Context is Everything: Harnessing Data to Tap into Today’s Consumer Trends

Whether its startups stealing share or consumers flocking to new purchase channels, the consumer goods industry is facing more competition than ever before. Among the many things CG brands need to juggle are demand planning fulfillment, trade fund investment and DTC promotions. And did we mention there also happens to be a pandemic occurring that’s wreaking havoc on, well, just about everything? With all those balls up in the air, now may be just the right time for consumer goods companies to evolve and refine their data strategy to achieve greater wins across their entire enterprise. In “Create a Roadmap for Success in a Data-Rich and Digital World,” presented by CGT and sponsored by EdgeVerve, we discuss the importance of harnessing data to tap into today’s meta trends, as well explore the emerging tech that’s required. The full transcript and presentation slides from the conversation are below. Alarice Rajagopal: Good morning everyone, and welcome to our webinar. Please note that while the materials in this webcast have been reviewed by our editorial staff, the views of the speakers and their organizations are their own and do not necessarily reflect the opinions of CGT or its owner, EnsembleIQ. Our first speaker for today will be Praveen Kombial. Praveen is the global product head for business applications at EdgeVerve, which is a product subsidiary of Infosys that enables its customers to traverse the digital transformation journey through IAI and automation. Praveen's mandate is to leverage artificial intelligence and emerging technologies to build and take to market products solving for specific business problems for industries globally. At EdgeVerve, Praveen leads a highly talented team of engineers, product managers, and solution consultants in building an ecosystem of products addressing industries and functions such as supply chain, financial services, CPG, and more. Prior to Infosys, Praveen spent many years at GE Capital and GE Money in India in different business functions and roles across retail and commercial finance and risk management. Joining Praveen today is George Lawrie. George is vice president and principal analyst at Forrester, serving application development and delivery professionals. He brings to Forrester more than two decades of experience, deploying global enterprise resource planning applications and complex multinationals. During his five years with Forrester, George has led research into topics such as SAP deployment best practices, ERP consolidation, IT investment prioritization, global data synchronization and trade promotion management. Our last speaker will be Suresh Bharadwaj, AVP lead product manager for TradeEdge, EdgeVerve. Suresh has vast professional experience in product management strategy and business process consulting in the retail and CPG industries. He has led successive technology startups to significant value by creating globally recognized products and platforms in the areas of warehouse management, logistics, and sales execution, including as a co-founder of Manhattan Associates, a global supply chain solutions company. He is currently focused on driving innovation within the consumer goods value chain. Suresh, Praveen, and George, thank you all for joining us today and with that I'd like to go ahead and hand things over to get started. Praveen, go ahead, the floor is yours. Praveen Kombial: Thank you so much Alarice. Thank you for the invite and all of us, Suresh, myself, and George we're delighted to be here. So look, I'm just going to kick this off and then really hand it over to George. You've seen some numbers out there, it's just a reflection of our world is changing, and probably for a change I'm not talking about COVID. The consumer goods industry was in the midst of a systemic shift any which ways, and if anything the pandemic has only accentuated the shift. Retail growth is shifting to online channels. In an effort to personalize, new product introductions have ramped up significantly across all CPGs. Meanwhile, disruptive startups are trying to build very deep consumer relationships, and maybe just one category in beauty and grooming last year had about 70 new startups was some research that was done. With the change in buying behavior, now the media mix is changing too. Traditional media expenses are going down, stunting growth in digital media strength and that is already about 50% share. We all know the challenges of measuring effectiveness there. So really amongst all of these trends, the one constant is that the data ecosystem is becoming even more complex. The webinar today with Principal analyst George at Forrester and Suresh and I from EdgeVerve, is really focused on trying to trace the importance of harnessing the data to tap into these meta trends. The emerging technologies that are needed, a maturity framework on channel data and kind of a journey map to try and navigate these trends, so let's gets started. It's my pleasure to invite George, so George over to you. George Lawrie: Thanks Praveen, and thanks everybody for coming to listen to this session, in which you heard about the challenges and you know them of course, you're living through them. But what we want to do, is to try and give you some idea about how to create a roadmap for success in this digital economy. I think you probably guessed, it's really all about the data, it's about managing data. Now I promised not to mention COVID but here it is. The direct-to-consumer was already a booming market before COVID came along. Generally speaking these kind of tipping moment, or critical events like this, they actually strengthened trends which are already underway. We're seeing more and more of this, that, as already mentioned, the mix of the marketing mix now has to have digital elements in it, but so does fulfillment and so does that relationship directly with the consumer themselves. Not too surprisingly, we see people are using their