From insights to foresights: predicting what consumers will need next with Kerry-Ellen Schwartz

Listen

Watch on YouTube · Apple Podcasts · Spotify

Summary

Kerry-Ellen Schwartz, Senior Director of Global Foods at PepsiCo, on why hands-on, in-context consumer research still matters for transformative innovation, why she remains skeptical of synthetic data and AI personas for forward-looking work, how authentic brand conversations on platforms like Reddit are reshaping brand tracking, and why the enduring job of insights is to advocate for the real, weird, nuanced consumer.

Guest

Kerry-Ellen Schwartz — Senior Director of Global Foods, PepsiCo · Brand-side

From this episode — top claims

• Synthetic data is not a current fit for forward-looking innovation work: it is a model of past behavior, not real current consumer behavior, and it cannot predict the next big trend — for flavor foresight she must instead watch away-from-home channels, restaurants, and chefs and track how those signals trickle down to CPG. — Ep. 63, Kerry-Ellen Schwartz
• Because synthetic data is built on historical models, relying on it means perpetually replicating the past and cannot project drastic, unforeseen change driven by economic, political, or natural-disaster forces — COVID is the canonical example of consumer behavior no model could have predicted. — Ep. 63, Kerry-Ellen Schwartz
• The signal that an AI vendor tool truly adds value is that it speeds up the analysis of real data; the warning sign is AI personas and synthetic respondents replacing real people, where it is unclear whether anything is actually saved or whether the answers can be trusted at all. — Ep. 63, Kerry-Ellen Schwartz

aytm's take on this conversation

A model trained on the past replays the past, so synthetic respondents will miss the next shift and the clean-eater-with-a-junk-drawer contradictions real people carry. Keep AI for speeding analysis of real data, and watch chefs, restaurants, and live consumers for the trend coming next.

aytm's perspective, voiced by Molly.

Full transcript

Auto-generated captions — speaker labels aren't always available and wording may be approximate.

0:00 If you're just looking at historical data, and that historical context, you're always just kind of replicating what used to be or what happened in the past. You can't project, like with these with synthetic data, like a drastic change that could happen. You know, there're economic climate, there's political climate, there's, you know, natural disasters. There's all these things that are happening, again, outside of, you know, us brands that are happening to consumers that have an impact on their behavior.

0:31 Hello, fellow insight seekers. I'm your host, Molly, and welcome to the Curiosity Current. We're so glad to have you here. And I'm your host, Stephanie. We're here to dive into the fast-moving waters of market research, where curiosity isn't just encouraged, it's essential. Each episode will explore what's shaping the world of consumer behavior, from fresh trends and new tech to the stories behind the data. From bold innovations to the human quirks that move markets, we'll explore how curiosity fuels smarter research and sharper insights.

1:05 So, whether you're deep into the data or just here for the fun of discovery, grab your life vest and join us as we ride the Curiosity Current. Today on the Curiosity Current, we are joined by Kerry Ellen Schwartz. Kerry is senior director of global foods at Pepsico. Kerry is one of those insights leaders who sit right at the intersection of innovation and real consumer understanding. She's worked across brands like Lay's and Doritos, leading efforts to create entirely new product experiences, not just optimize what already exists.

1:39 Kerry's work focuses on blending hands-on consumer interaction with the merging tech, helping Pepsico navigate how insights evolve in a world where AI is accelerating everything. So, today we'll explore how research methods are changing in a fast-moving innovation environment, where synthetic data fits and where it doesn't and what it takes to stay grounded in real consumer understanding as technology reshapes the landscape. Carrie, welcome to the show.

2:08 Hi, thank you so much for having me. We are so glad you're here and if you're amenable, we're just going to jump right in. Um one of the things as we were doing, you know, research on you that really stood out is how hands-on your approach is. Consumer testing new innovations in home, in context observation, real-time iteration. And it seems like it's happening at a company and scale where maybe I would expect the default to be something more formalized like stage gate framework.

