Avoiding the Validation Trap with Conagra Brands' Bob Nolan

Speaker 1 0:00

I'm Bob Nolan, SVP of Growth Science with Conagra Brands, and you're listening to the CPG Guys Podcast.

PVSB 0:08

Hello and welcome to the CPG Guys Podcast, set at the intersection of commerce and tech. Your hosts, Sri Raja gopalan and Peter V. S. Bond, explore how brands and retailers engage consumers in a digitally driven world. And now, here are the CPG Guys.

Speaker 0:27

Hello and welcome to this episode of the CPG Guys Podcast. I'm of course Sri, your co-host and also CRO, co-founder at ThinkBlue Consulting, your trusted partner in your omnichannel development journey, where you can get in touch with me at Sri at thinkblueconsulting.co. Please do listen to my older daughter's music at www.rhearaj.com. Follow Lara Raj, my younger daughter is a member of the world's largest global girls band, Katseye, double Grammy nominee, just performed at Coachella, and what a wild weekend it was. Where Lara also got to DJ multiple aftersets. And this guy, yeah, believe it or not, was up till 6 a.m. three nights in a row, two weekends back to back, and I'm paying the price for it. But joining me today is my co-host and co-founder, Mr. PVSB, who also moonlights us at an industry and client engagement at Flywheel, Commerce Acceleration Division of Omnicom. Mr. Bond, you had something special going on last weekend. You were buying golf clubs. Tell us all about it.

Speaker 3 1:25

My daughter, Nadia, who's seven, came up to me a week and a half ago and said, Dad, I want to learn how to play golf. Apparently, her best friend has an older brother who competes in tournaments, and she saw them on TV and she's like, If I did that, could I be on TV? And I'm like, Yeah. And she's like, Daddy, I want to play golf. So we went off to Dick's Sporting Goods, and it was fascinating because she writes with her left hand shree, but we didn't know does she swing left or swing right? So we went to the golf section and I posed this to the expert there. And he goes, We have a way to solve for this. Follow me. And we walked across the store to the batting cages. And he put a softball on a T and gave her the bat and stood on either side and he said, Take a bunch of swings and tell me which you like better. And she clearly hit for more power from the left side and showed a level of comfort. And he goes, She's a lefty. Okay, now we know what kind of clubs to order her. So we have a set of clubs. They're in the color she liked. That was very important to her. And we've already set up golf lessons for her. Sri very excited about the journey that is about to begin.

Speaker 2:29

I hear the level of scholarship available for golf for females is a five, five is to one ratio versus men. So I can see the preparatory classes already beginning 15 years out, 13 years out. What is it? 12 years out?

Speaker 3 2:43

Yeah, about that. And you know, we uh so one of my colleagues at Flywheel had said to me, Peter, there's this actually really cool scholarship called the Evans Scholar that gives you four years, full tuition, room and board books and everything. And it's a caddy scholarship. It's called the Evans Scholars. And I'm already doing my research on that. So yes, there are huge opportunities for scholarships coming from the golf space. So that's a nice thing to know, too.

Speaker 3:10

Would you alright qualify for the caddy scholarship?

Speaker 3 3:14

We are, I think uh, well, we did watch Caddy Shack back in the day. So I think that may that may either qualify us or disqualify us as the case may be, Shri.

Speaker 3:24

Again, what does he do? He dates us. But in any event, make sure you're subscribing to our podcast on your preferred listening platform. We can get our latest episodes and go back and consume some of the 590 plus episodes were already published, which means, folks, we've got a blockbuster episode 600 coming up soon. But let's jump into who we've invited on the show. This is a very special episode. We've been waiting years to do this one. And uh, we got a great conversation lined up with a leader who's helping shape the future of one of the most essential categories in retail, needless to say, food. We're joined by none other, straight as a follow-up from his presentation at Cagny 2026 down in Florida. Mr. Bob Nolan, who I have actually had the pleasure of working with together at PepsiCo Bank. Now it's been a good 20 years, but I remember those days we were insights, geeks, and nerds together on one team. Yes. And he's now with Conagra Brands, a company behind some of the most recognizable and trusted brands in the grocery aisle. Bob brings a deep perspective on how brands stay relevant, how innovation actually lands with consumers and what it takes to win in today's rapidly evolving retail landscape. From navigating shifting shopper behavior to driving growth across iconic portfolios. Bob's got a front row seat to the challenges and opportunities defining CPG right now. But excited to dig right into it. Bob, welcome to the CPG guys. What a pleasure to do this with you finally.

Speaker 1 4:46

It's always been my dream. I follow your uh content ravenously over time since you first launched about six years ago. I also uh follow all of your posts. My favorite posts you guys do, no offense to you on this, is your daughter's stuff. I cannot get enough. At first I was like, why is he posting this? And now I'm absolutely hooked at it. I read every one of them, I look at all your pictures, I love your backstage, all the crazy stuff you see. I mean, so uh keep doing that, keep doing this too. This is wonderful.