smartphones, mobility is very important, this is creating a vast wealth of data. The company that's able to understand this data and to exploit it is the one who is best positioned really to capture these new kind of we call them mobile moments or new moments of truth as they used to be known. So the interesting thing is that this data is turning everything upside down. The important thing is to be able to figure out how to exploit it, how to use it. One of my clients said something very perceptive. She said, “You know, the consumers don't expect you to meet their demand; they expect you to anticipate it.” You have to anticipate it by looking at all the places where they are and mobile is one of them. But actually their use of social media, their search engine arguments, all those are clues that help you to anticipate what they really want. Now, it's not too surprising that, particularly in the U.S., people have had a very strong relationship with brands and with consumer packaged goods companies, and it's why they always used to be called national brands. They've known about them for years. If I go back five, six, seven years, we have loads of data that showed they would always check the manufacture's website before they bought something, which is interesting because you'd think they know everything about it already. But sometimes they were looking for coupons, sometimes they were looking for new ideas and quite often they were trying to understand the innovation that, that CPG or that brand was driving. Here you can see, that U.S. online adults in general are interested or in fact very interested in buying directly from brands or from CPG manufactures, and that I think is a very interesting thing and it's slightly counterintuitive. You can see in other parts of the world, there's more and more interest in private label. But actually in the U.S. certainly, they're interested in the owner of that brand or that label. That's not to say there isn't a terrific amount of disruption going on, you already heard Praveen talk about it. We all know the story of the so-called billion Dollar Shave Club, and McKinsey and others have been famous for saying, actually the real risk is for new upstarts who understand how to use this data, and understand how to drive that relationship and that personal experience and also the personalization of the product or service offering to be able to capture these consumers maybe to get them on a subscription basis. Gradually to build more and more revenue to the detriments of traditional CPGs. So those disruptive brands I think we all know have been challenging the status quo. But it's not to say that the status quo isn't fighting back. Even if you look back three years ago, this is one very large company that aimed to build on this idea of a one to one consumer relationship. So we hear people talking about this, a market segment of one, which is addressable with digital technology. But more and more that's also affecting the way in which we will go about fulfillment and one day, the way in which we'll go about the manufacturing as well. Many of you will have seen examples, perhaps of a customized pet food for your own dog perhaps according to his condition, and it won't be too long I'm sure before we're doing this for people as well. Some of you may remember Coca-Cola's brilliant experiment with printing people's names on the side of the can. Of course, it wasn't actually personalized to you, but it gave you the idea or the direction in which they're going. So that said, we've got a lot of disruption but we've got established companies also thinking: How are we going to use data to drive this new kind of relationship? The one-to-one relationship, the market segment to one and one day, perhaps a batch size of one. We can already see it in some types of manufacturing. So Harley-Davidson, for example, has probably about 6,300 different combinations, and you're able to create those combinations yourself online and then to watch your own motorcycle being assembled. So you can imagine the brand relationship and your relationship is not a product, it's something that you already created and that will happen more and more with personalized recipes and personalized solutions to problems. Now the important thing is that using this data and becoming a direct-to-consumer disruptor, it's all about understanding a latent consumer demand. So who knew previously that there was a demand for things like pet food on subscription? But the first people to understand that, to exploit it, made a much tighter relationship with the consumers. If we think about that, we think about how will we build that direct-to-consumer relationship? Well, we have to start by, really I would say, segmenting or thinking about the type of brand or CPG that we are. We think about this, I suppose all analysts do, in terms of a matrix. We think about the prestige, the desirability of the brand and we think about the frequency of purchase. This has an impact on the data that we collect, but also the way that we relate to those customers. You'll see in a minute, it also has a relationship to let's say what kind of revenue we can also expect from this digital channel. It's important to remember that consumer packaged goods companies and brands generally, they're still setting through their existing broken mortar and their existing outlets. It just so happens that the fastest growing revenue is in their digital channel. But they have to understand the relationship and the mix between them. The point I really wanted to make to you, though, is that depending on the purchase frequency and depending on the emotional connection that people have with that brand or that CPG, that tends to determine what might be the percentage of total brand revenue that they might get from a digital channel or direct-to-consumer. I've been looking at this now for probably more than 10 years. I know when I first surveyed about 100 different CPG executives and I asked them about lifetime customer