2:39 And I'm curious what if you had to unlearn about more traditional research approaches to take this kind of approach and if I can hit you with a double-barreled question, um what does it yield in terms of consumer understanding that something like a benchmark score from a survey might not? Yeah, that's a great question. I think um in the in the type of work that I'm doing which is really around um transformative innovation, trying to get into white space, traditional methods are helpful. They we definitely still utilize those. So a formalized stage gate process, yes, we're still doing all that type of stuff. But sometimes like I tend to want to take things a little bit further, especially when dealing with things like white space. Like there are there are instances where, you know, you're trying to innovate a product and it's not just a new flavor, you know, it could be a new format or it could be a new packaging type. Like it's really important to get consumers to touch and

3:36 feel those things. Um and it's very easy for us to get lost, you know, broadly within insights um in data. You know, data, you know, comes to us uh in Excel sheets or something and we can look at it and kind of think we know the answer, but um sometimes you have to kind of go outside, touch grass, actually talk to consumers, really get their feedback, and really, you know, hear directly like what's working, what's not working. And I find it, you know, my job with as an insights leader, as an insights person, and a person passionate about insights is to remember that it's not just the data on a slide that we're presenting to like senior management or executive leaders, it's about the consumer, right?

4:16 Like we need to make sure this is working for the consumer, making sure that we stay consumer-centric in everything that we're doing. And if we don't have that story, if I can't, you know, with confidence say like consumers love it because they've actually tried it or they've tasted it they've touched it, then I I don't think like all the data in the world will will solve that. Like you really need the consumer to kind of help guide the process. Yeah, AI can't exactly tell you what something tastes like just yet.

4:46 I hope not ever. Fair enough. Fair enough. And to support a lot of this work that you're doing, you're very intentional about the partners that work to support all of your innovation work. And instead of going to the biggest firms or the biggest start-to-finish off-the-shelf process, you often work with smaller boutique teams that can move quickly and stay really scrappy throughout that. Um, so what is it about those scrappy partners that make them the right fit for you in this kind of innovation work that perhaps the bigger, more polished brands don't have?

5:22 Yeah, like um, I I'd say there's there's a fair room for both, utilize both. I'd say the bigger firms that have their own internal frameworks and their own, you know, processes, it's also hard for them to get out of their own way sometimes and and think beyond that. And sometimes a boutique firm will say, "Yeah, you know, let's, you know, let's try that. Let's go pilot it. Let's go see if it'll work. And if not, we'll we'll pivot and do something else. Um and I just like the ability to have a couple people in a room, think, you know, think through a problem, try to solve it, test it out. I I I truly believe in test and learn. Um and and just saying like, all right, we can go to a mall and talk to a bunch of consumers and and get them in a room and see what they say. And, you know, and if that doesn't work or if that's not kind of leading us down what we were trying to solve for, let's try something else.

6:14 Let's go into people's homes. Let's, you know, do more kind of on the ground type of research. Um so I I do appreciate the the ability of of smaller firms to kind of not only think outside the box, but also be willing to pivot and and also be a scrappy partner with me. Um I don't have all the answers and that's why, you know, I lean on vendors to kind of help me solve it, but I'm also looking for partners who um can come up with scrappy solutions as well.

6:41 I love the still doing the mall intercepts. I think that's something that's refreshing for me to hear. That is to this day one of my favorite methodologies. And I don't see it get used a lot. Like, yeah, after uh you know, like after COVID, I feel like like it started to trend down, but yes, because that cross-section of people is pre or it used to be, especially pretty representative. And you get them you're getting them in real time, like intercepting them. I don't know. It was beautiful. I used to do that a lot and I loved it.

7:12 Yeah, I did that a lot, too, like way back when. Um but it you're you're right. Like you there's something about just kind of running into person. They're not prompted. They're not You kind of put them on the spot. And so you get like really raw responses and you get like truly human behavior um if they're like willing to talk to you. Um I will say in like entertainment, I have friends who are in entertainment insights and, you know, they still do that as part of like movie screening and test screening. So it still exists, but it tends to be more um for things like where where consumers are like directly interacting with the end product and less so on, you know, within CPG world.

7:50 I think now, apart from like AI and synthetic data, I think a lot of the shift has gone to kind of an online quant survey route and if not that, then IDIs, focus groups, things like that tend to be the default versus intercepts. For sure. And and you hit on this a bit already, but I want to come back to it cuz it's it's all in this realm of what we're talking about. The idea that like access and understanding are not the same thing. We have these AI-powered tools and and even digital tools that are clearly expanding access. They allow you to reach more diverse consumers more quickly than than ever before. But when it comes to understanding something new, especially something experiential like food, access is only the first step. It does not equal understanding. And I'm curious, how do you think about that trade-off between breadth and depth, access and understanding when

8:47 you're designing a particular piece of research? Yeah, I think it sometimes it comes down to like how much risk, you know, the team is willing to kind of take on when it comes to like a launch plan. So, if it's something like you're changing a flavor of a product that's been in market for a while, you know, your R&D team's probably done a ton of sensory. Do you need to do more than just a concept test at the end of the day to kind of, you know, push it over the finish line? Concept test probably fine.