Speaker 5:15

But I love that. Did you see the latest one? She made my dream come true. Nine Inch Nails is one of my favorite bands. Saw that, yes. He had me hang out with Trent Tresner, which kind of made my coachella. And uh, especially since he called me out when he was singing on stage. But Bob, we're honored that you've given us the time today. This one, as I mentioned, we've been waiting for it. So hopefully we're able to coach the world on how to think from a demand science perspective. But in the digital line and not so this episode, we'll include links to your LinkedIn profile and of course Carnegie's corporate websites for listeners to access while we go on with our conversation. So I'm gonna jump right into it, Bob. Obviously, the first one, you know, I'm going to start with something we observed at CAGNY. You famously shifted your entire insights budget from traditional validation, which used to be classic, syndicated data, get the share data, things of that nature. 9, 10 used to be back in the days when you and I worked together, it was three weeks behind schedule. And it was really checking what has already happened and cannot really be altered in any short duration. What was the breaking point, Bob, in the industry where you realized that asking consumers what they wanted was actually leading the company towards failed launches, incorrect launches, or simply not satisfying from an insights perspective?

Speaker 1 6:30

As you know, I come more from a customer orientation background, category management, customer insight, shopper, all those things. And you know, if you're in that world, it's very much about the real evidence. What's the real behavior? What's the real data say? Customers have no time or patience for nonsense stuff, right? So I think I grew up my whole career on that side. And I think when I started to switch over and the company gave me more responsibilities around brand insights or shop or consumer insights, you know, I had those same questions about is hey, how are we getting a better outcome from this research? And we did all the stuff everybody does. We did all the focus groups, we did lots of survey work, we did the product concept testing, we did the pre-testing and advertising, you name it, Conag was doing and spending money on it. But the business results were terrible. So I so we quickly realized, hey, we're spending all this time and everything's telling us everything's great. Products are all top box score, consumers love it, you know, it's exactly what they want to see, but the results weren't there. We've always done this research, but it's been no predictor of the outcome. It keeps telling us it's all wonderful. And then all of a sudden things fail. So they had a frustration, they didn't have answers. And for me, it was about leaning back on what I knew. So is there a more predictive way? Can we get better data? Can we get new data that lets us see over the horizon more? So that was really the start of our journey. And I have to say, I give them a lot of credit because they gave me the permission not just to do the new work, which we had already started, but to retire the old work. Because I think that's uh, you know, I I know you guys are very familiar with a lot of the the research methods that are out there in the world, and they give people a lot of comfort. They may not give them good answers, but a brand leader or a brand management person loves the fact that they have a score and the score made the decision whether I launch or not. Not me. And I and I I think, you know, in a world we're in, there's plenty of real evidence that will give us not perfect predictions, but at least will reduce the risk and give us choices that make more sense. But you also have to use your experience to make the decisions. We're never gonna, I don't think, be in a world where the computer is the AI, give us the answer and we just do what the answer is. I think they're gonna give us choices that we have to make as humans.

Speaker 3 8:51

Bob, welcome to the podcast. We're excited to have you here today. You know, the answer you just gave made me think about every time I ask a large language model, what does it think of the idea I've come up with? It always tells me that's brilliant! You're phenomenal, of course that's going to succeed. And I just I'm a little skeptical. So I understand why you wanted to find something that was a bit more of a an accurate predictor rather than uh than relying on some old methodologies that may not actually get you the outcomes you're seeking. Uh you've been quoted in the past as saying, and I love this, consumers don't do what they say they'll do. I always like to say that if you ask someone who they are, they'll probably end up describing the Brady Bunch family. Uh, but if you look what's in their basket, they look a little bit more like the Simpson family. In the context of your CAGNE uh 2026 presentation, I'd like you to specifically dig in on something you refer to as the validation traffic. Exactly what is it and why can that be dangerous for legacy CPG brands uh today?

Speaker 1 10:06

Yeah, I think it's not the consumers are trying to mislead us or lie to us. Even though one of my favorite books is Everybody Lies, which I think illustrates a lot of this, uh, but it's not intentional, they're lying to us. You ask people questions, they will give you the best answers they can. But we're pretty complex how our brains work. And I can ask, I could ask all of us here, who's trying to eat healthier this year? And you know, we'd all say, Oh, yeah, I'm trying to eat healthier. But if I actually got all your food service food food service receipts, I got all your scan data, you know, it might reinforce that, or it might tell me a different story about that. I think people always have good intentions, but they always answer with their aspirations. And same with, you know, what do you want? So I so we've we kind of I would say I know I know uh you know I get quoted sometimes and say we don't do any research at all. We still do some research to supplement and you know, like ethnographies and shop along, those types of things that are observational. But the two areas we've really tried to consistently avoid is asking you why you did something or asking you what you're gonna do next. And I think that's where we really fall down. I could ask you questions around circumstances. Hey, how many people were at dinner last night? How much time did you spend? You know, what appliances did you use last night? Any, you know, you can ask people and they'll give you answers and they're very accurate, especially in short-term memory on there. I can't ask you, is like, why did you pick that for dinner last night? And you might justify that or get to some answer on there, but but it just didn't lead us to outcomes that were positive. We, you know, that research validated what we had in our minds. You know, we already had, we think that people want this. Let's show them that, let's ask about that, and then they'll tell us, oh, yeah, we love that. That's exactly what we want. Or we would sometimes even go in with a blank slate. Hey, what's the next flavor you want to see? Right. And they'd give us all kinds of wild different answers, and then we'd find out in the marketplace, wow, that didn't sell at all. I mean, so you know, we we didn't cut the budget to nothing. We got rid of the 15 million of traditional research and we used all that money to buy the data. And that money funded us to get all the food service data. I have all the menu data, I have all the other channels, like the things that I didn't have before, like the natural channel. I could my goal is to try to see real behavior that's a little further over the horizon versus the old world. Let's look around myself on the shelf. Oh, what's Nestle doing? You know, what's craft doing? And then let's just, you know, they they know something, let's get into that. That that doesn't lead to an outcome that, you know, is going to grow the business.