value. I said: “What do you think would be the lifetime customer value in your best brick mortar channel? Then in a digital channel and finally the mobile channel?” Interesting, and I can remember this as if it was yesterday, they said that in the mobile channel it would be twice as much the lifetime value of the customer as it would be in their best brick and mortar channel. Now, imagine how much money you spend on trade promotions, how much money and time and effort you spend on your sell-in, your sell-through team, nurturing that brick and mortar channel? Yet if you could track them to your mobile channel, you'd know where they are, you'd be able to address them at home, when they're out shopping. You'd be able to develop a much more personal relationship with them because it's on their device. So I just wanted to explain this little segmentation; you can see that those which have a very high purchase frequency, not too surprisingly tend to have a higher percentage of total brand revenue available for a digital. But if we look at the, let's say more or less commoditized and high frequency purchases which we tend to map a little bit into the CPG area, you see their 5% to 10% of revenue. But you can see nonetheless it's a pretty interesting opportunity for any CPG and the key thing as I mentioned, is that it's growing faster and that the lifetime value of these customers ultimately will be higher. So it's investing in your future. Now thinking about the way that brands and CPG sell online, they kind of sell in different ways. So I'm concentrating a little bit here, much more on direct to consumer. But some of them sell to household partners online. Of course quite a lot of them sell through market places, it's no surprise to anyone that Amazon is the first place that people look when they're shopping or when they're thinking about how they're going to fulfill their need. So it's important to also to think about where you fall in that segmentation and what kind of channel you are going to use around your direct-to-consumer. Now, I've talked a lot about the importance of data and how that really is critical in first of all forging the relationship with the consumer, but also maintaining it and enriching it away into the future. We've been thinking a bit about this and we've been identifying some different technologies that we've seen people use to do this. Not surprising the technologies revolve around things like artificial intelligence and machine learning. Why is that? It's because these technologies enable you to do that anticipation that I talked about, to merge together multiple different streams of data. Now in the past we had to rely on surveys, we had to rely on a TLog or point of sale data, perhaps we relied also on data that we got from aggregators and syndicator of data. But today there are many, many hundreds of different sources of data, including public data, including Google Analytics. There's lots of data that we can use; the critical thing is can we can have a technology and can we have the expertise to blend them? So a couple of years, I was talking to one of our clients, the clients told me that they have no less than 17 different sources of data that they acquired. They were running a mixture of a digital channel and a brick and mortar channel and they told me, “You know, we used to really understand when we just had a brick and mortar channel, we understood our share of shelf. We understood our tactics and what to manipulate. Now we have a digital channel as well, we're buying in all this data, we don't have any way to merge it, we don't have any way to analyze it and we can't drive intelligence in time to take action, particularly in the digital channel.” “But even in our brick and mortar channel our promotions last on average, 14 days. We don't have time to analyze that data and to drive any action out of it. So what do we see really happening in CPG?” They have three really large problems. One is they're solving for the last mile that I just talked about. Two is addressing emerging markets, the next billion consumers are not in North America and Europe they're in other markets. Those markets are already using mobile even more intensively, so they need to be able to manage that data and reach those consumers as well. Problem number three is around product personalization in mass. So problem number one, they're solving for the last mile, what do we see them deploying, to collect and to analyze this data. Right at the top, the most important one is the ability to predict at extreme scale and to have local optimization, to help them to understand how to build those loads, how to find sometimes to crowd source that last mile delivery. What else are they doing? They're pushing some of that decision making to the edge of the network and they're using distributed trust systems, in order to be able to manage all those different delivery, particularly if they're crowd sourced it. Problem number two, what do we see right at the top again? It's data. It's all about a really vast scale of data. Many more people, much more data, much more complex algorithms that they need. Some of the others here, look at the mobile location based services, very important because these populations are already very highly engaged in mobile channels. So we need to be able to reach them where they are and to make recommendations based on data. But also, contextual to their location, to the time and to what we already know about them, we've already accumulated in terms of data. Problem number three, again right at the top is this idea of using machine learning, artificial intelligence to really anticipate what they want and when they want it, so that we can personalize what really matters to them. So I think you can see that these three problems, all three of them are very tightly related, to data. The important thing also, is that it's not really one day, one technology that they're using, they're typically using combinations of technologies and the key