Transcript Disclaimer: This transcript has been generated using automated tools and reviewed by a human. However, some errors may still be present. For complete accuracy, please refer to the original audio.
00:00:00 GOVIND ETHIRAJ: You know Nestle India as the maker of Maggi noodles and Kit Kat chocolate wafers, India is now the largest market for Nestle's instant noodles and soup brand Maggi and the second largest for chocolates. Now there is of course coffee as well and a host of other products in the dairy space. The company is launching new products or extensions all the time and thus creating a repository of data which it hopes to better mine to understand what to serve customers present and future. So the challenge and opportunity before companies like Nestle and the leadership is in first capturing and using that data and then the intelligence over it to reach those customers of today and tomorrow, including via newer distribution channels like Quick Commerce and second to fuse all of its existing product and brand strength with advanced technologies almost like re engineering the organization. My guest to talk about this journey and transition from Nestle's perspective is Suresh Narayanan, Chairman of Nestle India. Narayanan joined Nestle in 1999 or 26 years ago, holds a master’s from the Delhi School of Economics and is a career FMCG or fast moving consumer goods veteran, having worked before that with Hindustan Unilever and Colgate Palmolive.
00:01:14 GOVIND ETHIRAJ: Mr. Narayanan, thank you so much for joining me. So I'm going to talk to you about consumer products but from the context or lens of artificial intelligence in specific and technology in general. So is there a point that as we've gone through these last two years when we've looked at how AI is entering our lives and maybe transforming some of it that you said, oh, here is something that maybe can address a problem that I've been grappling with or we are grappling with as a consumer products company at any point, whether it's at the production, the sourcing operations, supplying distribution and here is maybe what it could do.
00:01:47 SURESH NARAYANAN: A very good question Govind, thank you for having me. First, you know one of the challenges of a large consumer company is that there are literally oceans of data that we sit on. You know, consumer data, customer data, manufacturing data, sourcing data. And one of the challenges that we faced as a company was what do we do with all this data and how can we make the data work for us in terms of information and prediction. And this is what excited us with AI, the usage of consumer and customer data in order to create predictive models on demand forecasting production, forecasting sourcing requirements. This is where AI really takes on a new life. And that is what we are working on. And I think, I think I one of those believers. Goindra it's easy to spell the doomsday scenario for a technology. I think AI has got so much to offer in terms of positive aspects of operations. Imagine if I'm able to forecast better. A 1% improvement in my forecast capability will multiply itself in terms of my efficiency and profit delivery. And that I think is what is the exciting part of this whole journey.
00:03:12 GOVIND ETHIRAJ: So let me pick two commodity linked products. So coffee and cocoa, both of which are critical to your operations and success and both of which prices have been going up and down in the last year. Thanks. For various global reasons and maybe climate reasons. So how would you apply it here?
00:03:27 SURESH NARAYANAN: Look, I think, you know, India especially let me take coffee. While there is a global market which is much larger than the Indian market, but India plays a significant role in the exports of coffee. Now every time we go with data and forecasts of events that have already happened. So we know that the Brazilian crop for example, is coming in a little bit weak. We know that the weight of this crop is affected. We know that it creates a price impact that is going to be significant. In fact, you would have read today's newspaper that's talking about 50 year highs as far as coffee prices are concerned, which is going to pose a threat not only to the industry in terms of conversion to coffee because it will make the end cup so much more expensive, but also threaten the livelihood of farmers. Because most of the farmers who are growing, the 12, 13 million farmers across the world who are growing coffee are all farmers, many of whom are on the borderline of poverty. So they are not the affluent farmers. Now, if I am able to use AI predictive models, I will be able to determine as a company when should I intervene into the market for buying my coffees. What kind of hedging policies should I have, what kind of stocking policies should I have in a proactive manner, which means that six months before the event, if I am able to put through my algorithms that gives me the benefit of purchase value, stocking consumption, I'll be so much the better. I'll be able to give not only myself the benefit of better sourcing, but hopefully the consumer the benefit of a cheaper cup of coffee. That is ultimately for me the golden grail where this technology can work for us.
00:05:18 GOVIND ETHIRAJ: So if you were to take a step back, or I were to take a step back, you're saying that the larger challenge for you and the potential application for technology is before the product leaves the factory, so to speak, at this point, yes. So I'm going to come to post factory as well. But you're saying as a as from a top level today for a large company like yours, and as someone who's leading it, you're saying this is where.