Speaker 12:36

Bob, before I jump into the next question, panel data, consumer panel data, survey-based data is so popular in the industry. It's been around forever, and people are using left under, like you said, to even incubate innovation and and kind of self-justify many things, almost like a check the box. What I hear you say is the consumer doesn't do what they say on those panels. Not a lot 100% of the time, but in most instances. Because if you look at the actual receipts, et cetera, the story says something else. What is the future of consumer panel data? Is there a role for it, or is it slowly you think will phase out over time?

Speaker 1 13:15

I I think the traditional attitudinal parts of those will phase out. They need to phase out. They're not they're not useful. Like when people self-profile their interest or what their aspirations are, you know, that's not useful. I do think panels play a role. You already kind of hinted at it. The idea of someone scanning their receipt, because their receipt is the evidence of what they've done. I think those panels, as they've grown in size, I think of you know, multiple companies are providing those. We're always using those data sets to drive what we're doing. There's use in that. Because what's missing from the scan data is who are they? How does it look longitudin over time? Right? I want to know a household for the year. I don't want to know them just for one basket. I think there's definitely use of that data, but it's got to be grounded in real behaviors. And that could be food service behaviors, you know, scanning those receipts. It could be grocery stores, could be all the e-commerce receipts, right? All of those are, I think, are reliable. And we could profile the people, because those are facts, who you are, how old you are, who's in your household, all useful. But I think us getting rid of all that other noise that's in there of asking them, hey, why did you put what was your motivation in the moment when you bought this basket? Or what was your goal on there? It it just doesn't, it in fact, if anything, I would say it misleads you down the wrong path versus giving you something that's predictive.

Speaker 14:34

Ab, you're pulling me to ground earth, you know. I talked to Peter, I always profile myself as the Maharaja of Gai Pajama, but then my receipt says I'm a CPG guy on ground earth. But uh, in any event, one of your biggest wins was identifying the bold trend, which is the healthy choice power bowls via behavioral data while the rest of the industry was still going the old way with trays. How would traditional validation have killed that multi-million dollar insight and the outcomes you ended up getting in the marketplace?

Speaker 1 15:03

Yeah, if you were using just the scan data, looking at the category, looking at the aisle, you would have said there's pretty little evidence that both make sense. Right? All the big behaviors, all the time. If I pulled the top 10 SKUs, they were all traditional trays. Right? And some of them are ours, some of the competition. And you would have said, hey, there's a lot of risk us doing something that's not being done. And we had a lot of work that had to go behind that, like supply chain making a bowl is a little different, right, than doing a tray. Adaptable, but you know, not the same thing. So we would have certainly said none of that makes any sense, but that was the year we moved to the merchandise market. If you've ever been to the merchandise mark, one of the largest buildings in the world, the second floor is where the food's at. And we would go down and for lunch and the food, and this is observational research, and everybody was selling the food in bowls, mixed dishes. I mean, I mean, that's that's when Chipotle was really catching on fire, starting to rise. I mean, that's how people were eating. That's how the younger generation certainly eats. And then we just needed to find, okay, how big is that? What's the behavior? And that's getting that restaurant data. For us, we buy everything in food service. We buy all the menus across all segments from restaurants to QSRs to Fast Casual, and we mine that. What are the dishes? What are people customizing on the dishes? What are the top proteins and flavors in there? You know, all those attributes started to give us confidence that, hey, this bowl thing, while not in the store yet, is what everybody's doing. And what and if that's for me, real evidence is always going to trump looking back in the rearview mirror, as you said earlier. And for us, you know, we launched in Healthy Choice with Power Bowls as the first thing. We quickly converted Marie Calendar, a lot of the dishes to bowls that made sense. Some things make sense in trace, right? And there is, I will tell you this, there are consumers out there that like that component, especially older consumers, that component-based meal. You know, you take that dessert out of the hungry man dinner for them, that brownie, they aren't going to buy a bowl.

unknown 17:01

Right?

Speaker 1 17:01

They're going to find something else that emulates that behavior. We we found that out sometimes the hard way. So for us, knowing that audience, knowing where the future's going to be has been key to our success. And we quickly scaled that across the portfolio. So every place we brought a bowl, it's worked on every single brand. And it was also making the food real. I think the bowl alone probably wouldn't have gotten the job done. I think the other inspiration that we took from food service was the flavors, the global cuisines, you know, the amounts of protein. We saw that clearly in food service at the same time. So if we had done, if we had put in Salisbury steak in a bowl, you know, I'm not sure that would have wowed Gen Z or the millennials at the time. I think the food also had to be modern and relevant.