thing is trying to drive this insight and this anticipation so that they can drive a more personalized relationship with the consumer. What else do we see them doing? They're typically spending a lot of time trying to understand how to aggregate different views of demand but also, different elements of a supply chain. So many of them using some kind of control tower, to try and understand, let's say what's available to promise or capable to promise, market by market, location by location. Of course, they're looking at agile manufacturing, of course they're looking at let's say, driving predictive analytics for demand. But also using these same capabilities to synchronize not just local production, but also very importantly local fulfillment and local delivery. So these key things, it's insight into demand, but it's also insight into supply. Another, let's say, data problem, that requires significant amount of analysis. So we've been looking at a lot of this around the demand side, looking at the kind of data and timed capabilities that you need, depending on, remember we talked about the frequency of purchase, depending on whether it's a low frequency interaction, a high frequency interaction or a ultra-high frequency interaction, what kind of capabilities are more or less important. I think the important thing to draw out of this is, yes, we have still sales and operations planning that we thought about previously, but look how important analytics are as part of this control tower. When you've got really high frequency, you've got to be able to do that, and you've got to be able to respond and be able to react to the way in which supply and demand need to be brought together. What else do we see? We're seeing very significant change in the way people think about marketing. So I mentioned my client who said, yes years ago when they were just brick and mortar, they really had all the analytics they needed. They understood every tactic. As soon as they started through a digital channel, their big question to me was, “How do we know? Should we be spending more or less on Google Analytics and does that drive an interaction in the store? Does stuff we do in the store, does that drive the total wallet share of this consumer or this consumer segment?” So the relationship between these different elements really means they need to start thinking a lot more about their marketing management, which has always been a big problem. But a marketing mix now is more complex than ever before. We're seeing more and more applying AI to that problem. So the traditional shouting match that you might have had perhaps, between the people looking for trade funds and the people who are spending at the ground level, we're seeing more of that being resolved by data. By saying what does the data tell us about this and where should be spending that money? So, your action plan should be around let's say securing the data at the right level of granularity, at the right level of frequency, so that you can use it to shape demand. To use your tactics, whether they're digital, or whether they're in store in order to shape their demands. So there's a whole range of work to be done around pulling that data together and aggregating it and driving out of it, the intelligence that you need. This for example is work that we did, around boosting the return on trade spend. All of you will know how important trade spend is in the CPG world, but the question is, what kind of data do you need? What kind of insight and how does that work in a mixed digital and offline world? If you kind of put all that together, if you want to delight these new digital customers, you've got to have a great deal of data and superb insight and you have to use that to orchestrate, all the touchpoints that you have with them, which by the way include mobile touch points and include in store and also fulfillment touchpoints. They need to know, they're already going to change their order when they've made it. They need to know where is my order now and when is it coming to me? Well I've drawn on these reports in order to put together this short presentation. If you would like them please let me know and I'd be very happy to share them with you. Now I'm going the hand over to Suresh, thank you for listening to me. Suresh Bharadwaj: Thanks George. What I want to do is really share some of the work that we are doing and a maturity framework that we've been using with our customers to assess where they are in their data journey, right? As you might see you know people are not dependent on their shipment data to do their planning, but they're definitely leaning on syndicated providers for data which is typically two to three weeks out and when you talked about the granularity of data and syndicated latency which you just referred to in your presentation. Two to three weeks maybe good enough for medium to long term planning, but when it comes to operational planning and execution that is way too late. You really need to have, what I call T plus one visibility, which is yesterday's sales today. The second thing that we want to get to know is, the granularity of the data. What is the store selling at an item level at an SKU level on a daily basis. That's really as good as you can get in terms of getting visibility to granular sales and inventory data with speed, you know that could be near real time. For that you really need to work directly with your retail customers, because they're the ones who have access to the data, they're the data source. Collaborating with them will actually take you to that level and that's no longer a pipe dream anymore, a lot of our customers are actually doing that today. In one of my previous presentation I alluded to this as much that retailer data is not necessarily a replacement to syndicated data, you'd still would want to get syndicated data for other kinds of metrics that they provide you including market share, you know competitive information and so on. But this is something that could