00:05:39 SURESH NARAYANAN: The real, this is where the procurement. Because you know, if you look at my P&L, more than 50% of my P&L is cost of goods. And that is entirely on three or four large commodities, right? Milk, wheat, oils, coffee and cocoa. Now any benefit leverage that I get here in terms of predictive models or in terms of interactive models, I'm able to get that much more benefit in the end stream. And the, the money to be extracted or the value to be extracted is before it leaves the factories. Because after it leaves the factories, then there are channels and there are channel plays there where also there is a clear use. And one of the big use cases that we are establishing now is on the demand forecasting and integrating. The dream that I have for my supply chain folks is very simple. It's almost like the princess and the pea. That 20 mattresses and 20 feather beds and one pea below could still affect her sleep. It's the same that we are looking at for our products. That a small variation in a offtake of a brand in a particular shelf is able to trigger production and procurement plans across the factories. That I think will be the kind of sensitivity and there is value to be extracted. We all think that we have extracted enough values from our value chains. I mean as a company, each year we extract about at least 1 1/2 to 2% of value from existing operations. Imagine if I use AI and I've used predictive models, how much more I can get, right?
00:07:30 GOVIND ETHIRAJ: And given that how unpredictable things are on all these factors. You talked about milk prices, cocoa prices, coffee prices, wheat. And you know, we're again looking at a wheat shortage right now because of again unpredictable. We've had a low rainfall for winter rains and we were looking at a shortfall in wheat crop. What is it that company the best, I mean the sort of the best of the class companies or people who are in this space. What can they do to maybe better anticipate? I understand the predictive part and this is what your goal is, but what could change? Because it does appear that some of this is just beyond our control today.
00:08:04 SURESH NARAYANAN: No, look, I think a link between AI models and agronomy practices. For example, at the crop stage in for example coffee or cocoa, the number of pods or the number of beans per plant that is being generated, Is there a variation in that? If that Data is captured early enough in the season, you will know that this year the flowering has been lower, for example, in coffee. So you know that the output is likely to be a little bit lower. Today. We know it after the event has happened. I mean, there are estimates being made. I won't, I will not be, I will not be exaggerated to say that nobody knows it, but the extent to which we need to know is much less. Similarly for wheat, right. If I know today there are drone technologies that are available that estimates the crop and estimates the kind of approximate yields that we will be getting. If I were to integrate that with AI models, I'll get a much sharper picture on wheat, on milk. I mean, we deal as a company with 100,000 dairy farmers in the Punjab. Now if using data on cattle and cattle yields, if I was able to predict what is likely to be the yield of the farm in the coming season and what is likely to be the impact on milk prices, I think I'm so much the better off. So there are multiple use cases for technology. The question is to link it to the source. I think once you're able to link the algorithm and the data source to the source of the issue, you're able to do a much better job.
00:09:45 GOVIND ETHIRAJ: And it's interesting because you're saying that basically we still have to work on the data points themselves even as we figure out the mining and the AI on top of it. And we have a lot of opportunity to do that, including in India in all these areas that you spoke of.
00:09:57 SURESH NARAYANAN: Look, I think one of the biggest challenges, Govind, and you'll hear this from many people, the challenge of AI rests on the quality of the data. Data integrity and data quality has become so important because it's garbage in, garbage out. So if I put in suspect data, I will get predictive models to work on that algorithm because it's on the basis of learning. So it learns the wrong stuff. It's almost like children. You teach them the wrong values, they turn out to be lousier citizens as compared to children who are taught the right values. It's the same with these models. And there is no gut feel there. If you are true to the AI models, you will accept what they throw up because your brain is too feeble to be able to do those kind of computations. So that's where the challenge is.
00:10:47 GOVIND ETHIRAJ: Right? So let's talk about now, post factory. What are the kind of opportunities that you see for technology and specifically AI?
00:10:54 SURESH NARAYANAN: Look, I think the, if you look at the conversion cycle, there is the procurement cycle. There's a storage cycle, there is a manufacturing cycle, there is a logistic cycle, there is a delivery cycle and there is an offtake cycle. So after the factory there is a logistic cycle and there is a delivery cycle and there is an offtake cycle, typically in any consumer goods thing. Now in logistics today, we are using techniques which are able to give us considerable savings on logistics costs by better modeling of the predictive demands and these supplies that we have to make in order to make it happen. So movement from factory to the distribution centers. Today we've got models that are available to us. We call it the T hub, like a transportation hub. So this itself, I can tell you, in the last year or two years has given us almost 400 to 500 billion rupees savings in logistics costs by just planning the movements well and by routing it well.
00:12:04 GOVIND ETHIRAJ: Can you illustrate this with one of your product for example, Maggie Noodles.
