Exploring the Future of Retail & Consumer Goods Analytics

Today’s retailers are eager to cruise right past the recovery stage and instead situate themselves just a hop, skip, and a jump away from the future. CGT fielded some inside perspective of today’s analytic investment priorities through a chat with Servco Pacific, an automotive company that also has the ownership stake for Fender Musical Instruments. Read on to learn how it’s looking to enhance its analytic capabilities, how it leverages consumer data for a competitive edge, and just what's topping the list when it comes to concerns about retail data sharing. Tim Denman: Welcome, everyone, to “The Future of Retail and Consumer Goods Analytics” webinar, which is hosted by RIS News and CGT and presented by Amperity. I am Tim Denman, Editor-in-Chief of CGT and RIS, and I thank you for joining us. Both the retail and CG industries have endured unprecedented disruption over the last 18 months. We all know that, but we've emerged more resilient than ever. This reimagining of the enterprise as a nimble entity able to pivot to changing market conditions at a moment's notice is only possible thanks to a commitment to data, analytics, and insight-powered decision-making. With us to explore the future of analytics in both the retail and consumer goods industries are Tausif Islam from Servco Pacific and Chris Chapo from Amperity. I'll allow each speaker to briefly introduce themselves, their company, and their roles. Tausif Islam: My name's Tausif Islam, I'm from the Seattle area, but have been in Hawaii for two years now. I work for a company called Servco Pacific, Incorporated. We are an automotive retailer, but we have some pretty interesting business lines, including our startup, Hui, which is a car share program. We also have quite a large footprint in Australia, and have the ownership stake for Fender Musical Instruments. My focus is specifically automotive retail, working with our dealerships and distributor business to optimize our business and increase the customer experience. Chris Chapo: Nice to join everyone today. As Tim mentioned, my name is Chris Chapo. I work at a company called Amperity, which is an enterprise customer data platform company. I've been here a little over a year and my role is, within our product team, to help to bridge the gap between the raw data that customers provide us and the value they want to extract and generate for their business. Prior to Amperity, I spent my entire career on the other side of the house in companies leading analytics teams, retailers, and others. Denman: Thanks, guys. Before we jump into the results of the study, I want to briefly touch on the methodology and demographics. In the methodology, we sent out a survey via email to the RIS and CGT audiences in April and May of this year. They were slightly different surveys to each audience, but they largely mirrored each other, with slightly different terminology for the retail and CG side. Most results that you're going to see are tabulated as retail and CG separately to allow for a direct comparison between the two industries, but in some cases, like the word clouds we're going to look at in a minute, responses from both industries were combined. On the demographic side, about one-third of respondents are from the C-suite. For CG, the biggest cohort are from the CPG category, that's not surprising. For retail respondents, grocery and specialty were the biggest. For both industries, the primary core business function was IT and technology, which is typical for our audiences, and for revenues, too then, a large percentage are from Tier 1 ($5 billion and up) organizations. We've always looked at IT budget for part of this study, and last year, we saw CGs pivot quickly toward it, investing more of their budget in analytics in the face of the pandemic, for obvious reasons. Retailers were a little slower to react but have closed the gap significantly this year. We still see CGs investing more aggressively in data and analytics, especially in the 20% and over range. There are no retailers that report that kind of commitment currently, but I know it's understandable. Retailers are always allocating a large portion of their budgets to in-store technologies, which CGs aren't usually hampered with. Chris and Tausif, why do you think CGs typically allocate more of their IT spend to analytics, and what do you think they're doing with that increased investment that retailers aren't able to? Islam: I think you brought it up as you were speaking. This question is a little bit more nuanced; it's not as simple to answer. Retail is a really operationally-driven industry, so there's this layer of hardware that needs to be maintained. What's interesting is when you look at the actual issues that are for CG and retail, budget is really one of the top five. So, it's not like either one is really moving forward with analytics as I think they would hope to, but I think they're having to fight some of the constraints within their business. Most specifically with retail is the fact that they're operationally-driven, that they have hardware constraints, that's most likely pulling away from their analytics budget. On top of that, they actually have multiple channels, where most, I can't say all, but within CG they can be a little bit more focused in their effort in terms of selling products to retailers. Whereas retailers have multiple channels that they need to work on and optimize. Chapo: I agree with almost everything that Tausif already shared, but the one thing I would add to it, to a little bit more nuances as he mentioned, was when you think about retail and going back to this concept of being an operational machine. Sometimes the analytics, particularly if you need to invest heavily in it, have to go hand-in-hand with change management of how the organization's going to run business processes, and that can be to some