How Nestlé Gets Close to the Voice of the Consumer

Consumer expectations and choices are at an all-time high. The proliferation of data – from star ratings, to reviews, to feedback forms, and more – mean consumers are hyper-informed before they make any purchasing decision, and are growing wise to where and how they choose to shop. To win the loyalty game, brands must do all they can to gain visibility into what shoppers are thinking – no easy task, given the ever-evolving nature of consumer tastes. They must also develop a robust yet agile data strategy, one that not only gathers key insights, but puts them to work as well. Nestlé is one brand that’s getting it right. In this edited CGT webinar transcript, Julia Ford-Roumani, consumer experience data and reporting manager, Nestlé, and Gautam Kanumuru, co-founder and CEO, Yogi discuss why the brand has prioritized measuring customer sentiment, and the specific tools they’ve implemented to gather meaningful data in a way that serves the needs of Nestlé extensive product portfolio. Ford-Roumani and Kanumuru also nail down how to move from insight to action, bringing their experience to life with a series of case studies and success stories. Read on to learn by example from Nestlé about how to deliver richer consumer journeys and ensure your customers feel heard. Maia Jenkins: Hello, everyone, and welcome to this How Nestlé Measures Customer Sentiment webinar, which is hosted by CGT and presented in partnership with Yogi. My name is Maia Jenkins. I'm an Editor at CGT and I thank you all for joining us today. So now more than ever, having an in-depth and accurate view into customer sentiment is critical for consumer goods companies. Brand loyalty is continuing to fall, competition is at an all-time high, and shoppers today are much, much savvier with access to nearly limitless information about products. And as we'll learn later today, the way people think about and interact with these products is also evolving all the time. So in this conversation, we'll be exploring the ins and outs of measuring customer sentiment through textual analysis and looking at the strategies for getting closer to the voice of your customer. We'll be doing that through shared perspectives from thought leaders in this industry. And specifically today, over the course of the next hour, we'll be looking into why and how brands like Nestlé have prioritized measuring customer sentiment, where you can start if you're just beginning this journey yourself, to take advantage of particular methods and data sources, real examples of how brands are using this information to gain tangible ROI. And we'll also get a snapshot of how customer sentiment technology is changing in this space and some of the opportunities that we have there as well. We invite you to bring your questions to the speakers we have here today and get ready to head into 2023 with a better understanding of what customers are thinking, what they're feeling, and along with the hyper-informed choices that they are making. So before we get started, I just want to remind everybody that this webinar is interactive, and let those in attendance know that feedback is important to us. So please do use the tools that you have at your disposal here to interact, to add questions, send us comments at any time during the webinar. It'll be much more fun for all of us and informative if we make those contributions. Be sure to keep the comments flowing. The webinar today is going to take the form of a conversation between two speakers with questions led by me, but we're keen to hear from our audience as well, so please do add in any questions and I'll pick those up at the end, as many as possible that we get time for. So with us today to talk about this topic, we have Julia Ford-Roumani and Gautam Kanumuru. Julia and Gautam, thank you so much for joining us. I'm just going to invite you to take a moment or two to introduce yourselves and describe your roles at your respective company. So Julia, I'm going to come to you first. Why don't you go ahead and introduce yourself? Julia Ford-Roumani: Definitely, thanks for the intro. So I've been with Nestlé Canada for three and a half years. I'm within their marketing team, and really my key role there is to bring together and aggregate and distill insights from all of our voices of the consumer sources that we have. Really, we're trying to go beyond just transactional data to dig deeper into sentiment and themes, and really understand and get closer to the heart of our consumers so that we can understand them better, understand their journey better, and resolve any pain points and make our products some of the most loved on the market. Jenkins: Brilliant, thank you. And Gautam, you're up. Gautam Kanumuru: Yeah, for sure. Thanks for the intro. Yeah, my name's Gautum. I'm one of the co-founders and CEO of Yogi. So yeah, I have been here since my co-founder and I thought of it in a tiny New York City apartment. My background is actually in the natural language processing and AI and deep learning side of things. So prior to Yogi, I used to work at Microsoft on Cortana and kind of Cortana on Windows, Cortana on Xbox. So I like to tell everybody that if you ever want to hear startup stories or if your Microsoft product isn't working correctly, I probably know the person that owns the feature. So just shoot it my way and I'll do my best to get it fixed. But yeah, excited to be here. Jenkins: Great, thank you. We're excited to have you both here