definitely augment what you're already deriving from syndicated data as a value. However, direct retailer, past data, is not necessarily going to come easy, not so much in terms of getting access to the data, but then how do you, given the variety of data format, giving the variety of frequency and granularities that retailers would like to share data with you. Handling all of the data and ingesting into your data itself could be a huge problem. Having said that, a bigger problem is probably the mismatching product hierarchies between how a retailer identifies those products and how you identify within your master data systems, right? So that is in most cases you know people tend to think that, hey, I do have UPCs and GTINs and therefore in a market like the US that's not a problem. You're probably right to a good extent, but I also want to point out when you look at the GS1 standards, there are a few exceptions that are made, which actually create inadvertent problems as well. For example, if a product is different only from a pricing perspective, you know and for a short division then GS1 standards will actually make an exception for you, from having to use a new GTIN number for that. Consequently, what happens is that you will have multiple SKU codes within your organization that's going to the same GTIN. So you know when you get a point of sale data, back from your retailer while he may be reporting the sales at a GTIN level, but you really don't know which SKU was actually sold because a lot of these skills would have been customized by the retailer banner or by key account managers based on pricing or whatever have you. So it's not a trivial task trying to map what the retailer is telling you in terms of what is sold in his store today with what you think he sold. That exercise of identifying and classifying products and be able to have those same reports, that you can meaningfully converse with your buyers, is actually quite a tedious part-time consuming activity. In those markets where you don't have, unique identifiers like UPC or GTIN, by the end of whatever, whatever have you, it becomes even more difficult. In that case you're actually dependent on product descriptions and that definitely calls for using more artificial intelligence and machine learning. Again you know the result are not going to be 100% accurate, so that will certainly require a certain element of teaching the system what to do. But once you've done that one time it's going to be fairly efficient going forward. What I want to do is really share a few examples of what our customers have done. So there are few examples and like I said, just a minute ago, you know knowing yesterday's sales today, is actually quite a huge step forward. Not many of our CPG customers are actually doing that today except to the leaders who are actually leading edge or even maybe even the cutting edge. But like I said, in my past presentation as well for them a COVID like situation was business as usual. So it didn't really take them by surprise. Yes, they were out of stock for maybe a couple of days, but they were able to recover from that very quickly where you saw many other brands being out of stock for weeks. So, you know they were able to adjust their short term forecast and they were able to react to the situation with speed. They were able to create those replenishment orders and then get back in business in a matter of days and not weeks. So that's a huge benefit that they've been able to get with that T plus one, or even with T plus two. But that's how it's going to help you, when you are able to get to that level of latency which is really short. I still call it a foundational capability, simply because you really want to be there before you can start thinking of, what else I can do to create a differentiation with my peers, right? Those who have actually managed to get to that next level have actually also been looking at not just their own product sales through these retail channels but also have access to competitive products as well. Of course, you know this is not something that's going to be available to everyone, it's available typically to category captains. But the question is I said you have enough challenges trying to map your own products from your retailer with your internal product hierarchies, just imagine what it would be to actually start receiving competitor product data as well and you have no clue what those products are, because you don't their GTIN or UPC numbers in your product, you have to build them from scratch. Number two for you to be able to look at their product description and then map it as closely to one of your SKUs or variants, again it is not a trivial task and I don't need to over emphasize that. But if you're being able to do that, and which I am saying many of our leading customers are already doing that, you'll be able to come back and design promotions that actually take into account competitive activity as well. While you're being thinking of looking at past sales, the kind of media spend that you've done, et cetera, competitive activity when it presents itself in the form of this data coming through your retailer point of sale, visibility actually elevates your ability to design better promotions to a significantly higher level. On a separate note, most of my supply chain execution point of view, having that visibility on all the levels will actually help you predict what delays at the time of receiving those orders itself. You can definitely relate this to the penalties that some of the leading retailers are enforcing on delayed shipments where they've reduced the window of acceptance from say 72 hours to 48 hours or even 24 hours. Predicting out of stock is definitely something that I have and again already alluded in a situation like we are in today. This would come in extremely handy and it actually comes down to knowing where your consumer off take is happening. So it's not only about being able to predict the overall forecast