Speaker 17:45

Peter Howe interpreted what he said, and I think this is an important one. You know, if you want to fight for share in the category, you look at the past data. And then you're fighting for share, you're eating from the same pizza, and basically you're transferring one slice from somebody else. If you want to shape the category and grow share, then you don't stick to just the traditional old values. And the trace story is that validation trap. You are trying to shape the growth of the category and hence grow your share.

Speaker 3 18:15

Bob, I I love the food court at the merchandise mart. Having worked there and eaten many times, what I can tell you is that as much as I love ordering from the diverse food options there, I always feel that magnetic pull down to the end to the Billy Goat tab. I knew you were going to say that. It just sucks me in. All right. So getting back to this discussion about research methodologies that actually help you with your understanding and measuring the outcomes. Let's say you've moved away from those traditional methods and you've reduced your investment in those to zero or virtually zero. So, what do you do now, kind of a as a pre-flight to a major innovation, like you've recently introduced Rebel Roots Talistics and the Dolly Parton line, which you talked about at great length in uh at CAGNE, without what would be considered uh, you know, the proverbial safety net of a focus group that everybody has relied on for time immemorial?

Speaker 1 19:24

All the work goes in the front end. So if I have all this behavior-based data, you know, us mining that, us educating our brand and market partners and what it's saying, us debating. Because there's still, even though it's real behavior data, there's still interpretation, there's still the storytelling that has to go along with it. What are we really seeing and where do the dots connect? So for us, all the effort is into that. We're super involved with the RD team. So unlike where I've worked before, where the RD team was kind of like a handoff, and then they got back to we're intimate with those. They sit by us, the leaders right in the office next to mine. We're at our kitchens here in Chicago and Omaha all the time, going through and then trying the food, tasting the food, make sure that food does the job to be done that we designed. Because it's easy to stray from what the insights tell us to like, hey, we should compromise this. Or maybe people don't want to bowl. Maybe it's just a tray with one compartment. It's easy to get off track based on kind of like, hey, we just do this thing well. So we say, unlike kind of the research of the past where you would lop the research over the fence or lop the scores over the fence and hope they came to like, my team's involved in everything all the way to the end. And that's and that's not just with the RD, that's with the commercialization, that's with our sales teams, right? We've changed radically how we teach the sales folks about how to sell. We call it, we call it progressive selling, where they're using the same insights my team put together on why we're designing this, they're using that with the customer. I mean, I remember the old days when I was in the some of the category jobs, people in sales would say, Hey, why'd we come up with this item? I'd be like, I have no idea. Do you not know? You know, and we would guess why somebody invented this thing. We take the data from step one, everything we've talked about, it flows all the way through. So whether we're talking to the shopper, marketing folks at a customer, they get to see why did you do this? Who's it for? What's the job it's doing? What are the attributes people care about? We we keep that fluid throughout the entire process.

Speaker 21:21

I can relate to why do we create that item, Bob? You and I probably 20 years ago debated that heavy enough, especially as we were putting into the marketplace. So I'm gonna bring us back to traditional testing, is now down to zero from what we've heard for you. How do you know pre-flight a major innovation like I'm giving some examples over here Rebel Roots, Tallow Stakes, or the Dolly Parton line without the safety net of a focus group that I referred to?

Speaker 1 21:49

So for us, so for so for us, it's the confidence in all the design work we do. So, you know, we're big uh proponents of jobs to be done, laws of growth, and those concepts. All of that goes into design. Our team helps write the brief. RD executes it to make and we make sure it's on spec with the brand team. And then we launch it. And it's and at first, when we converted this, you can see where the rub might be. People would be like, hi, it's what you just said. How do I know this is going to work? Right? People were nervous about it. I need my score. I can't. How do you tell me to invest in slotting and support on this? And I have no idea if this is going to work. From the top down, people are like, the work goes in getting it right. The work doesn't go into after that. So we get it right and we still miss it. As I said, is our hit rate dramatically better? It is. I mean, the industry, as you guys know, you know, 80% failure rate for food launches over a two-year time period. We've gotten that down to half of that. So we but we still have failures. Sometimes we're too early. Sometimes we make execution compromises that don't, you know, bring it to life in the marketplace the way we want. So we're still not perfect by any means, but the hit rate's higher. And all that getting rid of that validation is about being faster in the marketplace for us, because you you know how slow those tests are. You put it in, you need four weeks, six weeks, you got to get a read on it, you got to come in and try to interpret it. I mean, that's time we don't have. We're already working two or three years out on innovation on there. We can turn, we just turned a huge success for us with the mega breakfast bowls in the marketplace. We turned it in two months from the insights and the data and the jobs to be done to on the shelf. So if we had validated that, we'd be another six months talking about it, debating what we learned from the validation. I mean, it doesn't make the product better to validate it. It makes the product better to get it right from the start.

Speaker 23:40

Couldn't agree more. You know, validation is a little bit more of an ego check versus assuring ourselves that we got it right from up front, right? Can we talk about project catalysts to modernize Carnegra through AI? Because you did speak about it at Cagny. How does AI-driven demand science replace the human element of traditional market research? Are we reaching a point where the algorithm now knows the consumer better than the consumer knows themselves? There are a few things around that, Bob. The LLM is directly married to the quality of the data that feeds into it. So have you all gone through the challenges of already creating the right data sets with privacy and security around that?