with us today, and we will get time to hear a couple of those stories later on. With that, I will kick off with my first question, which is for you, Julia. So what were the drivers and what were the pain points that prompted Nestlé to first partner with Yogi? And maybe you can tell us a little bit about the background, when did you first partner, and what were the goals? Ford-Roumani: Yeah. We actually attended a webinar and heard about Yogi a few years back now, and really decided to get started with them because we had a couple pain areas and we knew our ratings and reviews data was so valuable and there was so much more potential in it that we were looking to unlock. So for the most part, prior to Yogi, we had just been looking at ratings and reviews in a straightforward, "Hey, this is how many we received. This is the average rating. We get so many four star, so many five star reviews." And the text always was a little bit of a challenge without having text analytics in-house. Especially being part of a marketing team, not a technical team. Standing up a system to do text analytics is incredibly complex and not straightforward, so Yogi really helped us get started with that. I think the other question we were always challenged to answer is, internally at Nestlé we have a benchmark set for our products, that we want to see our products above an average rating of 4.0, which is a great target to have. But then, you always have to step back with any data analysis and really ask yourself, "Compared to what, though?" If Nestlé products are sitting at 4.0 and our competitors' are at 4.5, it becomes a little bit of a problem quality-wise. So really, we didn't necessarily have the competitive data for ratings and reviews, and Yogi has unlocked that for us as well. Lastly, I would say another big challenge in this complex e-Comm world is, and especially with Nestlé, we sell on so many different websites, have so many different rating and review partners, ranging from Amazon, ranging to our own ratings and review programs, ranging even to other e-Commerce sites. So really distilling those and centralizing all of that data into one source and ensuring that you're not duplicating data can be quite cumbersome and takes some time, so Yogi really helps us with that as well. Jenkins: Thank you. Thank you. Yeah, I think that really raises interesting points about benchmarking yourselves against competitors as well as just internally and then separating useful information from distraction or noise when you are privy to such vast amounts of information. So turning to you now, Gautam, how would you say customer expectations have changed to make insight into customer sentiment more of a necessity? Kanumuru: Yeah. Yeah, it's definitely an interesting question. I think if I were to sort of boil it down to just one statement, it would probably just be that customers are now just, they have more data available to them and they're moving faster. And obviously, e-Commerce is a huge, huge piece of this. But if I were to take a step back, I would say there's probably more than just that. If we just look at the fact that everybody now has easy access to things like smartphones and the internet, we're sort of in this age now where information is instantaneously available to us. So if we want to connect with friends or companies or celebrities, pretty much anybody or anything in this world, that's now easier than ever. And then, there's a level of convenience that's also come with that. I think the analogy that I was honestly talking to a cousin about during Thanksgiving is, when I used to check for snow days, let's say growing up, I would have to wake up at 6:00 in the morning and watch TV and watch the bottom banner to see that my county showed up. Now, people these days just wake up to a text message that says, "Hey, school is canceled," kind of thing. And so, that's just maybe a bad example of just the increased convenience and information that people have. But the way that affects brands and manufacturers is that at the end of the day, customer expectations and customer choice is at an all-time high, right? People expect the best from products and they have a lot of data before they even hit the purchase button about whether this product would work or not. And then, literally pretty much every product that I could possibly want can show up on my doorstep within probably two days, but definitely within a week or two. And so, because of that, everyone pretty much has the tools and data to make the right decisions before they go through with a purchase. And so, that's where from a brand and manufacturer perspective, it becomes really important to be on your toes and really have constant sort of data to ensure that you're making the right decisions. And customer sentiment ends up being a huge, huge piece of this. So like Julia was alluding to in the beginning, reviews and ratings is a great example of this, and it's why I kind of said Yogi focus on reviews and ratings is pretty much our core data sources, because that's really one of those prime examples of the type of data that users have at their fingertips before they make a purchasing decision. Jenkins: Yeah, absolutely. Yeah, I think that's such an interesting point about the intersection between choice and expectation and speed and volume of both those elements, I guess, how one really informs the other when it comes to customer journeys and how they interact with products in these really different ways, even