9:17 You know, you're not changing anything too drastically, so I wouldn't I wouldn't push too hard on like going outside of like our traditional methodologies there. But on the flip side, if it's a brand new product in a brand new category in a place we've never been, um showing up it to consumers in a in a way that they're not used to seeing us, that's when I would say like we need a more robust plan including talking directly to consumers and and kind of getting their direct feedback and making sure they're trying and touching and tasting these things um so that we really understand like yes, this is working or no, it's not working. But I think as part of that process, you you can kind of slot in your more traditional stuff um and traditional methodologies um in between the more kind of scrappy on the ground kind of work. So, for instance, um it's a new concept and it's a new idea. You can do a quant concept

10:16 test there. Um now you've got your feedback, you know it's a great idea, you've optimized it a little bit more, you've got new flavors, you've figured out the format, you've figured out the production, all the supply chain stuff. Now you have an end product. That's when you need to kind of before you build your launch plan like again, take a second to kind of again get feedback. Is this still working? Do consumers still like this? Does it taste good? Is it kind of meeting all those kind of key metrics that that we have from like a flavor profile perspective?

10:47 All of that kind of on the ground insights needs to come back into play. Um whereas like again, you know, my earlier example of it's just a flavor swap, concept test probably fine, some R&D, sensory testing is probably fine and you know, you're you're good to go. But if again, if you're going into like a new category, new space, the risk level is much higher and so I try to balance out like what the risk is if we don't have those answers.

11:15 Makes a lot of sense. That's I think an important thing that you bring up about what's what's the risk if we do or don't have the answers. I know sometimes research can especially with larger projects can get very bloated very quickly. How are we staying to the poignant questions, the things that I absolutely have to know, and the impact of those versus what's a nice to have. Yeah. Yeah, exactly. And at the same time you're talking about the evolution of different types of research, uh brand tracking is also evolving.

11:45 We're moving from standard funnel metrics towards a more integrated type of analysis that captures more of an authentic brand conversation on on platforms like Reddit. And when I think of authentic interesting conversations, the Wendy's X account, Twitter account, always comes to mind for me, and they were really the first to do that and stand out in that way. I know, it's incredible. It's so entertaining also. Um what what does that unlock um that traditional brand tracking never really could do?

12:18 I think um I started off my career doing brand tracker studies a million years ago, and I lived in those funnels. And I always, you know, as a new research kind of wondered like is this real? Like are is this truly how consumers feel? Is this you know, this doesn't feel right to me because how much do consumers actually think about any brand? You know, like there's very few brands that consumers truly truly think about on a day-to-day basis that like can like tell you they love or hate or whatever. Um especially CPG brands.

12:53 You like there're a handful of brands that people are super passionate about and like they truly want to talk about the brand, truly love the brand, and you'll find them on spaces like Reddit. Um but like how many people talk about toothpaste? You know, like truly like really recommend toothpaste to their friends. Like not a lot. Um unless they had some crazy experience they just want other people to to have. Um so I think as social media, right, has has expanded and has become, you know, this other thing, people you can actually hear those raw conversations, and you can really understand like do people care? You know, off off the bat. Like do people actually care about our brand? Do they care if we do this thing differently? How are they thinking about us? And what you'll find through those conversations like through social listening is unless you're really messing up in some way probably not in the conversation which is good. Like you don't want to be in that kind of

13:48 right cuz it's messy. Um or if you've done something really funny or cool like the Wendy's example great. But does that you know, does that increase sales? Do more people go to your store, buy your food? Like I think that potentially is up for debate. I think I would need to see more kind of analytics of like how scrappy a person is on social media or brand is on social media, how it relates to actual consumer behavior. Um I think we see this more in like political realm, you know, like like um Democratic and Republic social medias like are having a heyday right now.