Speaker 1 24:20

Yeah, so we're kind of we've kind of gotten uh lucky that the AI came when it did, because you know, as we've been talking, we've spent the last eight years building that data foundation, right? Because the AI is nothing without the data and the quality of the data, how it's stored, the cleanliness of it. Because a lot of data, as you guys know, comes in pretty crummy from the different suppliers and stuff. So, but that's been our bread and butter using this data to make decisions, right? So we've been building that foundation for a long time. So about three or four years ago, as we started to acquire every data set we could get our hands on, it started to get harder for people to connect the dots, right? I can take two or three data sets, triangulate them in my brain, kind of, you know, dig in, roll my sleeves up, and that that's worked pretty good for us. A lot of these, we talked about the bowls, that's how we did it. But we've got 10 times the data sets. You know, we're mining the video data from Instagram and TikTok. We've got all the reviews data from all the different retailers' sites, right? We've got so much more unstructured data that it's become pretty hard for folks to sort through it all and you really get into the nitty-gritty and see the patterns any longer. So that's why we started down the AI journey, I said three, four years ago, you know, brought in, we're the ones that pushed the company to, we put a team together on my team that was going to help lead the way on this because it's new technology for most folks. Uh, and that team pushed the company to get ChatGPT in-house, you know, behind the firewall so the data is protected and safe, but we can still benefit it from this outside world as well as all the data we've acquired. And we've been using that the last two years to assist us in design. So, you know, it would be like more like still humans powering through, and then the AI has been a check. But I can see these last two years, we kind of flipped that around where the AI has given us the first pass on all the innovation and stuff. And then we still, you know, do a lot. It comes down to the skill of your prompting. So you're digging in and iterating. It still doesn't give you answers, step one, that are ready to go to the market. It gives you areas that have interest, it gives you some direction, but you've got to prompt it to get it closer. And then we'll still throw other data against it to say, okay, you know, and then part of this is also that our team's really been focused on how do we have a clear box from the AI. So as we've designed, we have a product uh project called Pi, so our product innovation engine. And it's not good enough for people to say the computer said this is the item. They want the backup and the story still. So we've designed a lot of this interface for us where you can still then click and it says, these are all the evidence pieces that got me to this conclusion. Because they still have to sell it a little bit. You've got to get management aligned, you've got to make the sales guys feel confident. So we've tried to be using the AI as much more of a clear box support tool with humans still in the driver's seat to make the final decisions based on their experiences and based on the volume of evidence.

Speaker 27:15

A reminder to audience, we're speaking today with Bob Nolan, SVP of Growth Science at Conagrabands. Over to you, Peter.

Speaker 3 27:23

Bob, I love hearing that you are using AI and large language models to take advantage of very rich data sets like customer reviews. I've talked about this on the podcast in the past, but I'm particularly drawn to one study that was done by a couple of professors, one at MIT and one in Northwestern, that hypothesized that just about everything you get from expensive focus groups, you could probably get by just reading reviews at Amazon. So they took six categories where they actually had focus groups and compared to the reviews. And what they found consistently is about 97% of the concepts that were surfaced and prioritized during the focus groups were available in the reviews. And so it's all out there. It's just your ability to collect, manage, and analyze it. And why not take advantage of that? So to follow up with on Sri's question about demand science, you've spoken about it being a key growth driver. How does moving away from validation allow companies like Conaggra to be more provocative and actually take risks that traditional research would have said, no, no, no, stay away. Dangerwell Robinson.

Speaker 1 28:35

Yeah, I mean, I think part of that's driven by the changing world. As you think about the new generation and what they want. You know, you heard me talk a lot about the how this Gen Z is really different in terms of what they want in food, even though their purchasing power is smaller, right? And it's easy. I think in the past we were always like, let's find the biggest behavior and let's go into that, because there's probably a lot of room for people in that one space. But the reality is that's still very backward looking, right? For us, you want to get to new demand. You know, that is finding those smaller pockets. It's finding the channel launches, getting the quality of data so I know what's happening in a dollar, club, e-commerce. You know, that granularity of data helps give us confidence on some of these launches, even though the behaviors may be smaller today. Right. And then that's and that's what the bowl, it's back to the bowl question. Because you know, the behavior at the time was small in retail sales on there, but the evidence was strong. And that in that for us to have good conviction on jumping into something, especially if it's a newer thing. I mean, you mentioned the rebel roots, the tallow fries. I mean, the evidence on tallow is up a thousand percent, but it's this big. It's still very small, but we know there's an audience that loves it. We know there's certain customers, especially some of the more modern customers that are leaning into it. So it gives us enough evidence to say that, hey, this is on the heart of the upswing. Can I tell you, is it a fad versus a trend? Maybe too early to know that. But but I know if you can get it in earlier on things that are growing, you'll have a better outcome versus waiting until they peak and are coming back down the other side. And we'll take some risk on those. And because we've we you know we've retired all these other metrics, the company and the leaders have gotten more comfortable using that evidence to feel good about the decision. And it took a while. It took us, I would say, two years. Because people, we would bring people from other companies in, they're like, hey, where's my score? Where are you giving me that? And and the culture really had to change for us, for people to say, I don't need that. Give me the right evidence and let's debate the evidence and then have conviction when we're gonna do something. If we're not 100% sure if we think we're on the earlier part of that adoption curve, let's make it a channel launch. Let's find a customer to partner with. Let's min, we can minimize the risk. I think big companies like us in the past were like, we've got to sell it to everyone and we've got to sell it everywhere. And that that world is gone, long gone. If you start, if that's the start of your conversation, you're not gonna have trouble growing because that's not where the growth's gonna be, as you know, all the bifurcation of income level that I've talked at length about, the different generational changes around ethnicity and global foods and cuisines. You've got to think about what's the demand, what's it gonna be, and how do I get in at the right time.