pre-purchase, like you say. It's not just about how they're experiencing the product once they've purchased it, but even before they hit that purchase button. So my next question here, it's really for both of you, but Julia, I'll have you take a shot at it first. So how have processes changed as a result of the partnership between Yogi and Nestlé? In what ways has the partnership enabled a greater, more accurate understanding of customer sentiment? Ford-Roumani: Yeah, definitely. I think this has given us a really brand new tool in our arsenal for, we actually run a program internally called Consumer Connections, and it's where we use Yogi in conjunction with all these other data signals that we're collecting about our consumers, and go to the brand teams and really ensure that they're optimizing that consumer journey. And I think, like Gautam was saying, the speed and the efficiency that the consumers expect today, they also expect to be heard by their brands. So just making sure that feedback comes full circle, that a consumer went out, bought our product, and these are now the feedback points or the pain points or the most loved points that they're saying about it, really to make sure the marketers are hearing that and can connect that back to the material that they will put out for future marketing. Because I think just making that full circle just delivers that richer consumer journey, consumer experience, and really can improve the consumer experience overall with the brands. For us also, it really unlocked insights about, as I said, meeting that benchmark versus exceeding the benchmark versus what are the features. So not only stopping at the rating and the traditional method of analysis for ratings and reviews, it's always been word clouds, but what do those words really mean? And going back to even the sentence level insights has really helped our marketers get closer to our consumers and understand them better. Jenkins: Right. Great. Yeah. Yeah. And Gautam, do you have anything to add to that, what Julia just said? Kanumuru: Yeah. Yeah, I think on our end, I would say our partnership with Nestlé has definitely helped push us to scale as well and be as agile as possible. I think the awesome thing with working with a company like Nestlé, and I know Julia and our team talk about it all the time, is you're dealing with categories ranging from ice cream to coffee to frozen pizzas to premium water. So each one of those divisions or teams, they have their own questions, their own needs, their own requirements. And so, it really is on us to ensure that the platform is set up to meet all of those needs while also giving Julia and her team what they need in a timely and agile manner. And I think that's really a critical piece not only for Yogi, but also for the companies that we work with, because as we alluded to, the other thing that comes with this convenience and this more data is that the world just moves faster now. And so, it really is on us to not only be that agile sort of cog in the machine that, hey, if I need an answer in five minutes, the platform is positioned in a way to get the answer in five minutes, but also help to see if there's other ways that we can just, through best practices or what we see kind of in the market, other ways to push brands to move faster as well. And so, that's always been a key piece for us. Ford-Roumani: Great, great. I do think I want to tease out something you said about the different businesses, because that's so important at Nestlé. As Gautam mentioned, we play in the ice cream space, frozen pizza, even right down to infant nutrition. And each of those consumer needs states and use cases are so drastically different, that just doing plain text analysis and getting out the themes is not always so helpful. But what I really find helpful in the Yogi platform is the bucket to custom themes for each area. So for instance, in pizza, we're really focused on, "Hey, are they satisfied with the toppings or the crust?" Versus in ice cream, you want to talk about texture or the usage moments. And really being able to tease out when our consumers are using the various products and what themes they're talking about is really important. Kanumuru: And I think Julia makes a great point. And so, probably one piece of context on kind of the Yogi platform is that; it's just the themes that Yogi is able to kind of go through the text and recognize themes dynamically. So people talk about their frozen pizzas completely different than they talk about their coffee creamers, for example. And so, making sure it's not just one generic, one-size-fits-all, taste kind of theme is not useful to anyone. And then, you also can't do things like word clouds and stuff like that, so that becomes a super important piece for sure. And I think it also gets into the subtleties of exactly like Julia was saying, the questions that the frozen pizza team has is completely different than infant nutrition. And they also run on kind of different schedules: different innovation schedules, different presentation schedules, and all that kind of stuff. So yeah, at the end of the day, I think really what it boils down to is probably two core pieces. One is depth. It's not just, "Hey, this is your average star rating" or "This is how it's trending over time" kind of thing, or "This is what your volume is." It's, "This is why shifts are happening. This is what people are talking about. This is how they're talking about it," kind of thing. That that's a 