14:25 They're going crazy. They've they've tapped into the zeitgeist of what social media behavior is. Um but will that resonate at the polls, right? We don't truly know. Um same thing for brands. Like if you're really funny and cool as a brand showing up on TikTok does that make more people buy you? Probably not. Um but maybe. I'm not sure. Like I said, I don't have have the data in front of me. Um but I what I think it's important to tap into those those conversations people are having authentically about your brand in those spaces if they're having them at all. Um there are certain brands I think that that do play a role in people's lives enough that they'll they'll talk about it. Um I worked on Mountain Dew uh in a previous previous role here at Pepsico and people love Mountain Dew. Like they're super passionate about it. They're fans of it to the point where they do have a Reddit, you know, subchannel and they do talk about our flavors and they they speculate on what the next flavors will be. They I look at emails from consumers

15:25 directly saying like, "Why can't you bring back such and such flavor? It was so good." I love that talking. Yeah. There are some brands that consumers are super passionate about that sometimes you'll pick it up in a brand tracker, but you really you really need to be listening directly to and look and observing those conversations to understand like, "What is it about your brand that's driving these conversations? What is it about your brand that consumers care about to talk about amongst one another without you facilitating the conversation." So, I think I think as as as brands, it's really important to kind of track those types of like authentic places where people are are having these conversations and really see like what that equates to from like a a metric standpoint for brand tracking.

16:13 I think it's a I don't know. Maybe I you said you used to work on Mountain Dew. Here's a a piece of insight for you. When I was in college, I was so obsessed with Mountain Dew Code Red that I actually took a bottle of it to one of those make-your-own ice cream places and I had them I had them put it in the cream and like on the freezer stone, I had them make me Code Red ice cream. So, there you go. We had like consumers like tell us like you guys on that brand like, "You need to do candy. You need to do ice cream.

16:44 You like this flavor is so good. You need to do these other things." Um and from an innovation standpoint, that's great because again, it's going back to consumer centricity. You want anything that you do, anything you bring to market to be rooted in a consumer need or a consumer tension that you're solving for or something consumers are directly asking for. So, it was really fun working on that brand because again, the the fans were so passionate that um it was just so fun to just like hear pitches of flavor ideas from them. Like it was so interesting.

17:18 I bet. That's great. Yeah, I love that. Well, and it's also I think the point you're making about, you know, I think the the thing that's so appealing about like a brand tracker is that you do get to produce that nice funnel. And that's what we all want to see, right? We want to see conversion and we want to understand it. But the reality is that there are things that happen around the brand that are not translating neatly down the funnel. It doesn't mean they're not valuable things, right? But understanding how they relate to the funnel can be it's like it's own that's it's own project, right? Like Yeah, and I I think that the problem I think with those traditional kind of funnel metrics is if there's a drastic change up or down, um it's really hard to diagnose directly what's causing that. Is it your comms?

18:06 Is it a new innovation? Is it some other kind of noise that's outside of your brand that's causing an issue. Like it was I just recall um like 10, 15 years ago trying to diagnose those those swings in conversion to something tangible and it was always such a a pain because you you couldn't there was no direct correlation, right? Like they're happening in the marketplace that don't directly involve you that could be having an impact on those conversions.

18:37 Yeah. That makes a lot of sense. Well, to take us to an entirely, you know, different topic. Um let's hit synthetic data if you don't mind. Um and I know I want to say before we talk about this, I know that's a loaded term with multiple definitions. Um but in almost every iteration of synthetic data that I can think of, it is by design looking backwards because it's built on models that are built on historical data.

19:04 And yet the kind of work that you're currently doing requires you to look for what not has full for to look for things that are not fully formed yet. Things that haven't been said before. So, when you're trying to surface like emerging trends, breakthrough ideas, what do you think has the potential to get lost if folks rely too heavily on models that are rooted in historical behavior? Yeah, I I think actually I think it it kind of ties back to what we're just talking about with brand funnels. Like there is noise outside of the historical data, right? If you're just looking at historical data and that historical context, you're always just kind of replicating what used to be or what happened in the past. You can't project like with these with synthetic data like a drastic change that could happen. You know, there economic climate, there's political climate, there's, you know, natural disasters, there's all these things that are happening

20:02 again outside of, you know, us brands that are happening to consumers that have an impact on their behavior. We saw that drastically with COVID, right? No one saw COVID, you know, coming the way it did. You know, you could have modeled things and said like, oh yeah, this could happen, but like could they have modeled everything shutting down and everyone being in their houses? They could have modeled toilet paper at least. You really couldn't model that, right? Cuz it was it came unless you're like academic like researchers in like, you know, viruses and stuff like probably knew something like would eventually happen, but they even they couldn't have modeled have modeled like the consumer behavior of of what have happened and what did happen.

20:50 And so, if you're constantly relying on like a synthetic model to to to tell you what to do, something will happen to blow it all up and it it won't matter anymore. I'm I will say like maybe it's controversial take, but I I'm not a huge fan yet of synthetic data. I think it's it's not real, you know, it's not real it's data, but it's not based on current consumer behavior. It's based on past behavior. It's so it's just a a model of like here's what happened in the past when all of these things were happening.