Speaker 31:21

And speaking of where the growth is going to come from and the completely changed demographics of our country over the last 10, 20 odd years, we've heard you use data from whole foods and sprouts to predict what will happen in traditional grocery like Kroger and Walmart two years later. Is it fair to say the natural channel therefore becomes a new version of a test market? And how does that data flow actually work?

Speaker 1 31:47

Yeah, we kind of have a I just hinted at it a minute ago. We have this uh product evolution curve that we use. So we try to take every food trend, whether it's an ingredient, a new attribute, a new type of cuisine, and we kind of plot them across this curve. You know, in the very beginning of the curve might be this exploration stage where it's like experimental restaurants. You know, it's uh it's you know, really out there. Or maybe you see it overseas. It's not even in the US yet. You know, maybe too early to get in. For us, the next one is really the understanding of restaurants. So that's a discovery phase. What are restaurants doing? And then we start to look at that's where the natural channel comes in, right? Because we see, especially around health and wellness and new attributes, they typically emerge there in retail before anywhere else. Right. So we try to look at those as early predictors of, hey, we've seen people that are eating this food, they like this attribute, they're starting to purchase it, still maybe very small in some of these other channels, and then we try to extrapolate the rest of that curve. Some things will die there, they'll get to that point and they'll go right back down because they were a fad. But other things will sustain from there and keep growing over time. So you think about, you know, protein really started, you know, restaurants first. Yeah, you think about the chicken craze. Chicken restaurants started rapidly growing because the food was interesting and provocative, but also people are looking for more protein overall. So, what does that mean in the food world? How does that translate to what we're doing in here? Overall, you know, that that's the key for us. We're gonna get dark. Yeah. Uh so for us, you know, I think the natural channel, restaurants, things that people are experimenting more with food before they get to retailers, where we spend all our time. We we we'll look at fully competitive in the current world. Hey, are we missing a size? Like PPA. PPA is a lot about fully competitive, right? Accessing channels. You know, our old employer, you know, that was their bread and butter, right? Getting the right size, the right place at the right time. We still do that work. But when you're talking about innovation, it's all about looking over the horizon and finding the behaviors as they're still moving towards retail.

Speaker 3 33:50

Bob, when when Sri and I were at CAGNE in 2025, it was readily apparent to us that Conagra and maybe one other company was set apart from everyone else, particularly around the whole GLP1 phenomenon. Most companies were either not talking about it or denying that it was having any impact on volume. You guys took it head on and said, no, no, we're reformulating our products to focus on the benefits. And to that end, it was like you positioned ConAgra as a beneficiary of GLP1 drugs rather than a victim of what was going on. So I want to kind of understand the behavioral science that underpinned that rather than to the point I made at the beginning, these, you know, asking someone, consumer surveys, how do I really help you realize that these users aren't eating less of everything, but are actually pivoting towards specific nutrient-dense frozen options, protein being a prime example of that?