10 to 100 times difference than just, "We got 100 reviews at a 4.3 average star rating." And then, the second is just to kind of keep saying the same buzzword over and over again is just agility and just getting the information out as fast as possible. Jenkins: What I'm kind of hearing from both of you is a sense of how granular it can really get when you get to gaining these insights and how valuable that can really be across the board. And also, on Yogi's side, Gautam, it seems the partnership has been instrumental in helping you scale and apply the tech to a broad spectrum of products and different categories. So we're digging into some more specifics now. Julia, I'll come to you first with this. What have been some of the benefits? You've kind of touched on this a little bit before, but how are you measuring success? And please feel free to share a couple more examples or anecdotes of the benefits that you've been seeing in Nestlé. Ford-Roumani: Yeah. I think I actually want to dive into two examples that I love to talk about because they're really cool, just examples of how we've taken consumer feedback and incorporated it going forward, and also how it's helped us prioritize or deprioritize certain topics. So one key area that we actually started looking at was our coffee enhancer, AKA our Coffee mate line, and really looking at the experience. And people always add it to their coffee, they talk about whether it curdles or not, whether it's the right level of sweetness. But what we actually discovered through the Yogi platform was in the usage moments, it wasn't just coffee that they were using this with, it was tea, it was hot chocolate, it was other beverages that were somewhat unexpected. And so, we've actually leveraged that insight to play it out into marketing materials. The second example that I like to bring up is I think sometimes when you're inundated with data and analytics and insights, it's sometimes important to know where the fire is and where it's not. And whether a data source is telling you, "Oh, yes, you have to go look at this urgently" or "This is a problem that maybe can be deprioritized." One of the things we uncovered on actually our infant nutrition side, was we were seeing a little bit of a negative sentiment on some of the formula-related reviews. And after digging into that, we actually narrowed it down to an e-Comm and shipping-related packaging damage. So a lot of the times these packages were arriving damaged, and that was really sort of hurting our overall average rating because nobody wants to get a beat-up looking package, especially when it comes to something as important as infant formula. You want to be assured it's safe, it's whole, it's good to give to your baby. So I think one of the things that helped us just scale back the urgency of that problem was really understanding that it wasn't just our products experiencing this difficulty, it was also our competitors. So we looked at Enfamil, we looked at Similac, and they were also having similar shipping experiences. So we did boil it down to an e-Comm problem as a whole, but then really, that allowed us to follow up with our e-Comm partners and have a discussion about, "Hey, is there anything that can be done to improve this?" Kanumuru: Yeah. And just to tack onto that, I think the second example is actually very... Both examples are awesome, but I think the second one also is interesting, given where we're sort of headed in 2023 based on, "Hey, who knows where the economy is going" and stuff like that. But I think the question of prioritization starts to become even more important, right? Because we want to make sure we don't hit the panic button too fast, or that given our limited money and human capital and time and resources, that we're spending that towards the thing that will actually move the needle at the end of the day. And so, just being able to look at that in context, right, of, "Hey, okay. Yeah, we're having some packaging issues, but so is the rest of the market. And we're actually not as bad as the rest of the market, so let's now spend our time looking at something else," becomes a huge, huge piece of just kind of what we see with our customers, but I would say customer sentiment measuring in general. Jenkins: Brilliant, brilliant. Yeah. Those are some really fantastic examples, I think, that really delve into the specifics of how you can really help with gauging what customers are feeling, what they're thinking, what's happening in the market on a broader scale, too, so thank you very much for sharing those. Looking ahead now to the future, Gautam, I'll come to you for this one. Could you give us a couple of insights into what you see as the future of data-driven visibility into customer feedback and voice of customer technologies? Kanumuru: I guess in kind of one sentence, where we sort of see the world going is better versions and processes of what exists currently. And so, what I would say is this whole Omnichannel, e-Commerce world is new to everybody, and so, where everybody's still relatively early in the journey. But the way that we see sort of things play out is that tools, and this is where we constantly push ourselves, are just going to be better at making information easily accessible. And so, it's getting you the exact answer you need at the exact time, even doing that proactively. So it's like, "Hey, we know there's certain sales conversations with retailers in March. Boom, here's the information you need in February to