degree a blocker to actually investing more. You can invest in the technology and the analytics, but if the business process doesn't change and if they're not actually willing to adopt and use it within their businesses, it's kind of a wasted effort. I've seen some retailers just take a longer approach to implementing analytics initiatives because they have to both operate the business and do change management. That's a lot to do at once. If they want to be successful, they tend to take a more measured approach. Denman: Most of the results that we're going to present today are quantitative, as most research is. It's numbers-based, but we did ask a couple of open-ended things, and then we did ask CGs and retailers to provide the one word that best describes the industry today and one word to describe the industry last year. For 2020, you can see that some of the most popular responses were, "Chaotic, challenging, adaptive, turbulent," which makes sense. They were going through a lot of upheaval, but for this year things are a little more positive, "Recovering, adapting, hopeful, transformative," as the one word they used to describe things. What one word would you use to describe the industry today, and do you think that word will be different a year from now? Chapo: I was trying to think through this. I think the word that really comes back to me is the word that's the biggest in 2021: recovering. The reason why, and then this is where I'm hoping, is recovering not just from the pandemic and all the changes, but honestly recovering in terms of the investments, going back to the belief that digital transformation is important, and really getting back to the basics on that. Islam: Tim, you may dislike me for this, I'm going to take that one word and I'm going to combine it and put a hyphen in between. I think for myself, I see it a lot in Servco and a lot of retailers that I look towards — it's customer-focused. The customer is really important during the pandemic, during COVID, and you see some retailers being focused on how to interact with their customer best during this time. Even during Servco, I could have picked transformational because we had this amazing change where we implemented our omnichannel strategy, so now customers can buy vehicles online from us. However, what you will find is that other retailers will have similar customer-focused journeys. This can be different for various businesses and it can manifest itself in different products. I love Target. It's a little secret of mine that I don't like to share. But now, you can go to Target. You can pick up your groceries or you can drive up and get your groceries put into your car. That is a great example of a customer-focused retailer. For the coming year, what do I think that word is going to be, or what are we going to see? We're going to see more resilient businesses. I think 2020 and early 2021 is like a weeder course. You go to college, get hit with organic chemistry, and find that some people just don't have the grit to get through it. As we move forward, we're going to see really strong retail businesses that have actually invested during COVID, whether it is analytics, different strategies that engage the customer, or the various technologies that they're investing in, email platforms, online portals, and so on. Denman: I would agree. I like the whole flow of “recovering.” I think we're past recovery now, right? We're now just building for the future, building tomorrow. I guess that's two words, but I'm going to cheat, too. The benefits of analytics, obviously, in CG and in retail as we've discussed, they can't be understated. Obviously, going into it and getting insight in the future is easier said than done. We asked retailers and CGs to pinpoint their biggest challenges, and we see a lot of alignment between the most popular responses. Both retailers and CGs named, "Lack of budget, absence of clearly articulated analytics strategy, inability to integrate data from multiple sources, and limited analytics tool sets," as key challenges. Unique to retailers was, "Company culture, resisting change," and CGs point to “not having the right staff in place.” Chris, do these challenges align with what you're seeing in the market, when you market among your clients? Chapo: It's funny. When I read this in preparation, I was almost saying, "How are you following us every day and through all the questions and concerns we have with our customers?" because this mirrors really well what we're hearing. It's pretty consistent. When I step back, there's two core categories that all of these roll up into. One being the technical challenges, things like, in this case, the limited tool sets, ability to bring data together from multiple sources, and that is really critical. That is actually one of the things that our company helps many with, particularly with regards to customer data. But the biggest challenge as I see it, is around people and process. I alluded to it in my last statement, but having the right analytic strategy: What do you want to do with it? How do you want to transform your business? Do you have the people in place to do that? Do you have the training available for those who you want to democratize this across your organization? Those types of challenges, you see them called out here, but those tend to be the ones that when I've talked to folks in the market, tend to be the hardest to solve. Denman: Tausif, what are the biggest analytic challenges that you're facing at Servco, and how do you guys plan to solve them? Islam: So, with most of my challenges, I dumb it down. For me, it's usually not enough time, not enough people, or not enough resources — you can probably take that and apply it to most other businesses. There's a really interesting dynamic in Hawaii, I've never really heard this term until I moved here. It was brain drain: the migration of intelligent people to different