question of that metadata, it can then visualize those insights just directly in front of them with no technical knowledge. So that’s super important is, first and foremost, go get that metadata of the partners that you leverage today in adtech and martech.

And it’s not only adtech and say that DSP metadata. It’s also, you know, some of the martech metadata. So if you’re using a CDP today and you have SDKs that are deployed and there’s curated site activity of visitors to your… to your owned and operated properties. Are you landing that data in Snowflake? Are you munging it with that paid media metadata so that you see the correlation of one or the other? Or at the very least, are you looking at that site activity and understanding where people are landing on your owned and operated and they’re spending a lot of time on different pages? Or they’re landing and they’re looking at a couple of different pages and then they sort of click out of that window. And what is the sort of impetus for one situation or the other? And how do you leverage that data to optimize the personal experience on your own and operated as well as to invest in more dollars to drive more qualified traffic that stays on there a longer time.

Diego Pineda (00:09:55)

So, where is all this headed? What will ad targeting look like five years from now? Denis Laboda believes that the push for privacy-first regulation in advertising will dictate how far advertisers can go when targeting audiences.

Denis Laboda (00:10:10)

The industry moves really quickly and we are at the mercy of regulations and legislation as we could see like around the world, right um GDPR in Europe, UK, adds a lot of conflict, I would say to how targeting and measurement is done there. So I think just over time, people you know advertisers maybe become more comfortable with less like one-to-one targeted advertising and maybe more cohort-based advertising or using more of alternative solutions such as contextual. I think like the winds are heading in that direction, right? One-to-one advertising will likely become harder in five years. But you know that inventory, that is going to support that will be worth more.

Diego Pineda (00:10:55)

And David Wells predicts AI-powered predictive models will become the new standard in media buying. AI will decide where and what ads to serve.

David Wells (00:11:05)

AI could power the decisioning, right? So in the ad server, as much as the DSP. How do we train a model to sort of dictate where we serve that next ad? And I think that’s super important in addition to sort of the segmentation or the identity that happens upstream. I think the other thing that’s going to happen is creative variables, right? So as of now, we’re working on using AI and data to power different iterations of creative variables that are then used in an activation. And so that AI can say, hey, not only serve this ad to David versus another individual that looks like David, but serve this creative variable. And it can learn over time as those creative variables lead to the outcomes that we want so that its next decision is that much smarter.

Diego Pineda (00:11:59)

The death of cookies isn’t the end of ad targeting—it’s just the end of lazy targeting. AI may turn out to be smarter, more privacy-friendly, and more effective. But to stay ahead, advertisers must do several things:

First, invest in first-party data – Build direct audience relationships. Second, leverage AI-powered contextual targeting – Move beyond keywords. Third, explore universal IDs and alternative identifiers – Stay ahead of privacy laws.
And fourth, prioritize AI-driven data strategy – AI is only as good as the data you feed it.

So, what’s next for The AI Advertising Podcast? In our upcoming episodes, we’ll explore AI-powered personalization, the top AI tools for advertisers, and the use of AI in emerging advertising channels.

Podcast Outro (00:12:55)

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.