make a good sales pitch," for example. And so, I think that's where things are going. And then, the other thing that we're seeing that I'd really say kind of Julia and the Nestlé team is definitely at the forefront of, is the processes around this customer sentiment measuring, right? So it's "Hey, here's the insight. Coffee enhancers are being used both in tea and hot chocolate and a variety of other beverages. But now what happens?" Right? Then the processes have to be in place so it gets shared with the right people, and the right people react as soon as possible to get that creative content out or the new marketing campaign out, because that full feedback loop or that full process is what truly makes the difference. It's not just, "Here's data" or "Here's data and an insight," it's, "Here's data, here's the insight, boom, business action, what happens next?" And so, we're starting to see those entire processes sort of get built out within our customers as well. Jenkins: Brilliant, brilliant. Julia, did you have anything to add on that or have another... We can move on to the next question. Ford-Roumani: No. I just think the last point Gautam was making, the insight without action, is it really insight then? Right? That's been the biggest thing for Nestlé in our consumer sentiment monitoring initiative is really connecting it back to, "Okay, the data tells us this. Then, so what? What do we want you to do? We do we need you to change about the consumer journey to make sure that we're being seen by the consumers as a responsive brand?" Jenkins: I love that, insight without action isn't insight. That's, I think, a really key point to take from this whole conversation today. So yeah, it really seems like a really exciting space filled with tremendous opportunity, both for consumers and for brands. But if we take a step back now, looking at some of the challenges. Julia, what were some of the challenges you encountered during the implementation of the project? And were any of them unexpected, and how did you overcome these? Ford-Roumani: I think, for me, the biggest challenge at Nestlé has always been the breadth of the portfolio. As Gautam and I were discussing earlier in the call, when you're working with a pizza business versus working with an infant nutrition business, they have very different needs states. We end up serving 10 very unique businesses and hundreds of marketers on each team, really, how you meet each of their unique needs and understand enough of their business, go deep enough, to recommend that action, right? Because as I said, insight, you need action. But you have to understand the business context to be able to recommend actions. So managing depth versus breadth versus velocity, that's been the key challenge all along. And I think, to Gautam's point earlier, tools, I have a stack of tools that we use for consumer sentiment, consumer behavior. Tools are critical to success because they just shortcut the process and make my job much more efficient in delivering those insights and actions. Jenkins: Gautam, were there any challenges that you had encountered? Anything you'd like to add and how you got through those? Kanumuru: Yeah. Yeah, I think the way that I think we sort of look at things is, "How do we be the right partner?" I guess is the way that I'd probably summarize it, where this isn't just, "Hey, here's the data, here's a platform, go have fun" kind of thing. It's truly the space is just moving so fast that it really is about, hey, having that constant interaction, us plugging in in various ways, us even brainstorming together of, "Hey, is this an area where consumer sentiment or Yogi data could be useful?" And then, just working together to experiment. Not everything will work out, but there are kind of the hits that you get that honestly end up making a difference for the business. So I think that's where kind of like Julia was saying before with some of the challenges. But where we look to kind of play the role is we have to make sure we're doing our part to tackle those as well, and making sure that we continue to push the envelope, because at the end of the day, it's about getting results, right? And so, it's not just us standing where we are. It's about us continuing to push for better and better results. Jenkins: Absolutely. Yeah, absolutely. Ford-Roumani: Well, and not just that, but also, Gautam and team has been very willing to pivot. You know that famous Friends episode, where they're like, "Pivot, pivot, pivot." Sometimes it feels like, especially during pandemic times, we've all had to pivot and pivot rapidly. So sometimes, I'll reach out to Gautam and say, "Hey, we need this insight next week or even tomorrow," and really pushing the team to keep up with the speed. So I think it's been a challenge, but I think it's been a well met challenge. Jenkins: Brilliant. Yeah. Your partnership will have taken place over some very volatile times in our industry. So I guess pivoting and that word agility again comes back into it, of just being responsive and willing to change. And I think what I'm hearing as well, is that a lot of... You mentioned not pressing the panic button. If you have those insights, then I guess that that allows you to better balance priorities, like things we need to solve now versus that innovation space that is so important as well, right? And you've been able to work through that together and find a good balance there. Ford-Roumani: Yeah. Definitely. Jenkins: We're drilling down on another