locations. We have fantastic kids in Hawaii, fantastic individuals, but they're leaving to the mainland for better paying technical roles and more interesting opportunities. What I've been having to deal with is sourcing the right data people, the right data-minded people, but with the skills as well. One thing that we're trying to do internally at Servco is, in December of last year we launched an analytics retooling program. We realize that it is very difficult to find the right person, but what I really appreciate in Servco is that we're finding the right individuals with the right mindset, and we're giving them the training to be able to work within our technology stack proficiently to meet the analytical needs of their organizations. For me, I would much rather work with an individual with deep business knowledge and give them the tools, augment their skill set with the right tools, and we did that. We're doing that by taking them through. Of the two individuals in our pilot cohort, one of them has an AA, the other has no experience in analytics. We're taking them through Excel to Tableau, to SQL so they have the ability, model, and pipe their data. They have the ability to visualize their data, and Excel just because it's the largest analytics tool in the world, it's how we communicate with our executives. We've found that we may not always find the right individual, and this is an issue. I think going through the report in retail, 45% said, "Right staff not in place." So, instead of going out and finding individuals, we have vetted Servco employees that we believe in, and we're investing in their future so they can actually work on the critical needs of the business — they don't have to ramp up. I could probably count on one hand the people in Hawaii with deep automotive knowledge and data skills. Denman: Those are two unique things to bring together, for sure. Now, still similar to this question about the top challenges, we asked CGs and retailers to name their top areas and for analysis. Again, we see some strong alignment between both sides of the house. "Consumer insights, demand forecasting, inventory planning," all ranked high by both audiences, and given the disruption the two industries face, that certainly makes a lot of sense. We need to understand consumer needs and wants, and figure out how to best meet and serve those needs consistently and seamlessly. Tausif, what do you make of these numbers? Are demand planning and consumer insights your biggest use cases right now? Islam: Interestingly enough, I think consumer insights is always our number one. If you speak with any of my direct VPs, it's always about the customer journey, about the customer experience. If we look back almost two years, I don't think we would have the capability to really understand our customers. It wasn't really until we started working with Amperity and connecting our siloed data sources that we were able to track our customers. Now, a customer can come in, I can look them up and say, "Oh, you created a lead at 1:00 a.m. You looked at this Toyota. Are you interested in these things? You have this old vehicle." We have a lot more insight on our customers, and it allows us to be strategic with our marketing and email activities. Most recently, we've been utilizing customer data to enable our retargeting strategy. We've also created thresholds. If we believe that our click-through rate isn't high enough or our subscribe rate’s cumulative-like run is exceeding our threshold, we'll pull off. We'll slow our marketing activity. It gives us a lot of understanding of what our customers are doing and how they're engaging with our content. Up until a few months ago, I had no idea that most of the email opens are usually within at least 24 hours for our group. Finally, what's interesting is that I see these stragglers. We sent out an email campaign three months ago — it’s as if someone came out of a sabbatical from the world and opened up our email. It's learning about these really interesting engagement points that I'm excited about, that my business stakeholders are excited about. What I will say about demand planning is that it's critical for Servco. Again, Servco is a very interesting company. We have brick-and-mortars, we are a retail organization, but at the same time, we are the sole distributor of Toyota and Subaru on the Hawaiian Islands. If you want to buy a Toyota or sell it at your dealership, you'll have to come to Servco and put in an order. Demand planning is very important because it dictates the inventory available for our customers. As of right now, coming into this role, one big aspect of my responsibility is building the foundation for our future analytics. When I think about demand forecasting, true demand forecasting, traditionally, it's a bunch of guys in the room doing this and saying, "I think we're going to need X and Y of this." We want to get to a point where we're actually utilizing analytics, specifically predictive and prescriptive analytics, to make smarter demand-planning decisions. As of right now, it is manual, but we're building that foundation to make better decisions utilizing analytics platforms. Denman: Chris, is there anything missing from these lists that you think might make an appearance in the next few years? Chapo: Honestly, I don't think so because this is pretty comprehensive in terms of the key areas of focus. The one thing that might actually change though, which is the under-the-hood for this, is the how by which this is done and what are the inputs that go into these types of areas. For example, Tausif was describing demand forecasting, right? Historically, people may have used historical sales data at a store or maybe at a store's SKU-level to help build a forecast for the future. One thing that I've started to see more often is, "Can we add additional inputs into that that maybe we wouldn't have traditionally