21:24 But again, it can't really predict like what the next big trend is going to be. It's not going to tell me, you know, right now, you know, a a big flavor is, you know, barbecue. What's the next barbecue? It's not going to tell me what the next barbecue is. It's not going to tell me what I should be gambling on or banking on from a flavor perspective. Like I have to be again listening to what's going on in away from home channels, what's going on in like restaurants and and what chefs are doing and how they're communicating all the things that they're doing on from a flavor perspective and how that all starts to trickle down to consumer and CPG goods.

22:03 Like I need to be paying attention to that. Um, versus like just looking at past data cuz that for especially for innovation, that's not really going to help me too much. Yeah. No, that makes a a lot of sense. Um, and I mean thinking about AI in in a totally different context um, that I know you have sort of thoughts about. There's something almost fascinating about the way that AI is starting to sit between between essentially consumers and brands. Not just influencing discovery, but potentially shaping the choices that people even see in the first place. Um, and I have that experience all the time. I I love the every you know, when when AI especially the LLMs first started being released to consumers, I was like, this is the best internet exploring like the best way to view the internet that I've like, yes, amazing.

22:56 But sometimes often now, I'm like, I just want want see the search results. Like where are they? They are door knick was, right? Like they're gone now. So when you think about chatbots like chat GPT becoming that kind of mediator of the experience between brands and consumers, what do you think changes most about how brands are competing for attention? Um honestly, I actually have thought about this a little bit. Um I I kind of just see a world in which you you pay to be at the top or you pay to be inserted into the answer, right?

23:33 Like kind of like the way sponsored um results come up now in like a Google search. So I imagine a world where a person goes to like a chat GPT and says um I'm out of groceries and I need chips, I need soda, I need water, I need poultry, I need all these things. Create my my shopping list with prices for me. And I would expect it all to branded because of thing to be part of the search, right?

24:04 Like it's like a new SEO or something like that. Um that's where I kind of see things evolving um because that's how they've kind of evolved previously previous this, right? So I wouldn't expect anything different unless, you know, like uh these companies decide no, we want you know, our AI chatbots to stay completely unbiased, but I just don't see that happening because they're companies um and this is it's still it it's ad revenue. Like we've seen it with all the other like um kind of new like new tech things that have come out, you know, with like social media. If you remember Facebook when it first launched, it had no advertising. It was nothing. There was nothing sponsored. It was such a different experience than it is today where if you start scrolling out, every other post is an ad or something sponsored. Um we've even seen that now with with Tik Tok. Like if you scroll on Tik Tok every third or fourth

25:02 thing you see pop up is is an ad or something like someone's trying to sell you. So I wouldn't imagine AI or these chatbots to be any different. Yeah, and I and even from a an earned perspective, there's a lot of buzz now about the emerging AEO and and that the AI engine optimization. So instead it's it's you know SEO for AI systems. And so even on our side we're we're building back-end pages to our website that are scrollable by AI so that when you search who's the best market research tech platform, it is able to more quickly pull those things from our website and list us. So even I mean there's strategies that companies are doing that is not even from a we're giving open AI money to to rank us, but we're trying to competitively get that earned edge in those systems already.

25:56 Yeah, yeah, exactly. It's already it's already happening. It's just going to evolve and eventually, you know, 5 years from now like we will forget what Google searches were like pre-AI. Yeah. Yeah, my brother-in-law had was saying the other day, he's like, I heard somebody say, I Googled that in like specifically instead of saying like, oh I searched that because now there's this inclination that by just saying I searched that you chat GPT'd it or you Claude'd it or you looked it up in AI versus you actually Googled it.

26:29 So you're saying Google has gone from being a specific term to a generic term back to a specific term. Back to a specific term. Yeah. Yeah. Yeah, because it's like it's a very specific type of search. I don't know. It it's very weird. I kind of wonder like what like I have three kids. One of them is a 14-year-old boy who is all over all of this. He's really into computers and tech and he he's the one who gives me updates on things like AI. And he and his friends are very kind of like anti-AI. Like they they're they're not really fans of it just because it feels so like not real. Like it's they know everything on the internet's like not real, but like this is like a step too far in at least my 14-year-old son's mind. It's like I wanted to figure these things out. Now you're just giving it to me and I don't even know where you're pulling it from. Like it just feels like

27:25 it's it's so much so quickly that he's like, I want to learn how to code. I don't want AI to code for me. Um I understand that it's a tool, but now it can do everything. Like so where's the human in all of it? And he he kind of he's maybe a little dystopian, but he's kind of like like what happens in the future if AI can do all this stuff? And now we're going to have robots that have AI. Like what what happens to all of us? Like where where are the jobs that we're supposed to be doing? Do we just service the robots or do the robots service themselves? Like is it all AI?