quantities correctly but also be able to deploy those inventories at the right location. When you're looking at the market segment of one, but even if it is not one to be able to get the product at a store near you is actually very important. To know that your competitor is actually running out of stock and therefore you have an opportunity to capitalize on that situation is going to be extremely differentiating for what you can do in the market. In the emerging markets we've also seeing other technologies including mobiles, where you know branch have been able to reach out to the retail customers and in those cases the retail outlets which are small you know mom and pop stores via the mobile in terms of the brand promotion. But also where the distribute alert coverage is restricted purely because of logistics reasons. It could be the size of the store, the value of the store or the distance from the distributor's warehouse. Enabling those retail outlets to place orders from mobile phones have actually eliminated the order taking cost and thereby help the branch increase their coverage. So these are some things that our customers have been able to do, you know leveraging both data and technology and that have created a very differentiating positioning for them. Moving on, you know I'll just take one example of what could be innovative. George did refer the last mile coverage and how do we do localized fulfillment? It's one thing for you to say, “Look, I've received these orders and I'm going to do my order allocation on SAP or whatever order management systems I have.” Then if I am running out of stock, then I'm either going to create a back order for this order and then send it as a second shipment. Or I'm just going to say, “I'm not able to fulfill this order at this time.” However, more frequently than you are aware, there is typically inventory available in your extended network. A lot of our customers don't have a visibility to that. While in the presentation today we've primarily talked about like getting visibility to your retailers’ sales and inventory. When you actually look across your customers and see where there maybe excess inventories across your network and that is not necessarily the metrics at your own facilities, but even your customer facilities or your channel partner facilities. Really extend that ATP process to include facilities from your extended network. What you can do is going to be very different. So now you're not just limited by your own inventory capacities, but you now have access to inventories that are available network wide. This is something that I really feel is cutting-edge and some of our customer are actually looking at this. But you have also seen a couple of news items in the last day or two where some of the leading retailers are trying to do this for their customers who are placing orders online and wanting their orders to be delivered at home. So they're marrying it with their inventories and also personalized shoppers to deliver that to you and me in the most efficient way possible. So hopefully this gives you an idea of what you can do when you have data, at the granularity of an item store or even an individual segment level and then you get to know that in real, real time. So moving on how do you get to that point? The first step in the journey is really to assess how can I get to the best in class situation? Number one, we definitely want you to assess all your channels specific customers if you will. Really come back with a value versus complexity assessment. Maybe needless to say, there are the ones that are going to be top priority where the value is going to be a clear, we know the complexity is not as much and then at the other end of the spectrum where the value is low but the readiness is very low and the complexity is also very high, we definitely to deprioritize those. So what this essentially means is it is okay for you to be best in class for certain segments and also be leading in other segments. The second thing that we want, that we're recommending here is, this is something that covered it too last time as well, was trying to do handle all of this data in house is going to be a humongous task. You're actually going to drown yourself and it could take months before you onboard your first retailer to the extent that you can start delivering value to your business. So looking at data as a service is something that you really need very seriously consider, because this is not going to compromise any of your core capabilities but it's definitely going to bring you speed and that is extremely important. The third element that I want to talk about here is implementing an element of data governance. I say this because again there are no easy answers here but you know the way you can handle or minimize creating those duplicate SKU definitions would greatly help you as well. Otherwise, all of this standardization actually comes back to nothing. But last of course is your cloud strategy. This is very important because while data as a service is something that you may embrace, the question is do you want to leave this outside your firewall or do you want to bring all of the data into your fire wall? Do you want the data as a service to be available as a co-located on a co-located cloud. So these are things that you want to think about and will certainly help you in further reducing latency. Once you do that, you know the benefits are there for you to pick and choose. Like I said the immediately low hanging fruit there will definitely be getting better sales reporting, KPIs that your business can trust. The ability to predict out of stock I don't need to repeat that, the ability to replenish your customer inventories with speed and really moving that, mind sate as soon as you could. Then of course coming with better promotional forecast and this is very important because in our experience the out of stock rates are actually much higher during promotions and that's a very well documented fact as