thought about?" So, an example of it could be customer-based inputs into a demand forecast. Example: What are the local search terms around this certain store for the types of products people want? Because you may not be able to measure that if you've never put it into a store, so it's a self-fulfilling prophecy of demand forecasting where you only forecast what you put in the store. But what are other inputs that you could try? That also goes down into the assortment planning component. I'm speaking for retailers who are actually using inputs: customer inputs, customer segmentations, who are the people buying in a place that informs what's been a product creation and the product generation engine of the company. So, to step back again, I see that a lot of this is probably going to be on how it's done and what the inputs are versus the subject areas. If there’s a question about inventory planning, particularly around the pain-point of the out-of-stocks that we've all personally felt. You know as you've gone into stores and heard about just why aren't people putting this higher on the list and investing more in it? It's a really good question. I stepped back to the demand forecasting pieces. First, you have to know what the demand is going to be, then you need to plan the inventory to fulfill it. But the other piece I’d like to add, at least I've talked to many retailers, to some degree it's like they've already bought the product that they've got. They've — for some specialty retailers where their horizon is 6, 12, sometimes 18 months out when they're buying product — it's already been purchased. So, what we're trying to do is optimize, based on the inputs, demands, and what we're hearing from consumer insights, the fulfillment of that product versus just the planning and buying components of it. Denman: That's great, and that all fits nicely into our next conversation: the how. How are you guys doing? How are retailers and CGs doing in particular categories here? We asked CGs and retailers to evaluate their analytic prowess against their main competition as well as the industry leaders, and on this slide, we're presenting their self-evaluation against the industry leaders. If you want to see how they ranked themselves against the competition, it's available in the full report. For retail industry leaders, it's Walmart, Target, Amazon. For CG, it's the Procter & Gambles of the world. Without digging into any one category too deeply, you can see at a glance that retailers and CGs consider themselves pretty woefully behind the pack leaders. There are only a handful of areas where any executive claimed to be significantly better than the industry leaders, right? Chris, despite CGs' ongoing analytic investment, they rank themselves as “significantly lagging to industry leaders” at higher levels than retailers in basically every category, specifically analytics strategy, data management, and data quality. Why do you think CGs perceive themselves as behind the leaders in these critical areas? Chapo: That 's a really good question, and I wish I had the Magic 8 Ball to answer it. Some of the inputs that I would suggest might be helpful for us to get a perspective on that comes down to the analogs that people are using to compare themselves against. Many of the success stories that we hear, particularly in retailers, are those that are really far ahead. What it tends to come down to is that they're building this because they have a direct, first-party relationship with the consumer. They can shape the customer experience, digitize that, and create better experiences. They have access to all this rich, first-party data. One of the driving points, particularly for CGs, where the data is at arms' length, is that there isn’t necessarily a direct relationship with the consumer. They're working and selling through retailers. For many this puts them a step behind from the get-go. So, one of the things that we're seeing with some of our customers is where certain CGs and retailers want to work together to have, in a privacy-compliant, very secure scenario, to be able to share data that creates better experiences on both sides of the house. To me, that is an example of one of the things to help bridge the gap between the perception of where people are today and what the future could be. Denman: The data sharing is obviously vital, and we're going to talk about that in a little bit because that's always been a big component of this study. Now, Tausif, how would you rank Servco Pacifico's analytic prowess in some of these categories against the Amazons and the Walmarts of the world? Islam: One thing I will bring up initially for this conversation is Servco is a 100-year-old company. So, a hundred years ago, I don't think anyone could say that they were a data-first company. Now, when we look at the Amazons of the world, they are a data-first company. That's what they utilize to engage the customer throughout their customer experience. So, can I confidently say we are on-par with these businesses? Probably not. They started off as a data-first company. However, if I compared ourselves to our direct competition, other automotive retailers, yes, I would give Servco a better rating. I can confidently say that because of the changes that we've done in the past two years alone. COVID dramatically accelerated our omnichannel strategy. We had a long and lengthy roadmap with a two- to three-year roadmap to implement our omnichannel strategy. Then COVID hit. We became super customer-focused, and were able to roll it out in less than a few months. What I will say is that we are working towards building out the capabilities that Amazon has. We're investing in the right analytics platforms and investing in the people to build them up and give them the skills. One thing I want to point out is: Why are we going after an Amazon-like experience and capabilities? Well, what we have found