Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Series 1 : Episode 3 of The AdTech Forum
29 Plays1 year ago
🚨 Series 1: Episode 3 of The AdTech Forum🚨
Big thanks to our Inaugural Series Sponsor, the
#creativeadtech king, Flashtalking by Mediaocean
In this episode, we talk with PepsiCo’s Ally Schnitzer on how she measures creative effectiveness, across the retail media landscape to drive sales.
#ai is going to enable faster creative testing" - Ally Schnitzer
Ally is wicked smart and has an edge working across multiple brands + partners. Adtech God, of AdTechGod, and Jeremy Bloom, of OhHello.io 🌞☕️, loved listening and learning to what Ally had to say about the effect of creative personalization across unique
#retailmedia initiatives.
This episode promises to deliver hard-hitting insights on an evolving industry and how creative plays a crucial role in optimizing #media #sales and #roi
Welcome to TheAdTechForum.com
Recommended
Transcript
Introduction to the series
00:00:05
Speaker
Oh, hello and welcome to the ad tech forum. Today is episode three of our five part series on creative brought to you by flash talking by media ocean. We're so excited to get episode three started today.
Guest Introduction: Ali Schnitzer
00:00:17
Speaker
Our guest, Ali Schnitzer. Ali loves the Texas Longhorns and telling people how to spend their money, which is always fun. She spent 10 years on the agency side in Chicago, working across various clients and verticals before moving over to PepsiCo four years ago.
00:00:34
Speaker
Ally currently handles all of retail media clients such as Target, 7-Eleven and Circle K. So we're super excited to have her here.
Exploring Retail Media Potential
00:00:42
Speaker
Ally, you're covering one of the hottest sectors in advertising. As the retail media lead across major, major partners, as ATG just alluded, walk us through some of the madness behind creative testing. Some of your madness, some of PepsiCo's madness.
00:00:57
Speaker
Yeah, retail media is kind of like the hot new girl at school, so everyone's really excited about her, but there's still a bit of mystery about it, which is kind of fun sometimes. But I mean, we get a lot of rich data, and it's a little bit of a blessing and a curse sometimes. There's a lot of opportunity to use that data in a variety of ways.
00:01:20
Speaker
But it's sometimes difficult to determine how do we actually decision appropriately given it's still, in some cases, very decent. And we're tying ourselves to actual metrics like sales and things. But a lot of the RMNs from what we've seen sometimes will lack the appropriate capabilities to give us the actual metrics that we want.
00:01:43
Speaker
We'll do a lot of testing or we'll try a lot of things, but the actual metrics that we lean on sometimes aren't necessarily what we would want or a little bit more directional. So it's sometimes you just have to make a gut decision like are we okay with using what we have or do we maybe need more time or different testing or different metrics to look at to actually decision off of it.
Personalization with Retailer Data
00:02:07
Speaker
What are some of those major metrics that you're decisioning off of?
00:02:11
Speaker
I mean, that's really like the biggest unlock, honestly. In some cases, we get it. In some cases, from some platforms, our favorite walled gardens and things out there, there are just a lot of privacy challenges.
00:02:27
Speaker
from a platform perspective as well as from a retailer. We're blessed, we're hashtag blessed to have the retailer data. The audience is, we're tapping into loyalty cards and things like that that provide a lot of that. Some are definitely more sensitive than others in terms of some of that data sharing that allows us to
00:02:48
Speaker
you know test a bunch of things and then use those log files and connect it back to the thing so we'll put a bunch of things in market and in most cases you know we hope are close with sales that will allow us to understand you know the incrementality piece of it but in a lot of pieces you know we're somewhat in a limited.
00:03:08
Speaker
So that leads us to our next question then, Allie. How important are creative personalization and optimization across the different partners? Because you just mentioned several different partners. Help us understand.
00:03:20
Speaker
Yeah, from a retail perspective, again, we're super blessed. We're, again, tied to your hashtag. Are we still doing that? I don't know. We have the ability to leverage awesome data. And it should kind of be a very low-hanging fruit, such a slam dunk for us. We know who you are. We know where you're shopping. We know what you're buying, whether it's in the category or our brands or something else. This should be a very easy exercise
00:03:47
Speaker
for us in terms of, you know, tailoring that creative based on, you know, are you near a Kroger and I know that you're a specific type of a grocery shopper or, you know, I know that you're buying genmerch at Target. How can I get you to kind of like upsell you into the grocery aisle, things like that. It should be very easy for us. But again, with some of those like
00:04:10
Speaker
Privacy concerns and things like that, we're a little unable to because again, we're leveraging a lot of the RMS. From a CNG perspective, we don't yet have like a true partner that is retailer owned. So we get to use third parties, which is really nice because
00:04:30
Speaker
they they're not tied to anybody and they have a lot more freedom and flexibility. So we are able to test a lot more and do a lot more personalization. It's not necessarily a sexy and a spawn, right? A lot of it, it's like, it's a gas station, it's convenience store. So it's a little bit trickier just because the behavior is a little bit different. But you know, if you're in an urban area, right, like I walk past the same time 11 every day and like,
00:04:55
Speaker
pick up something before I go work out. We should be able to, while I'm checking scores or something, serve you an ad because you're close to a 7-Eleven and we know that you're buying a Celsius or you're buying a fast Twitch or something. That's where we should be able to do that very easily. But again, we do have some challenges in that area from the proper RMNs. And what we've been seeing is just from a
00:05:21
Speaker
a resources perspective and roadmap. It's not something necessarily that the RMS themselves are building. It's expensive to build that technology out and to bring that in-house as in part of your own stack. But we also see some hesitancy to leverage some of these third parties because again, it does require data sharing. We're at this crux of, is it worth it to spend a lot of money to go create something that lives in-house that can do the personalization because we know it's where we should be going?