well. That's really want I wanted to share on today's presentation. I'll leave the floor for any questions and answers. Rajagopal: Thank you, Suresh. I'd also like to bring back George and Praveen just to handle some of the questions that are coming in. The first question that came in I believe is for George because it was during your presentation. It says: What are the products available in market which are good in planning promotions as well connecting individual customers in order to inform customers about promotions and market and give mobile app to buy? Lawrie: It's interesting because, in general the solutions really which are good at planning promotions the sell-in solutions and I'm thinking about ones that you probably know and use, so people maybe like the CAS solution that now Accenture has, Accenture trade promotions. Or perhaps if you're in the developing markets you might use something like Cyntergy or Ivy Mobility. Or again you might be using something like Exceedra, those solutions are typically for the sell in team. They're typically not very well connected to the kind of shopper marketing that you're talking about that alert the shopper. That's actually generally one of the interesting challenges internally is to say, okay we've planned all this stuff and normally what you've got then is something which makes sure you have availability wherever you need to have it for the consumers. So it's sort of a gap actually bringing those two together, the digital promotional shopper marketing that you might have digitally and the traditional let's say trade promotion management. I think it's one of the things that I see CPG Is really concerned about because to manage that themselves is very hard. I had mentioned that company that was buying data from 17 sources had no way to amalgamate it and no way to drive insight let alone drive alerts to consumers on their mobile devices. So there's not an easy answer to that question I'm afraid. Rajagopal: Suresh and Praveen, if you have anything to add and George as well, while I'm directing some of these, please just feel free to jump in. I'll start with Suresh on the next one: With more retail moving online, pick up and delivery, does it change the dynamics of data collection and analytics? Bharadwaj: I think it just means more data, needless to say. We'll definitely need to move away from more traditional ways of collecting data, which has been either through FTP or file transfer based mechanisms to more API route to acquire data, and then to a large extent even potentially start looking at receiving data through data streaming. So, that is where most of the data collection technologies are headed towards. You know as much I've talked about reducing the latency to T plus one in terms of getting data from retailers, again you know co-locating the data in the sense of with your data warehouse on the same cloud. Rajagopal: George, I don't know if you have anything to add with retail moving online. Do you see it changing the dynamics of data? Lawrie: I would actually echo what was just said about streaming data. So we look at engines which manage streaming analytics and we're seeing more and more of this actually people saying actually in today's world I don't even care what they did yesterday, I want to know what they're doing today. What's happening in either as it were. So their click stream is more valuable to me in a way than the basket analysis that I have going back five years. What I really want to do is amalgamate both of them, see the basket analysis, see their click stream, see where they are now. So I'd love to get their mobile location if they would disclose it to me and that will tell me a lot. All of us know that our shopping behavior is different at different times and in different places. But generally speaking you need someone to help you build that, there's nothing kind of off the shelf that will do that because you need to decide what pipes you're going to put in and analyze. There are great streaming engines that you could use but to actually create, let's say the applications that solves a problem that needs a bit more. Rajagopal: Suresh, would the direct POS data from retailers replace syndicated data? Bharadwaj: No, not necessarily. Like I said, the big value you get out of syndicated data is that you have access to competitive information as well. It gives you market share knowledge which you may or may not get especially I you are not a category captain within a retailer. So it is not going to replace, if anything it's just really going to augment what you get from syndicated providers. Then this is more useful for real short term planning and your ability to react with speed, which you will not be able to do with syndicated data simply because it takes at least two, three weeks before you get access to such reports. Rajagopal: George, do you have anything to add? Lawrie: I would echo that it takes there is a latency with syndicated data. But of course is if they're doing work with it. They're collecting it from different retailers who different formats. So they're figuring out the formats, they're putting it in a way that you can see what was sold location by location or GLN by GLN and see your own products and see other people's. So it's one of the reasons why people like syndicated data. They don't like it because it's a bit too slow. Now if you get a direct feed well you're getting a much more up to date situation but you have got to do something about handling that data and normalizing it so that you can use it. If you get it from multiple different retailers that might be quite hard work. I mean we all know this, that to let's say install point of sale a t-locator is quite noisy, it's got voids, it's got training data, you need to clean all that up. So you need to someone to help you with that really, it'll be too hard for most CPG companies to do that themselves. Rajagopal: This one seems as a sort of follow-up: If someone is not a category captain, can the competitor data