is that at least 60% of our customers make up the bulk of our profit. We don't have sole individuals propping up our business. We're not a luxury brand like companies such as Burberry where one customer can go in there and drop maybe $20,000 a day. With our products, we have a lot of individuals that are helping our bottom line, very similar to Amazon or Costco, for example. We cannot have a narrow product strategy and we are working towards building up the capabilities that some of these leaders have in retail. Denman: Tausif, you mentioned before about talent, we always ask about talent. It's always been a challenge, and it will continue to be a challenge, I imagine. We asked retailers and CGs how many internal resources that they have, and you can see here the average is down to around 16 per company. Both sides of the house leverage a mix of internal and external sources to address resource needs, with CGs leading heavier on internal sources than the retail side. So, Tausif, in your experience at Servco, how do you look to enhance your analytic capabilities? As you're doing that? Are you looking to grow your internal team as you talked about, or are you partnering with the analytic third-party companies? Islam: For those listening, I apologize. I personally hate when I get a, "Depends," answer, but what I can speak about is Servco. We're going through a very large organizational transfer — organizational and platform changes. Within the past few years, we've brought on marketing cloud, commerce cloud, Amperity, Tableau, and there's a lot of change happening. It's hard to manage that change when you have consultants and outsourced team members. As we've gone through this change, we’ve been hiring internally to make sure that we have the core team members not only to develop, but also support these changes. However, depending on where you are in your analytics journey, it's fine to outsource some of the needs. For example, would I develop individuals internally to do dashboarding? Yes. Would I hire a bunch of data scientists to do customer stitching? No. I've got Chris and Amperity — they do it better and faster. The key is to find a partner for your journey. When we looked at transforming Servco, one of our three pillars was making sure that we were data-focused to enable the customer experience. Working with Amperity, we've found that they're a customer-focused business, that's what they do. They do customer stitching. When you look at building out your team, think about what's core and absolutely necessary to supporting your business. If you don't have that skill set, find a partner that shares the same goals. If you find a generic consulting company, of course, they're going to do the work, but you may not feel that they're asking the right questions. Some of the best engagements that I've had with consultants is through bringing up a problem. It will come off the top of my mind like, "Oh, I'm sorry this happened," and they'll ask about it. Then they'll ask, "How can I help?" When you can find a partner that's willing to help you outside of the bounds of what is traditionally their world, that's when it makes sense to invest in outside capabilities, consultants, and so forth. Denman: Now, Chris, I think Tausif teed it up for you, but can you talk a little bit about some of the benefits of partnering with an outside source, best practices, and advice on how to get the most out of that relationship? Chapo: Sure. The answer is, "It depends." It's a bad joke, but I had to say it. Actually, Tausif mentioned a word: core. There's actually a principle that I learned a while ago, when I was working at a company called Intuit, where they use this phrase often of "core versus context" with regards to particularly the software or technology capabilities that the company needs. When I think about this, I say, "What's core to your business? What's unique to what you, yourself, can do or is differentiated?" That's something that you want to invest more in doing yourself. For example, it may be that you're a specialty apparel retailer and there's no one else like you out there. There might be a planning component you may need to do yourself. That to me is a great example where you might want to invest in your own capabilities, and I'll share how to do that successfully. First, if it's core to your business, you should be focused on it and building your own internal capabilities. Second is context. These are the things that are important, maybe critically important, to your business, but it's not differentiated for your business. An example may be Amperity, which as Tausif mentioned, has customer identity resolution as part of its platform. It's really important to businesses, but it's not unique to your own company. That's where partnering with someone who could actually deliver that capability for you in the world of this context is important. Things like your e-com platform may be another one, which is probably not a reason for you to build your own because others have done it, and you can probably get to market more quickly which is really important. Going back to the core piece, this is something I personally learned. As I mentioned at the beginning, I was on the other side of the house running analytics teams and companies. I will give you an example of where we were trying to do something core, but we weren't necessarily super successful with it. I won't go into the details of the problem space, but it was an area where we had outsourced a certain analytic component to a large consulting company and they delivered. They created some amazing analytics that were very impactful to how the business would make decisions, but they did it separately from the rest of our analytics teams. So, what ended up happening was that it was a one-time analysis. We were never able to refresh it, we didn't have teams internally understanding how to build it. What turned