challenge that a lot of companies will face when it comes to new technologies. What challenges have you seen in plugging in new solutions into, I guess, more traditional processes? Julia, I'll come to you first for this one. Ford-Roumani: That's a great question, because change is always difficult, right? I think in our program, we've faced the challenge of marketers are inundated with so many different data sources these days. And to expect them to be experts and do their jobs as well as to be data experts and be able to go deep, not every marketer has the time to do that, right? So one of the challenges we found was, "How do we give them the right information at the right point in time and surface that insight?" So for us, the way we've positioned the program and the partnership with Yogi is we've actually had a natural synergy, because we were already reporting on ratings and reviews. But we've partnered with our businesses to have them come to my team with business questions. We don't expect them to know what best source answers that question. It could be Yogi, it could be another source that we have internally, right? But really, we position our team as the data experts who know which source will best meet and best answer that question. Now that being said, you are always going to have marketers or more data-driven oriented individuals who want to go and explore the data themselves. And we've definitely enabled that self-service through marketing dashboards as well. So having both in parallel, where we can have a more guided experience and more of a service model of analytics has really helped us. But then also, in parallel, still offering the self-serve option, I think, has made us successful in meeting the challenge of embedding this new data source into the Nestlé marketing processes. Jenkins: Yeah, it sounds like an arrangement that's really been working well. And Gautam, related to this, how would you recommend or address issues like this within organizational capabilities that, I guess, are based in more legacy-based solutions? Kanumuru: I think as with anything that's kind of a major change or disruption in an industry, different companies and different orgs are at different parts of the journey, right? And so, it's all about just being flexible and meeting people where they are. So that's kind of the way that we see things. And so, it's not just the data and the analysis and the way that the insights are presented, but it's also, hey, the introduction of some best practices, let's say and stuff like that of, "Hey, this is what we're seeing in the organization. But maybe it's a company that's just starting on the journey that just needs the reporting and the data shared out to the org." So that's step one. So we should be talking to them about how some of the more advanced clients have already built processes out for making action from that. And so, for us, kind of the way that we were talking about before, there's the platform and then there's the partnership piece. It's all about just, again, kind of pushing people, where needed, across both of those fronts. But I think at the end of the day, what I would almost say is the important piece is just that brands and manufacturers are pushing towards getting measures on consumer sentiment, or even boiling it down, just being more consumer centric as an org with the decisions you're making and the data that you have. And so, that is sort of what I would boil down as the most important piece. And then, the journey will sort of mature as you take further steps. I think that's also something we've seen with some people that we know who are really excited about pushing this within an organization, where it's tough to jump immediately to step five and have success. Where, as unfortunate as it is, I love moving fast as well, but sometimes you really do need to do, "Hey, maybe we can do step one and two and then let that boil a little bit, and then do step three and four, and then let that boil as opposed to jumping to the end." Jenkins: Right. Yeah, absolutely. Meeting people where they are is so important, as you say, and remaining open to, I guess, what's realistic and what's possible within the realm of possibility, absolutely. So just before we open things up to the audience for questions, I have a two-part question for both of you. So the first part of the question is for you Julia, which is what best practices can you offer to companies starting out on this journey into using these custom insights to strengthen brand loyalty? Ford-Roumani: I was just going to say, I think really to start on the journey, what's really important is to understand all the different data sources that you know that you have you're collecting about your consumers, because there's a lot. And to really take those and map them back to the consumer journey at the different points, which different data signals do you get? And then from there, map that to, "Hey. Well, what action are we hoping to derive from this? What are we really looking to optimize? Is it the beginning of the journey? Is it the end of the journey? Is it the whole journey?" And really break it down into, as Gautam said, baby steps. We're often tight for analytics projects. And so, making it as easy as possible to have some achievement, and then deliver value is I think super critical to getting started. Jenkins: Fantastic advice, that's brilliant, yeah. And Gautam, coming to you, how do solutions like Yogi's help with this? Kanumuru: Yeah, it's