27:58 And like he he's kind of like he and his friends are are kind of like, no, we're not going to use ChatGPT. It's so dumb. Like they're they're he's trying he has needs to rebel against something cuz he's a teenager and this is the thing that he's choosing to rebel against. He and his friends and I just find it very interesting. That is I think that's fascinating. And it's also like good, right? And I say that because you want like I mean we I think we all talk about this quite a lot, but it's like I think AI can be such a powerful tool, especially if you're working in a domain where it's very easy for you to spot where AI went wrong, right? So like as a researcher, if I ask it to do things for me, it can speed up things for me so easily. It's so powerful. But I can also quickly skim through something and be like, here's where you got wrong, right? And how does somebody who's 14, 15, 25, how do they develop that skillset of discernment?

28:53 You know, if with AI in the mix, so it is it's such a thorny Yeah, and like I said, like it it's I mean, his group of friends, you know, sample of, you know, four or five boys. Um, but again, like teenagers need rule, they have an urge to rebel, most kids, and like he's putting his energy to rebel against AI. So, I'm interested to see if where that goes in the future. So, I'm I'm not discouraging it, obviously, but I'm just kind of curious like how it will evolve cuz I know schools have shut down, you know, using AI for anything as they should because kids need to learn. Um, but at some point, you know, we're all, you know, within our insight space navigating like where these AI tools fit in our lives and our careers and you know, younger generations will need to figure this out as well. So, it's it's an important conversation to have like more broadly. Um, because right now like it's at that cusp where it can like go crazy and, you

29:52 know, it could it go so so far in the future it's hard to even comprehend or it could just end up being a Google replacement, which is kind of what I use it for, but I know I should probably be exploring more broadly what I should be using these AI tools for. Well, it's better than what I rebelled against. I think I rebelled against Twilight, so it's probably better. That's not that bad, either. I was like, "What's what's Molly going to say?" Yeah.

30:19 Oh, god. Well, you you mentioned that you're navigating AI in your profession, and so we're absolutely at this moment where every single vendor is talking about how they have AI and agents and things in some forms, but that doesn't often translate to real value. And we're all highly versed in the marketing fluff that can sometimes come about with these type of in injunctions and in places.

30:47 Um So, when you're evaluating partners, um especially the AI space, but in general, what signals do you look for that indicate that a tool is truly going to accelerate your work versus is just talking about it? Yeah, I think that's a really great question because it's sometimes it's hard to see through all the marketing fluff even when you're a part of making marketing fluff yourself. Yep, same, same. It's hard, right? And I I get, you know, I'm working with vendors and partners who are utilizing AI in different ways.

31:20 I'm I'm right now, when it comes to like the tools and things that that partners are sharing with me, anything that speeds up the like the analysis, right? Like if you've got a whole bunch of data, and you can like feed it through, you know, a chat GPT or something, and it just helps consolidate it more quickly, I think that's great. And I think that is what we should be using AI tools for. Where I get a little bit nervous is when we start talking about like AI personas and like all these things that are like not, again, not real people, um where we're supposed to be getting feedback from. And that's where I'm like, uh I get hesitant and I'm like, okay, like we've made up this persona based on all this historical knowledge that we know about our consumer. Is that right? Is that what we should be doing?

32:04 Like, how much are we actually saving versus just talking to an actual person um that exists in this world and is not synthetic. So, that's where I'm like, you know, what's the value here? Like, it doesn't feel like we're saving time to me, um but maybe we are because we can shoot, you know, the AI a bazillion questions, and it'll come back with some sort of answer. But again, are those answers real? Are they good? Like, can they be trusted? Um Anything where AI is just replacing like a person speaking as a person is where I get nervous versus AI just doing math and doing an analyses. Like that's the the cut off for me right now. Again, I'm still learning, still open to seeing how these tools evolve, but I just need to be convinced I guess a little bit more on the value of not talking to an actual human being versus talking to a chat GBT AI persona.