00:12:06 SURESH NARAYANAN: Right. Maggi Noodles. The average ton, kilometer as it is called, so 1 ton. The average distance that it traverses from the factory to the source today is about 900 to 1000 kilometers. Using the algorithm of source factories, manufacturing plants and distribution center feeding, we are able to reduce it to about 6, 700 tons. 6, 700 kilometers, which means 30% reduction in the kilometer age per ton of product will yield logistics savings. And that logistic savings when computed on an annualized basis comes to about 450 to 500 million. So this kind of algorithms can be worked in in the dispatch routings from the distribution center to our distributors. Again, similar savings can be done. More importantly, if you've got better predictive models, you'll have better stocking models. One of the biggest challenges of consumer good companies is because of the multiplicity of SKUs. Our complexity is high and therefore we have multiple stocking of SKUs. And that multiple stocking of SKUs is working capital. So I'm blocking up working capital. For example, if an item has a salience of less than 1% or 2% but it is 5% of your stocking, it's a waste of money. Better predictive models and better control on the demand forecasting is able to therefore reduce the stock levels, make it appropriate to the rotation of the stocks, thereby minimizing on working capital, minimizing on financing costs.
00:13:59 GOVIND ETHIRAJ: But is that something that's shifting or is there a sense of nonlinearity in demand today? I mean, using Maggie Noodles again, because one would assume things go steadily.
00:14:09 SURESH NARAYANAN: No, I think today's model makes it even more necessary to get in third technology into it because demand is no longer linear. It is very fickle. Earlier we used to only face the challenges of competition and seasonality. Today is more than that. There are multiple category interactions. I decide to eat more of chocolates and wafers and something else. I will reduce noodles. If I drink more of tea and drink more of, of bubble tea and stuff like that, I will reduce the consumption of coffee. Now all this, somebody has to be able to put the maths together and do the sigma and say this will have an impact of 1% on your coffee sales today. We know it post facto after the event, we say why did it happen? Then we say okay, because of this, because of this, because of this. But if you're able to predict it before, it is more challenging. It is more difficult. But I think because of the nonlinearity, because of the volatility and uncertainty of demand and of consumer behaviors, over a period of time, it becomes that much more challenging, necessitating the need for technology and the need for algorithm.
00:15:30 GOVIND ETHIRAJ: Right. So let me go further down the distribution chain. So one of the things that's obviously changed in the last few years is quick commerce. And you yourself are attributing a growing percentage of your sales to quick commerce and commerce. Commerce is larger, of course. So how is that changing? Some of the things that we've just.
00:15:47 SURESH NARAYANAN: That is, that is doing two things. One is quick commerce is, is today more than half of my e commerce business is quick commerce. Tables are turned on the traditional ecom players today. The, the Blinkets and the Zeptos are much bigger than many of the other traditional ones. Two things are happening there. One is that demand and the response of particular brands and SKUs tend to be much faster and stronger on E commerce because the turnaround time is very short. I do an introduction within a week. I'm able to know exactly what is likely to be the offtake of this brand going forward. Or what is the, what they call the daily run rates. Right. So the daily run rates, if it's moving in one direction, you know, it's kind of, it helps. The second thing which is happening is because of the opening of the dark stores by the quick commerce players, they're also now they're saying, and I think it was posed by one of our customers to us, he said, I am delivering in 8 minutes and 10 minutes and you still take 48 hours to deliver to me. Is that even fair? Right. So we put a challenge to our supply chain team saying how can we use their Information and our capabilities and our factory locations and logistics to be able to do a faster job of turnaround times of orders to quick hours. Today I'm able to do it in 24 hours. But the challenge to the team is to reduce it further to maybe part of a day. Because if they are delivering in 8 minutes and 10 minutes, I think in order to sustain my business and to grow it is no longer going to be just brand salience, it is going to be contact efficiencies that matter.
00:17:36 GOVIND ETHIRAJ: But you're also sending out much smaller lots now in order to. So how does that stack up with the earlier sort of distribution and logistics infrastructure?
00:17:45 SURESH NARAYANAN: Look, I think there's a mindset change now. The mindset change is.
00:17:49 GOVIND ETHIRAJ: And the cost of it, I mean.
00:17:50 SURESH NARAYANAN: That’s the cost of it. So that is where the whole optimization exercise happens. So we try and optimize much stronger on the volume business in order to pay for the disaggregated business. That's how the model works.
00:18:07 GOVIND ETHIRAJ: So there is a loss somewhere, there.
00:18:09 SURESH NARAYANAN: Is a loss somewhere, there is a loss somewhere. But that is the growing segment that's, you know, because I will move goindraj ultimately where the consumer goes, if the consumer says I want my products in 8 minutes and 10 minutes, I'm not going to be stuck there saying that look, I will give it to you in 24 hours or 48 hours. And I think that's where companies are looking at ways and means. The whole distribution chain, the value chain, the number of vendors, the number of intermediaries in terms of taking decisions org structures are all getting redefined and I think with full blown AI coming in, they'll get redefined even more. I still don't know what the end point is going to be. All I can say is being fast, being focused and being flexible are going to be three behaviors and being fungible are going to be four behaviors that are going to be essential for any consumer good company to survive.