a great question. I think it really is giving the data that's required in the best format and in a timely manner. So that's probably sort of a summarization of some of the points that we've made before. But it really is, at the end of the day, I guess the way that I would say it is for brands, for manufacturers like Julia's team, there's a problem that is trying to be solved, right? It's like, "Hey, we need to know more about our customer and take this into account for a certain decision." So really, it's on tools like Yogi and other ones to give the data that is the solution, but also give it in such a way that Julia and her team don't have to spend too much time on the solution piece. They have the answer, they can move on to the business action that that's required. And so, that's one piece. And then, the other is just being as much of a resource as we can to help these brands on the journey that they're going, right? I think one advantage that we have is we see just across categories, across companies, across even the world, how different people are approaching this and just bringing the best practices as often as possible. Jenkins: Wonderful. Thank you. Thank you for that. Yeah. Well, thank you for answering my questions today. It was a really fantastic conversation. And before we move on to some questions from the audience, do you have any other final comments from either of you or we can move on. We have a couple questions in the chat to pick up as well. Ford-Roumani: Yeah. I think I just want to add one more point to Gautam's question is, so foundational to any data program is curiosity. And I think one of the things that we've still continued to work on here is even if we're not sure of the answer, let's challenge ourselves and ask the question. Just always be asking questions has been critical to really changing our understanding of the consumer, I would say. Jenkins: Brilliant. Yeah, absolutely. Always be asking questions. Thank you. Okay. If we are ready, I think we can move on to the questions that we have from the audience here. I'm going to start with a question that has come from... Yan Furhman at P&G is asking, "Is Yogi doing the text analytics in a supervised or unsupervised way to identify topics?" Kanumuru: I would call it semi-supervised, is actually the way that it gets implemented. So I think kind of the best way to think of the way that we do sort of theme analysis or topic analysis in Yogi is our AI is going to go in and find all kind of distinct groupings of conversations. So, "Hey, here's a bunch of people talk about one topic, another topic," so on and so forth. And then, what we have the advantage of is because we're analyzing across hundreds of different product categories, we can sort of map things and be like, "Hey, this seems very similar to a topic that was talked about in a different category." And that is called topics, for example. And so, that's where the semi-supervised piece comes in. And as new sort of topics or themes pop up, we have the ability to kind of label them accordingly. So yeah, because of that, I would kind of boil it down to semi-supervised, is the way that we look at things. Jenkins: Okay, great, semi-supervised. And we have another question in here. Let's see who this is from. This is from Mike Blythe that Havelock Holdings, who is asking, "What are the top three decisions both Julia and Gautam recommend in terms of speed to action, time from standup to real action?" Julia, would you like to take a crack at that one first at all? Ford-Roumani: Yeah, I can definitely take a crack at it. So the good thing is at Nestlé, we do have a little bit of a culture of experimentation. I think you have to be willing to take a small risk, start small, prove either the theory will work or won't work, and then fail quickly. In our case, we didn't fail, which is the good news. We actually had very good success. So I think just be willing to experiment. In terms of our timelines, actually, Gautam and team that were really great. I think our biggest struggle was narrowing down this portfolio of SKUs we wanted to input into the platform for analysis. And so, really within a month we were set up and running and having insights delivered to the brands. Is that fair to say, Gautam? Kanumuru: Yep, for sure. Jenkins: Wow. Wow. That's really quite amazing, amazingly fast. Gautam, would you like to add anything to that question there or has Julia covered it pretty much? Kanumuru: Yeah. No, no, no. I think Julia definitely covered a lot of it. I guess the only thing that I would add is when it comes to kind of from time to stand up to real action, I think the way that we've kind of seen this work well in companies is you sort of think of what ideal state is and work your way backwards. So it's like, "Hey, we'd love to get to a point where boom, we find an insight and the action is made within 48 hours or something like that." But you start to work your way backwards of, "Okay, ideal state." So that's ideal state. But maybe what's more realistic now is, "Okay, we have one or two people who are responsible to find, once the insight is found, to push it to the right team and just track what change they make." So I guess the way that I would kind of boil it down is you could just kind of think of it as, "Okay, what is the quickest way that we can build the insight-finding piece and the quickest way we can build something that will push it to real action?" And just having somebody be responsible for that is, I think