00:05:50
Speaker
Or do we relax a little bit on some of the privacy concerns and get into a good place where we can go contract with some of these other third parties using retailer data to ensure that we are going down the right creative path from a personalization perspective?
Automation in Ad Customization
00:06:08
Speaker
So with the right creative path, where does automation play into this with your retail media partners?
00:06:14
Speaker
Yeah, I mean, again, right now, it's like some of the like usual suspects, right? It's location, time of day, day of week, things like that. Are you you're serving to someone early in the morning on a Saturday while they're walking by a 7-Eleven? Maybe they, you know, need a different type of hydration product than if it's two o'clock in the afternoon, and maybe they're headed to a class or something. So automating, you know, whether it's a G Fit bottle for when your
00:06:42
Speaker
going to work out versus like a regular core Gatorade when you're a little baby or in the morning, things like that just automating based on those triggers as well as things like from an econ perspective like.
00:06:57
Speaker
We we can see what's happening from an inventory perspective right so like if they're out of stocks or things like that or if they're specific like promos or value offers during that time like we should be able to pull that in automatically and it should look different day to day you know if you know i've already
00:07:15
Speaker
I already bought a Pepsi one day. Don't serve me another Pepsi the next day. Serve me either a different Bev or a chip or something else. That's where that automation should be coming in is not relying on our team to say, hey, we know that you buy this and we think the pairings in your creative should be these two things. It should really be based on the data that we're using and what we already know and featuring it that way and again, pulling in
00:07:41
Speaker
pack shots from the site or you know, update different flavors based on what we know you already purchased, things like that. Ali, I have a question for you. So, you know, you talk a lot about data and personalization and customizing creative based off of, you know, time of day and location and potentially their mindset, right? Like, are they hungover? They just come out of work and need like a refreshing delicious drink or hydration for their skin.
Dealing with Data Loss
00:08:06
Speaker
How do you find some of this data loss, signal loss happening, affecting personalization, and how important is that in how it plays a role in creative overall?
00:08:18
Speaker
Yeah, I think one of the biggest places we see this is in our ROI models and our MMMs and things. We typically have seen really strong effectiveness where we've done personalization. But as you start to see a dip of that, you start to wonder, is it some of that more long tail
00:08:37
Speaker
Creative or some of those like longer tail impressions that are being served at a smaller scale That are so specific that just aren't being as impactful or have to be served, you know at an over frequency perspective That again just makes it like inefficient and ineffective So something that we've been trying to do is, you know looking at our whether we make it quartiles or something but like let's look at the top 50% of impressions that were served from a personalization perspective and focus on there and kind of remove some of those like
00:09:08
Speaker
creative pairings or some of that like specific personalization that is only being served like a really small amount of people that again just isn't being impactful. So optimizing that way to say like here's where the like the good scale is and like here's where we're seeing good performance and ensuring that we can keep the effectiveness as well as the returns you know as high as possible.
00:09:29
Speaker
Amazing.
AI's Role in Creative Testing
00:09:30
Speaker
As we're kind of starting off the year and there's been a lot of buzz about so many different terms the last 12 months, which one in particular do you think is going to be playing the biggest role for creative in 2024? What's your prediction there?
00:09:44
Speaker
Okay, I know I'm the third person to do this, so I hope that not everyone has already said this, and I hope I don't get canceled. But AI, I know there's, you know, it's a trigger word for some folks, and there's been a lot of discourse about it in various industries. I think in ours, and sorry to all of my creative agency friends, but I think it's going to enable the opportunity for faster creative testing,
00:10:09
Speaker
and just efficiencies, you know, every dollar matters. And so, you know, we at least think of ourselves a little bit as performance marketers, right? So when you're looking at nonworking versus working dollars, it really does make a difference. So again, I think the AI can help in some of that personalization and some of that automation, which
00:10:30
Speaker
It's a little bit of a creative destruction to some degree, but it frees up resources and time for people to be able to actually analyze the data or help actually decision off of it and look at it and say, where should we be investing? And maybe again, that's where we do some of the testing and we see, hey, it's really, again, these four or five things, then you can go to creative agencies or things like that and say, hey, let's focus on this. But I do think that there
00:10:57
Speaker
there is a need for it in terms of, again, just some of the performance gains, as well as some of those cost efficiencies and
Episode Conclusion
00:11:07
Speaker
things. I loved your feedback on all of these questions. We really appreciate you being here. I want to thank everybody who tuned in today for this episode of the Tech Forum, episode three, brought to you by Flash Talking by Media Ocean. We've obviously covered a lot of topics in a very short period of time.
00:11:25
Speaker
So tune in next week for another episode. And thank you, Ali, for being here. Thank you, Ali. Thanks, young.