out to be a really good opportunity never grew. When we actually approached it the second time around — and this is something that I would recommend if you're trying to build internally — we did a co-development activity with a third-party because they've got the experience, they've done it before. We had their team co-located with some of our data scientists building on our own platforms, our own environment, this new analytic capability. Once it was done, we were trained. We knew how to do it. We had a deliverable, it could operate, and be operational as the business. So, there's a couple examples of not just how to think about it, but also ways to try to make it more successful. Denman: Specifically around the idea of workforce. A very big, well-known company is facing this challenge: the inaccuracy of data limits the ability to leverage advanced analytics. Do you have any tips about how to build the business case to convince leadership to invest the time and resources required to get accurate data? Secondly, support the investment towards advanced analytics? Basically, how do you get the leadership to buy-in to analytics? Chapo: I'll take a first swing at it, and, Tausif, I'd love to get your perspective. It's a bit of what I would call, "Trying to win the hearts and the minds of folks," particularly in the senior leadership team. The first, I would say, is more the “heart” piece, which is trying to identify what your strategic opportunities are that the business is trying to accomplish. If you're a retailer it may be around improving inventory planning or margins on product. If you’re the customers' delivery experience company, think about improving your Net Promoter Score, or you're likely to recommend. Then build up the case to say, "Well, if we're at a certain point today and you want to get to the new point in the next year, here's the two or three areas that analytics could help you achieve that journey." Instead of spending millions and hundreds-of-millions of dollars to prove that out, say, "Give us six months." That’s basically saying, "Give us a small amount of budget, and a small, dedicated team, and let's prove why this new approach can actually make a difference." Then, get clear ahead of time, this is your success measure, so take a small bite and actually prove it to people. The second piece is particularly around the data quality. We often talk with customers about, "Well, if you want to enable this customer experience ..." For example, maybe I want to welcome people when they arrive at my hospitality company. When they arrive at one of my hotels, I want to welcome them. I want to provide them a differentiated experience because we know that the check-in process is super important. I want to make sure that if they have any problems while they're staying with us, that those get prioritized over others because they disproportionately contribute to our profit. If that's what you want to achieve, you can go through and say, "Well, here's how the data breaks out. Here's examples or areas where the data isn't even connected to be able to do that." Those are a couple areas, but, Tausif, I'd love to hear some of the success stories you've had in your journey as well. Islam: Sure. We talked about having small, bite-sized pieces of work that you can actually showcase, and that's important. It can be really hard to get executive buy-in, especially if your executives' goals and your goals and objectives are very different. Their goal as an executive or leader may be to grow the business and optimize current business processes. Your goal as a data engineer or a manager of a data company is to make sure that whatever is being utilized within processes are cleansed and ready to be utilized accurately. Depending on the analytics you're doing, dirty data can be a roadblock. At the same, some data is better than no data. When it comes to getting accurate data, it really is important. Everyone is doing their best to showcase and communicate the value, but it's not until you can showcase something that that value is actually visualized. This is a lot of what I do within my organization, my title is director of data visualization, and the second portion is analytics. I meet these fantastic data scientists, data engineers, and then they say, "Hey, I've shown my work to my executives, but I'm not getting buy-in." It's that one piece that is keeping them from communicating effectively with their executive — the presentation piece, the communication piece, the value piece. You may want to focus more on that: understanding how your executives are going to listen to your pitch and how they're going to derive value to drive specific business objectives and goals. This is where it's difficult to find commonality between overarching business goals and specific unit or organizational goals, but it's a battle worth fighting for. Don't stop. It's going to be difficult, but it's worth doing. As a data professional, it's the first thing I think about, but when I try to objectively prioritize certain work in my backlog — I've got some projects cleaning current data sources, optimizing marketing activities, optimizing planning activities — when I look at the value that provided to our customer, how it's impacts our bottom line, it kills me, but I am optimizing. I'm going to choose to optimize our marketing strategy. I'm going to choose to work on planning activities. It doesn't mean that I am not and will not focus on the operational activities of getting the data ready for use, however, sometimes it just takes a lot to communicate the value and how it's going to impact other processes. I find that executives don't really understand the value of data until they see it impacting operations. If you can communicate why it's so important to the operational and bottom line, go ahead and try and talk to your executive on what is important for them as objectives. Then, go