Speaker 1 34:57

You probably both remember that November of 23, there was an article that came out and they had surveyed consumers who were on the drug. And they said, hey, what are you eating less of? And how much do you think you're eating less of it? And it was like the end of the world. Like, hey, I'm eating 30% less sweets. I'm eating 40% less snacks. I mean, you know, and I remember the market that day and the food sector wasn't great. And we quickly regrouped it and said, hey, that's obviously not how we would find the evidence on that. And we quickly, you know, we we it was on our radar, certainly, right before then. But we quickly pivoted to was like, how do we find the real answers on this? Where's the evidence out there? So we reached out to all of our partners and some new ones that were tracking, we seek data, purchase data for people that are verified on the drug. What they were doing before, what they were doing on the drug, and if they come off the drug. Because some, you know, some a lot of people don't stay on the drug, right, for a lot of reasons. Like, and then what's their behavior afterwards? And once we started getting that data, it took us two or three months to get our arms around that data. And, you know, we didn't want to jump to conclusions, but we started to see a really a different story than what that survey had told everyone. We started to see, yes, there's some winners and losers across that. The food overall is not that much different. You know, they're they're eating a little bit less food. And actually, you might be surprised. I just got the latest date on this yesterday. You might be surprised they actually spend more on food than folks that aren't on the drug. A little bit. But but but that that's not what I you know would have expected. But that data led us to the conclusion saying, okay, there's some winners and losers in our portfolio here. You know, winners, thankfully, single-served meals, frozen, you know, frozen, frozen appetizers, vegetables, frozen vegetables, also great. So we kind of that made sense. These people are looking for portion control, they need more protein, and they need fiber and stuff to keep their gut moving, right? It's one of the things doctors tell them to eat. We were all a little surprised we started seeing, like, hey, they're eating more meat snacks and less salty snacks. Because if you eat less calories, every calorie needs to have a purpose. Because you're not craving really anything. So it's like, how do I make sure my body's working? They're drinking more energy drinks because they've their energy is not as high as the calories coming in or less. They're looking for more nutrition, they're eating bars. So we started to see clear winners. And every time we got the data set to this day, because we get this every month, it's been pretty consistent. Same on the downside, unfortunately. You know, some of the down categories have been very inconsistent too. Traditional sweet baked goods, you know, candies, cookies. Uh one of the ones that's getting crushed in the data every time we see it is alcohol. And that, you know, and unfortunately, I feel I have friends that work over in the business, you know, I'd say a double whammy, right? Because they're getting the generational change, which is not great, and then the GLP ones on top of that. So, but but there's consistencies. There's some categories that are really aren't affected too in the middle. So for us, it gave us confidence saying, okay, this isn't the end of the world for our portfolio, but how might we lean in then? And are there opportunities on some of our brands to reformulate and change them? Yeah, the first part of any new trend is do we have products that meet that demand right now? And how do we make sure people know about that? We can do e-commerce for that. We can use a lot of online digital marketing. Uh does the product have the right communication? So that's when we added the on-track badge to healthy choice. We didn't need to change the products at all. They had everything I just talked about. But maybe consumers were buying them more already, but maybe we could help them on that journey. And then the last step for us is like, is there any new food or redesigned food? Like, is there something we should add fiber to? We got a lot of protein in this product. You know, those folks could use five grams of fiber. That would that, you know, that helps them also. So we have been looking at the innovation part also. But, you know, for us, that it uh it starts and ends with the data. If we didn't have that data, I don't know what the panic would have been, right? I I think we would have, you know, really struggled. It's like, hey, maybe we should reduce the cost of products. Maybe it's about trade. You know, we would have been on the wrong path. The real evidence of real behavior is the truth that you need to figure out how to market.

Speaker 39:04

So um validation takes months. Social trends happen in minutes, if not days. How's the death of validation increase your speed to market? And talk to us a little bit about the use of social data as well, especially Instagram, TikTok, things of that nature. And then any examples of idea to shuffle timeline that would have been impossible using the old model of validation and syndicated data?

Speaker 1 39:29

Yeah, I said we're still building muscle around going faster. Big companies, it's still, you know, how do you streamline the process? We've gotten a lot better at it. And I think when we put our mind to it, we can get something from really the evidence in the marketplace to launch in a few months, especially for a channel that's open, doesn't have a reset time and those things. But we still got we still have muscle to build in that space. Uh, you know, I think in the past we would have done all that research work, you know, and we wouldn't have been able to react to the trend fast. I mean, we mine the TikTok data. So yeah, you know, the thing I love about TikTok, certainly the funny calf videos, that's number one. But number two is that it's all about demonstrations. It's like people making the recipe, show me how easy it is. It's about clothes, it's about toys, about all those things. And for us, how do we, if a if a trend rises really fast, you know, how do we convert that quickly? It could be a new flavor, it could be a new ingredient. How do we convert that to the marketplace faster than we have before? We take all the the my team tracks the top viral recipes off TikTok. We take that and we partner with their chefs and we have a session with them every quarter where they make all the recipes. They go through and we get to see how hard it was it to make, you know, if it's an ingredient item, hey, could we transfer it to frozen food? And we talk about, you know, okay, is that trend staying? You know, are people still downloading that recipe? Are they buying the ingredients? We look for more evidence. The TikTok maybe is the cue for us, but we still want to make sure there's supporting evidence behind that. And then our goal is to get to the marketplace with that. I said more and more frequently it's through different channels because we can execute fast in those new channels that are emerging. But it's it's also providing the context and the evidence for our sales folks to go get the customer excited about it. We've we were at the salesman, we showed them all the videos where these products came from. You know, the original viral video said, Hey, it had this 10 million views. And we, because they're not always exposed to all that. And for them, they can show those videos to their buyer and say, oh, cow, I didn't know that, but wow, that's interesting and exciting. And it said it gives us more evidence to sell what we're doing.

Speaker 3 41:39

Bob, I think you just reinforce something that we say on the CPG guys all the time, which is if you're a marketer, if you're in innovation and you don't have TikTok and Instagram on your phones, and you're not exposed, you don't have to be out there posting. You don't have to be a creator. But if you're not seeing what's happening, you're missing so many opportunities that can grow your business. You have to be aware of what consumers are talking about.

Speaker 1 42:09

You got to be careful though. I will add to that. My wife has not appreciated uh all the clothes I'm ordering her from TikTok and all the other things. She's she sampled the food with them, but I, you know, she I get so sold in the moment, and then you have to click the bean. I just hit a button, and then that thing's at my house.

Speaker 3 42:28

I'll tell you this. The the other thing is I was I was laughing the other day. You're talking about erroneous market research information. I saw an interesting prediction that said that there would be a debilitating impact on the laundry detergent business because GLP ones would cause people to buy smaller clothing sizes so they wouldn't need as much laundry detergent. I I don't know if I can draw a causal link to exactly that. I think I'm gonna pour the same amount into the receptacle at the top of the washing machine, regardless of how much clothing I'm putting in there.