33:05 Right, and I I think related to that it goes back to something you said earlier around risk. And that's kind of how I frame it right now. I think especially when I'm talking to clients is is around like, you know, is this a tactical or strategic question? Is this a right now question or a looking forward question? If you had the data in hand, but you and you like if you could go through all your historical studies and try to find an answer and you would be satisfied that way. Like those are all cues that maybe a synthetic data is a good option here.

33:37 But when we're talking about strategic work, highly forward-looking work, it those are the areas where the risk is high and I mean I yeah, I don't think we're at a point either where a lot of us feel very confident about putting our eggs in that basket, so to speak. Yeah, exactly. And I I think kind of framing it around risk definitely makes sense. Um and also value. Like again, like a lot of partners come to me and say like you like this will cut down on the time and the cost, which as you know, are the two most deciding factors when you know, picking research vendors, how quickly they can get it done and how cheaply they can get it done.

34:15 But I'm I'm not always like, you know, faster and cheaper is better. So I I'm still again not convinced on like is it truly saving me anything if I have to go back after this and and again talk to real people when I should have done that in the first place. Yeah, I don't think AI is at a point yet where it can capture how weird people are and just predict that the interesting things that people do that surprise us. I mean, that's to me what keeps life interesting is the weird things that people do when confronted with different situations and it it can every time humans do something weird, like we're talking about COVID and hoarding toilet paper, it can look back, but I don't think that it can actually take that information and apply it to what if there was something similar that happened that humans would do because would we do toilet paper again? I don't know. We might pick something else.

35:11 And that's a really good point cuz humans are so nuanced and weird to your point. Like I I just remember doing some work a few years ago around like health like a health and wellness and talking to people who are like who claimed they were like super healthy and they were doing all these things and you know, they're working out every day and you know, they only eat clean and all these things, but when we went to their house and you know, assuming they lived alone, like I remember going to this person's house, they lived alone, so everything in their pantry and their refrigerator was their own and it was full of like junk food, right? Like and it was full of like these things that you would not expect a person who eats really clean to have in their house at all, you know?

35:53 And so that's and that was so great cuz then we could push on that. Like, "Hey, you said you eat really clean. Like why do you have, you know, this soda in here that's full of sugar, you know? Tell us what that's about." And then you can hear like, "Oh, well, sometimes I need, you know, an emotional boost and this reminds me of home and so I have a couple sips of of soda and I I feel better." And like you would never get that from a chat GPT, right? Like it just wouldn't come come through because people are weird and they do things that make absolutely no sense on paper, but but makes sense to them and that's really the role of insights, right? Is pulling out like, oh, well, there's some emotional comfort thing going on that people still need to tap into. So, even though you can come up with the cleanest, most amazing flavored CSD, um if, you know, and and they'll buy it, but they'll still need something nostalgic to hold on to. And like, that's an important insight, right?

36:49 That's so so deep that, again, you would not get that from an AI chatbot. Well, and honestly, that's just a plug for for, I think, you know, qualitative and in-home ethnographic approaches too, because I would even argue that in surveys, we are engaged with an aspirational self, right? Yeah, and and to my earlier point, like, that's why, like, even if I'm doing a lot of quant, like, I will still push to use some sort of qualitative to talk to real people, cuz again, you need to hear the the stories and the reasons why behind all the data. Like, the data can get you so far, but again, if you're pushing into new category or pushing some new innovation, you really need to uncover like those those nuggets of truth that you can only get from actually talking to a human being.

37:39 Absolutely. Kind of related to this topic, but the the through, you know, it taking it truly all the way through to practice. Um you've said that one of the most valuable skills in insights isn't just finding the truth, but knowing how to bring that truth into the business in a way that senior stakeholders can really hear. And I think it's so important to shift consumer and customer-related conversation away from leaders saying, I think or we the brand think, and toward what the consumer is actually telling us.

38:13 And I'm curious, like, to you, what makes that ability, the ability to really ground your work, as you were just saying, in what the consumer is actually, you know, saying, so critical for especially I'm thinking about like the next generation of insights leaders. Yeah, um I think that that's so critical because often times in in any big company ideas sometimes come from the top, right? They're not germinated directly from consumers or even from like a marketing team. Sometimes, you know, a senior stakeholder will say like, "Hey, this seems like a really good idea.