the first piece. I think the other part of the question was also kind of getting senior management acceptance, which I think is a huge, huge piece. And the thing is, it's all about quick wins and ROI, I guess. And so, it really is just really pushing for, "Hey, we're going to find a thousand insights from this new process, but what are the ones that can be sort of a short-term change that you can very quickly point to something like, "Hey, this was the change that was made and this is the action that came out." That's one. I think two is just locking into trends and what other companies are doing as well. Everybody knows that e-Commerce is important and growing and is going to continue to be a huge priority for companies. So just being able to tie how, "Hey, us analyzing customer sentiment more is going to allow us to make better decisions in this new world." And just being able to tap onto, "Hey, this is an example that we've seen from another company of how this plays out." We've generally seen those are the two kind of pieces that become useful from a senior management perspective. Jenkins: Wonderful, thank you. Yeah, absolutely. I think from a senior management leadership level, like you say, ROI and quick wins will always be popular, right? So this is my last question here from the audience, and this is one for Julia. It is from Mihir Shinde at J.M. Smucker, and, "How do you quantify the extent of themes you find in the reviews? And what does the rectification/change pipeline on the Nestlé end look like?" Ford-Roumani: It's a great question, and it's a complicated answer, I guess, to say the least. So let's start with the extent of themes we find in reviews. I think back to my point earlier is really you can't look at any one data source in isolation. But, for instance, if there's a theme that's popping up in reviews that's also popping up in our consumer contact center as well as on social media, then it's worth digging into that a little bit more and really evaluating what the total impact is and trying to quantify that, and going back to matching it to maybe sales and seeing sort of complaints per million. There's a lot of different data blending exercises that we do here to really make sure that we're prioritizing and looking at the right urgent topics. The rectification and change pipeline on Nestlé end, that's an interesting one because obviously in the CPG space, product pipelines and changes are planned typically years in advance, two to three years. And so, we try and influence what we know we can change. So I'll give an example where maybe consumers felt preparation instructions were lacking for a pizza or that they were burning their pizza. What we do is feed that into where we can change relatively quickly, for example, content on our website to help our support hub. So if customers go to our website, they can easily search a topic and get a support article very quickly. So I think it's just about knowing what you can change and what you can influence and where is the right point in the consumer journey to address the issues. Now that being said, of course, if there's an issue coming up in ratings and reviews or our CES team that requires urgent attention in the nature of a recall, obviously we take media action and address those concerns with the highest sensitivity. So I think it's just making sure you don't miss the forest for the trees and looking at the patterns from across multiple data sources. Jenkins: Thank you very much. Thank you for sharing that example and again, reiterate that point of separating important signal from noise and reacting appropriately, I think is what one of the big takeaways I'm seeing here. So we're just getting ready to wrap up. Thank you to the audience for all these challenging but insightful questions that we received. And I'll invite Gautam and Julia, any last thoughts before we draw the session to a close? Kanumuru: Thanks again everyone for attending and the questions. Yeah, I think on our end, really the way that we see it is this is not a one-year journey, not a three-year journey, this is sort of a long-term thing. And I think really at the end of the day, there's going to be multiple steps. There's going to be challenges. There's going to be a lot that everybody in this industry is going to learn. But I think really just getting started and getting the ball rolling becomes super important and focusing on kind of the foundation to begin with: getting those processes in place, getting the right tools in place, getting the right partnerships in place. I think we've seen the fruits of that play out with kind of the companies that have started on this journey a few years ago. And so, I guess that's kind of one big takeaway or closing point I would say on the Yogi end. Jenkins: Thank you very much. Ford-Roumani:I don't have anything closing wise to add, just to say thank you for hosting this webinar and the conversation has been great. And also, thank you to Yogi for being a great partner. Jenkins: Yes. Thank you both for such a great conversation. Thank you to our attendees, and thank you to our sponsor today. And this is just a quick reminder that the recording of the webinar will be available within two hours for on-demand viewers or anybody who wants to catch up or catch up on anything they missed. And I hope you'll glean some insight today. I know I certainly did. And I wish you all a great holiday season, and have a lovely rest of your day. Thank you again, everybody.

Related Topics