to your team and try to align your data goals to your executives' business objectives. Denman: A little more here on the internal operational structure, which is what we were just talking about. We asked both retailers and CGs, "Who's currently responsible for business analytics in your organization, and ideally, who should be responsible?" and the results were interesting. By a 3-to-1 margin compared to retailers, CGs reported that analytics is managed by one main analytics department, and for retailers the most popular response was, "Analytics is managed by the IT or technology department." When we asked them, "Who should be responsible?" Both sides agreed that ideally there should be a shared analytics department/center of excellence. This study has evolved from a data sharing study that RIS and CGT used to co-produce years ago, but we continue to track data sharing across the two industries today, and it is a major part of the study. We asked retailers how often they shared this critical data with CGs and their trade partners, and as you can see on this slide, there are some key areas like online customer behavior, loyalty, and CRM data that a majority of retailers are not sharing with trade partners. Data sharing has always been a challenge for retailers and CGs, and we're seeing that it continues. Why are retailers particularly unwilling to share some valuable data with their trade partners? Islam: Data sharing, interestingly enough, has been going on for quite some time, however, it's been vocalized in the past few years. Companies like Snowflake are touting their data sharing capabilities, and it's important to be able to share data with your business partners. At Servco, we have an interesting business — we're the distributor and the retailer — so we don't have an issue in terms of utilizing the data for the greater organization. However, I can see that to be a large pain point for retailers, traditional retailers, and traditional CGs. Why do I think that this is an issue? Well, data privacy concerns are a large one. When you go to a store and engage with that store, that retailer should have a level of integrity around the customer data. Most retailers respect that social contract with their customers. This is changing, the laws in the U.S. are changing. In the U.S., you're able to readily buy customer data if necessary. Retailers and their partners are understanding that in order to optimize the customer experience they're going to have to start sharing information because they're going to be lacking the data necessary to do worthwhile demand planning and inventory planning. Previously, retailers were comfortable doing that themselves. Back in the day, they were comfortable and capable of doing so, but now as retailers and CGs are becoming more competitive, you need to form alliances. You need to realize you can’t compete against an Amazon without actually sharing core data between both industries. Data sharing is important now because people are dealing with very data-savvy competitors. To move forward and be a resilient business, it's something that's going to be necessary for their future. The big one is data privacy. A secondary piece could be getting the right people, setting up the connections, setting up the data pipeline, and understanding the intricacies of connecting that data. There's platforms that do that like Dell Boomi. Boomi is basically like a visual ETL, but it helps organizations that are doing mergers consolidate data from various different organizations. It's a fantastic question: Why aren't we sharing data? Well, it's hard even in your own company to do the data modeling, the pipelining, and having a good data dictionary. Now imagine if you needed to share and communicate everything about your data to a different company or a larger organization, it's not easy. I don't think we have the technology, the people, or the platforms in place to make it quick and efficient. Chapo: I 100% agree. I actually had three key reasons. First, the data is a competitive advantage. People say, "Data is the new oil," and so having access to it and being able to use it is the competitive advantage because if someone else has this, they may use it to take advantage of me. That is one of the core pieces that historically has happened, but as Tausif mentioned, people need to rethink what's happening in the larger context. The second is privacy. Making sure that the respect and honesty — I call it the stewardship of the data — happens properly. Third, and we're starting to see some emergence for this, it's difficult to do this on an ongoing, systemic basis and be able to take action on it. Sometimes, people will do a one-time, clean room analysis. If I want to be able to take action on this and actually create different experiences, such as cross promoting for brands, being able to potentially offer benefits and loyalty programs, there really aren't the platforms out there to both do the analytics and then be able to action it. We're just starting to see some emergence because there haven't been the platforms to enable it. Denman: On this slide here, there are some key findings from the report, which we touched on most of these. One thing you'll see here is there's a couple of key findings about AI and machine learning. There's a statistic that wasn't in the presentation, but AI and machine learning is one of the lower-level investments so far for both “currently capable of” and “currently investing in.” Further down the pipe, there's going to be more money headed towards it, but the question from the audience is, "Why are we seeing such a small amount invested, or such a small interest? Why are retailers so slow to implement AI and machine learning into their analytic capabilities, and are we going to see that explode in the next few years?" I'm going to answer part of it because I definitely think it is, and I think we're seeing it ramp