be bought from syndicated providers? Bharadwaj: Yes, that was my point too. I would say you would need syndicated data especially if you are not a category captain because then you don't have access to competitive information. But if you are, then there's a good chance that you have as good or as maybe a better quality of data and access to this data. Lawrie: Yeah, which is why there are certain people who don't like syndication. So they're already in a strong position they don't want to be let's say make it transparent to other people. Rajagopal: The next one is for Suresh: How long does it take to connect with and start getting POS visibility from a retailer? Bharadwaj: So, the answer is it really depends on the retailer and your relationship with the retailer. You know, as the data as a service provider we have typically onboarded a retailer in less than two weeks. Even a lot of that is actually related to getting the various parties to coordinate with each other in terms of agreeing on what data they want to receive and how they'll be able to ingest and test that what's coming in is good or good enough. The technical onboarding itself is a matter of days, it could be a day or two at the most. But a lot of times it is about coordinating and a lapse of about two weeks this is what we've been seeing yeah. Rajagopal: George, how do you discover innovation in competitor products? Lawrie: That's a good question because we already heard, didn't we? That you can get an exception from GS1 you don't always have to have a new GTIN. Though in theory you should if you have a promotional variant you should. But to come back to the question, how do you discover these new products? Well, actually, quite a lot of people use engines to crawl the web to see if they're being displayed and they use AI to interpret the image. But also the text associated to see does it have different features? Is it something new? Normally when people are listing a new product, they will list what the features are. So people sometimes in the CPG world use that to keep an eye on what innovation is going on. That said, a lot of it is quite let's say marginal, so it's kind of marginal imprint. If you think about let's think I'll just choose an example from, think about Magnum ice cream. When that first came out, so the first one was maybe milk chocolate, it was only a few weeks before there was a dark chocolate then one with a mint stripe all that kind of thing. Actually if you're competing with them that's very hard to be sufficient agile. You might find out by looking, by using a prowler an AI engine to see what's been listed online. But it would be very hard to respond to that quickly. Rajagopal: Suresh, did you want to add anything? Bharadwaj: Like George said, the GTIN might actually completely miss the innovative aspect of a product. So you either look at the image or the product description itself. Another example that comes to mind is you know years ago when Ketchup used to be in a glass bottle and you really had to shake it hard to get it out, then a few years later when it was put in a plastic bottle and a squeezy bottle in that, and when they turned the bottle upside down and made it table top. Something like that would get completely lost in a GTIN description, in a GTIN number or in a UPC number. The only way you can figure that out is looking at the product description itself, right? For that you need NLT, you need text mining, so those are the technologies that you need to see, is there a new keyword that I have not come across so far? So, that is one example that I can think of. Rajagopal: I have one on the maturity curve: Does the movement across the curve have to linear? So foundational to innovative or can we cut straight to the end state? I'll start with you, Praveen. Kombial: Interesting question. In our view, the foundational element like Suresh and George spoke about is definitely necessary to start getting the data feeds. But it's also about prioritizing the right channels and the right 80/20 principle to get most of it in place. But it's also not just linear. So you can start on the innovative layer, you can start on the differentiation there using the AI and machine learning. It's all about getting started. A lot of people hold off to get perfect data, and there's no such thing, as the full list of features and variables that you might think about for modeling and the data associated with that. It's also about getting started out. So if you take a step back, we believe the foundational element prioritized to the right channels is absolutely almost necessary to get started, so that is, that's why it's called foundational. But otherwise you can also start elements of innovation and differentiation with the parts of new activities that can get focused and take it from there. Rajagopal: Any last closing thoughts? I always like to leave the audience with some kind of how to start a best practice. Kombial: I guess it's also about getting started. Most CPGs are at different stages of that maturity curve. It's a good time to take a step back and particularly with the pandemic impact. Such basic things like granularity and velocity of POS data have come back center stage whereas it was kind of getting back stage, right? So getting started with that foundation there to our minds is an absolute necessity. To see that that granular level is a good place to start. But like I said, the innovation layers are pulling in and data more so is at the heart of particularly of the changing trends of our consumer preferences. Rajagopal: Absolutely. I think that's a great way to end because it is all about getting started, and sometimes that's the hardest part. I'd like to again thank our speakers for giving us their subject matter expertise today. I'd also like to thank EdgeVerve for sponsoring today's webinar. Finally, I'd like to thank all of you, our attendees for devoting some of your very valuable time to be with us today, so we hope you found it worthwhile.

Related Topics