00:19:05 GOVIND ETHIRAJ: So just to ask you a sort of present day question, we've obviously seen a shift in the nature of the market. Some people have been telling me that this is including from your own industry that this is not something they foresaw. The slowdown in demand, for example, in the last year or so. I mean the role of technology is something that I'll come to later. But my question is really is there something so fundamentally changing in the market that either through intuition or technology we find it difficult? And therefore how does one look ahead? Let's say 2025 and the next couple of years at the market and the consumer.
00:19:38 SURESH NARAYANAN: Look, there are two or three problems, two or three things that are happening at the same time. And let me just disaggregate it for the sake of simplicity. The first is economic issues. Food inflation, inflation, unemployment, limited real wage growth is stunting demand to some extent, right? That's one part of it. The second part of it is multiple category interactions. See earlier the consumer space was defined by choice and, and category interactions were limited. So somebody ate a noodle or ate a chocolate, he usually ate that, right? And there was not very much else that put play. Now there are multiple categories that are coming in. So when I make a choice of eating a Kit Kat, I make a choice of not eating something else. And there could be five or six different categories that I might not touch. Similarly, if I don't eat KitKat then there are so many other categories that I might touch. So the capability of the marketeer or the capability of a commercial organization to seamlessly understand the interfaces that are taking place is also an important issue of demand. And that is where some of us who are legacy brand owners are trying to make changes. So I'm trying to make changes in all my offerings because I can't take it for granted with the new age consumer that because I'm KitKat or because I'm Maggie, that I'm the salient one. Because they'll say yeah, but this is like as old as the hills. We like it, we love you, but there are other options as well. And the third thing is, I think which is happening slowly but steadily, a lot of things are being attributed to D2C. There is probably some bit of truth in that. But also the category wise consumption spends are also changing Today, for example, travel and leisure has become a much larger part of the Gen Z budgets as compared to what it was in my generation. You know, in my generation we went for, all day once a year for 10 days and largely went to your parents home and came back. It was considered to be a holiday today every four weeks. Want to take a breather, a break? Right. Obviously the money has to come from somewhere so that money will come out of the discretionary items that is there on your, on your budget. So these are two or three things how this will play out. I mean I'm, I'm not saying that, you know, India is still a consumption economy. Some things will get consumed more than others. The question for, for, for consumer marketers is how relevant am I, how differentiated am I, how Sustainable am I. If I answer these three questions in the positive in terms of my offering, I think I still have a business.
00:22:31 GOVIND ETHIRAJ: Okay, last question. So what's the. Is there a new product that you would like to bring to the market in the next, let's say year or so or category, if you can talk about that.
00:22:40 SURESH NARAYANAN: Look, we are, we should perhaps address some of these things. We are on the throes of launching the Nespresso boutique in India. Yeah, right. Premium coffee.
00:22:51 GOVIND ETHIRAJ: Finally, as someone, some people would say.
00:22:52 SURESH NARAYANAN: Yes, it's a premium coffee. And you, of course, I don't have to explain to you, I think I'm excited by it. I think India is premiumizing quite rapidly and that is where the play for Nestle will be the strongest. So more premiumization across my categories. Coffee is just one example. It's happening in milk and nutrition is happening in, in foods, is happening in, in chocolates and confectionery as well. So I think that the, these are exciting times. And the second is I think the digital capabilities of the company to link with the consumer. I think today the role of influencers is much larger than the role of television advertisements. Right. And this generation is moving to, in, in in sync with companies that have got good quality brands that do good for society and that ultimately resonate with them. And I think that is the exciting journey that I see ahead for consumer companies in this country.
00:23:48 GOVIND ETHIRAJ: So supplemented to the last question, so the brand message, if there is one, it'll obviously differ for every product. Do you see that shifting? In some ways the brand message will.
00:23:58 SURESH NARAYANAN: Never change Govind, because it's a bit like you having the name Govind Ethiraj and me having the name Suresh. It will not change. But what we stand for could change a bit. What we stand for could change a bit. So if I was talking about two things in the past, I'll talk about three things now. Third thing would be, for example, the societal impact or the environmental impact of what I do. I think these were not questions that were so germane to the brand choices some years ago. They're becoming more.
00:24:28 GOVIND ETHIRAJ: And they matter to the consumer. Mr. Narayanan, thank you so much for speaking with me. Thank you.
00:24:33 SURESH NARAYANAN: Thank you very much.