Speaker 1 43:01

I think you're absolutely right.

Speaker 43:02

But uh Bob, you know who's gonna love who's gonna love your purchasing clothes on the TikTok shop? As funny as it is at the time of this recording uh episode out in the marketplace, is Patrick Nomenson from TikTok shop, who gave us a uh pretty deep rundown of how it's done and why there is so much activity going on in TikTok shop. But anyways, back to you. Well, food's coming.

Speaker 1 43:24

I mean, I mean the act is there's a lot, people are already using. I mean, when I got married, somebody gave us a Betty Crocker cookbook. You know, we we didn't know how to cook, and sometimes we'd like, where are we ought to make? We went to that. Everybody nowadays is going to TikTok. What's the trending recipe? And and I think for food, it's going to quickly convert to clicking that button to fulfill that order from whatever your favorite supplier is on e commerce. So yeah, right now it's not so much food, maybe some specialty uh D to C food, but uh it's gonna be just a moment and we're gonna be ordering full recipes ready to go.

Speaker 3 43:57

Yeah, I think I think it a lot of it comes. Down to the nature of what the product is. Like we were talking with the CMO of a major non-alcoholic ready-to-rink beverage business last night. You know, let's be honest. You still have to think about supply chain and shipping freight for uh beverages in TikTok shop is not ever going, at least in the foreseeable future, is not going to be a viable commerce channel of any meaning to certain companies. So you have to you have to be a little realistic about that. But let me let me close this out, um, Bob, because I think you can offer some some good advice to uh other brand managers and innovators out there that are still have trepidation about launching innovation without the proverbial green score from the uh the blue chip testing agency. What's your message to them about the risk of not evolving past the old mechanism of validation?

Speaker 1 44:58

Yeah, I think it comes down to finding the right evidence, having a lot of debate around that evidence, right? Because you can lead to many different conclusions and outcomes, and then having confidence in your experience. And I think in the past, people didn't want to be on the hook. I think the scores were a safety net saying that, well, I launched that because of that score. I, you know, I, you know, how I knew it was a bad idea, but you know, the score was good. I think people have to have confidence. We hire great people here at Conagra, they have amazing experiences. Use those experiences, combine them with what we're giving you on the evidence, and make the call. And sometimes we'll be wrong, but let's do it fast then. Let me let's fail fast, learn from what we missed was at execution. We didn't understand the marketplace correctly, and then move to the next thing. And I think that that took us a while. That that people change, the type of people we hired, the backgrounds. It took us a while to get a culture where that's the way to do it, right? And then you're not you're not going to be judged on that. You're going to be judged on how fast and you know, the body of work that you launch in the marketplace. Your innovation's been our bread and butter. And I'd say that's because we've been not afraid to jump into things and try things.

Speaker 46:14

But let me remind our listeners, you can find all of our content by simply going to a web browser and typing cpgguys.com at the URL. That easy. If you or someone you know something to contribute to this ongoing discussion on the CPG Guys, please do send us an email at contact at cpgguys.com to our audience. I want to thank you for the clicks, likes, comments, direct messages, meeting us at trade shows, coming to our events, recording episodes with us and our sponsors. Thank you, thank you, thank you. We are grateful for you. The show doesn't exist without all of you. You work with us all you, and we're grateful to have you as your audience and partners. Peter, fun doing this with you again. Give me that big takeaway.

Speaker 3 46:48

Yeah, Shri, it's uh, you know, I think about how much we live in interesting times and the fact that if you uh with all the change transpiring, how we consume media, how we connect with other can uh other people and the people that we trust, how we consider items for purchase, it's just so changing rapidly that if we as marketers don't recognize that that legacy mechanisms of validation may not apply to the new world and we aren't constantly challenging how we inspect what we expect, we're not going to get outcomes that are truly reflective of where growth opportunities are. You've got to be willing to question whether methodologies are appropriate given all the other things that are changing around us.

Speaker 47:43

Yeah, fairly straightforward, Peter. End of validation science history, birth of predictive science. As Bob says, it actually happened a couple years ago. We've seen him a few years in a row at Cagney, kind of using predictive science. To all of those people in Insights listening. And, you know, Peter, you and I at the time of this recording will be keynoting at the Ignite CPG Insights conference on May 5th. I mean, we're gonna take a playbook out of what Bob said. You know, predictive analytics is the way to go, and validation is more a check the box, ego satisfying, and a little bit of CYA, you know, in terms of that's why I did so, because the BC score for this was expected to be A, B, and C, you know. So maybe we can, but the way the world has changed and volume growth has completely metamorphosized over COVID in the post-COVID world where you can't take anything for granted, and the factors influencing the outcome at the shelf, online, offline, are so significant. The traditional syndicator-only mechanisms of validation are a thing of the best, and I'm gonna say actually dead for the most part. So, Bob, pleasure and an honor having you on the podcast. We certainly enjoyed it. Thank you for joining us.

Speaker 1 48:57

Thank you. It was always a dream of life.

Speaker 49:00

That's a wrap of this episode of the CPG Guys.