38:45 Let's do it." And like everyone's like, "Yeah, okay, we'll figure it out." I like to in my mind, like the role of insights is to say, "Is it a good idea? Like, let's kick it over to consumers. Will they buy this? Will they pay for it? Let's have them tell us the reality of this thing versus trying to, you know, shove it down their throats in a way. Um let's just hear hear them in their own words." And that's why I I love qualitative research so much because it's great to have a number on a page, but I love to have a number on a page with a video of, you know, consumer either talking to that point or kind of saying like why this may not work or where we could potentially optimize. Like, it's it's important for the role of insights to just be the consumer. Like, your job is to represent your consumer. So, if you're in a meeting and you hear something come up that seems crazy to you from a consumer perspective, you

39:42 raise your voice and say like, "Well, you know, the consumer probably isn't going to go for X Y Z. They may go for A B C instead. Let's, you know, try to pivot or or try to think through like how to position this in a way that makes sense for the consumer." So, I think it's really critical for the next generation of insights leaders to remind themselves that their role is the consumer. You know, that that's kind of your stakeholder at the end of the day is like cuz these are the people buying our products. Like, these companies wouldn't exist if people didn't buy the stuff.

40:14 So, you need to make sure that you're representing the person who's going to buy the end product and making sure that their needs are, you know, heard, represented, we're solving through their key tensions, their needs are being met. Um they're willing to pay for for this thing. Um especially these days when when money is is super tight, you know, is it something people are going to put value against? And if not, how do you how do you fix it and make it better?

40:44 Yeah. That voice of the consumer and being the advocate for the consumer, especially when it's not potentially something that additional senior leadership wants to hear, that's when it can get challenging, but to stay rooted in your advocacy of them, I think that's a really great takeaway. Yeah. And just to to build on that last point, it's not that ideas coming from the top are bad because, you know, sometimes they're great and it it's just that sometimes ideas coming from the top, like they still need to be optimized, they still need to be kind of vetted through the consumer. Um I think about like Steve Jobs and Apple when, you know, the iPhone was coming out.

41:19 Like no one was asking for the iPhone. Like if you had told people about the iPhone before it existed, they would have been like, "Why would I need this? Like I have a computer." Faster horse, right? It's not to say that these ideas that come down are are bad, but sometimes you just need to figure out like the right way to position them and the right way to make it make sense for the consumer or the right way to make it seem like the consumer is missing out by not having this thing, right? So, and like I just I just wanted to make that like, you know, super clear. It's not like ideas should always come from the consumer because consumers don't actually know sometimes what they need or want or whatever.

41:56 Um so, ideas have to come from all different places, but the point is as an insights lead, you are you know, again, you know, you said this to to advocate for the consumer. So, if this thing doesn't make sense, figure out how to make it make sense for the consumer, right? Like they're in work with the consumer to find out what that that angle is to optimize the product to make it better. So, Carrie, this has been such a wonderful conversation. Lots of insights. Um you've made me think about a lot of things that that I don't even think I've I've been thinking too much about lately, so I really appreciate that. For someone listening who wants to drive meaningful innovation like you do, but feels overwhelmed maybe by how quickly tools, expectations, AI, and consumer behavior itself are changing, what do you think is the single most important principle that you would tell them to sort of ground themselves in?

42:53 Again, I think it just goes back to the consumer centricity. Like we're all people at the end of the day. Um we're trying to sell these things to people. What are What do they want? What are they How do they feel about it? What keeps them up at night? What are What are the things that you could potentially be solving for to kind of help them through this product, right? And and sometimes it's it's it's not as simple, you know, the consumer wants this, so we build the thing. Sometimes it's, you know, tapping into an emotional need, you know, over a functional need, or maybe it's a nostalgic thing that you can tap into.

43:30 There's some like human truths and and kind of humanness that you can always tap into um as an innovation leader and as an insights leader that can help explain or help, you know, storytell what you're trying to do. Like it it you can kind of ignore all the the tools and all the other stuff if you can just make that connection to a person. That's a perfect way to close us out.

43:58 And Carrie, thank you so much for your time um taking the time to chat with us today. We've covered a lot. Um my takeaways and our how to evaluate AI and actually the purpose of it versus the fluff. How to engage with people more authentically and to keep that as the centric part of insights moving forward and also Mountain Dew. So, thank you again so much. The Curiosity Current is brought to you by AYTM. To find out how AYTM helps brands connect with consumers and bring insights to life, visit aytm.com. And to make sure you never miss an episode, subscribe to The Curiosity Current on Apple, Spotify, YouTube, or wherever you get your podcasts.

44:43 Thanks for joining us and we'll see you next time.

Produced by aytm

Curiosity Current is made by aytm, the consumer-insights company. Guests speak for themselves; the synthesis is ours. About aytm