up right now. But the question to you guys, why are we seeing retailers and CGers a little slow to put AI into their analytics? Islam: It's a combination of a few things. It's executives and business leaders not understanding how to manage and what kind of projects to select utilizing analytics. Also, analytics is this broad term that's thrown around for everything: traditional business intelligence is now analytics, reporting is now analytics, I've seen someone pass me an Excel sheet calling it an analytics product. So, what leaders and executives think of analytics is very different. Finally, when we think about retailers, traditional retailers, they've been around for quite some time. They're not data-first companies, companies like Amazon. I wouldn't even call Apple a data-first company, but they have really transformed and are utilizing data in very smart ways. Traditional retailers are playing catch-up. Here at Servco, because I talk about what I'm doing within my organization, I bring up the fact that I'm building a foundation. It's hard to do worthwhile analytics projects when you don't have a foundation. When I think about building out an analytics team and organizing the team to build-out core components of our analytics strategy, the foundation is having good, clean data. The walls are a good, useful, descriptive analytics tool. The roof? It's going to shield you from future issues like rain, storms, and that's your AI and machine learning. So, how can you say, "Let's build a roof," when you don't have a foundation or walls? It's something that you can't immediately jump to. What I'm thinking is these retailers, these other big players, they're building the foundation towards moving towards a worthwhile machine learning and AI product within their industry, within their stores. However, there's just a misunderstanding of what we can utilize AI or machine learning for, especially when you sit down with a data scientist. On top of that, there will always be that communication barrier. If you don't have a business leader, maybe that comes from a research background, it can be really difficult to make sure that you're using the same verbiage to understand and direct whatever analytics projects you have to the business objectives of the company. Chapo: It's an interesting question, and I loved everything that we just heard. The one piece I would add is that the action concept oftentimes, the low-hanging fruit is not the fancy AI, ML pixie dust that people say you sprinkle on the world and magic comes out. The reality is you can actually do a lot with some more basic things, and for many companies it's a good place to start. However, the opportunity when I think about this, particularly for some of the AI and ML things, are there tasks that by leveraging these tools you can automate and remove costs or friction from your overall business process. Those are good places to start because it actually proves out why some of these methods could be useful in other places versus saying, "Hey, I'd like to get to a place where I have AI predicting all the planning for my entire inventory and business." People aren't going to that day one because they don't understand it, to Tausif's point. They may not believe it, and honestly, the business is not ready to operate on it. So, two things. One is probably there is low-hanging fruit that you can get to market faster with more traditional analytics. The second being applying it to some of the more mundane friction points and cost points in your business first to prove this out so you can actually tackle your bigger projects later. Islam: What Chris said is fantastic. I didn't bring this up, but a lot of the things that I did when I joined Servco was just building out reporting in support of the business. A great example is that we automated this report that took an analyst eight hours to complete — eight hours. I joined the company, had no idea what the report was for, and took it down to 20 minutes. The reason why it's still considerably long in terms of an automated report is because we have to wait for an ERP system to kick out an Excel file. I didn't see this report as this fantastic analytics product. For me, it's a report utilized to understand the inventory in place, it's in a planning report. It gives us an understanding of parts available that are necessary for us to build out vehicles. When distributors get a vehicle, it's probably a generic vehicle, and then customers will say, "I want X, Y, and Z things. I want this badge. I want this kind of exhaust tip." So, as a distributor, you need to have parts on-hand to build out that vehicle. I had no idea, but this report was actually utilized to understand the inventory of parts necessary needed to build-out vehicles for customers. Again, I was not using AI, I was not using machine learning, I was utilizing a combination of Alteryx and Tableau, and descriptive analytics tools to curate a report much faster than previously. This report alone went from 800 hours per year, down to about 50 — eight hours to 20 minutes. By automating this report, we were able to realign labor and increase the customer experience because there were no back orders. Imagine you bought a vehicle, however, for whatever reason, it's not built-out in time because we didn't have a report available. It's the low-hanging fruit that you'll find the largest change. Here at Servco, we have a really small footprint for AI and machine learning, but we've seen dramatic increases in terms of business operations and how we're able to utilize that data. Denman: Well, thank you, sir. Unfortunately, we are out of time, but we could probably talk all day about automating and augmenting analytics, which could be a report unto itself. Thank you, Chris and Tausif, for joining us. Thank you, everyone, for listening, and thank you to Amperity for sponsoring the report and this webinar.

Related Topics