CVS Health Corporation (CVS) Earnings Call Transcript

Welcome, everybody. Welcome to the -- one of the probably final sessions for you of the NRF Conference. Hopefully, everyone is enjoying their experience here. I know it's been eye-opening for me in a lot of areas. We have an amazing panel today, and I hope you're excited for this. Snowflake here is happy to host several of our customers to talk about a couple of key topics. I'm Ross Armstrong. I run the Retail and Consumer Goods vertical at Snowflake. And this is many years at NRF and excited to be here again this year. The group of panelists we have are from Albertsons Media Collective, Niel IQ, which some of you might remember as NielsenIQ, the Republic National Distribution Company, CVS Media Exchange and LiveRamp. So it's going to be a busy session. And what we're going to do is, we're going to do several fireside chats, and so we will have people coming in and out. But hopefully, if everything goes seamlessly, you will get a lot of great information. And as I always try to think of these things, you're going to have a lot of questions walking in. If you walk out with more questions than you came in with, the better off we are. You can't -- can you hear me okay? All right. Thank you. So without further ado, I'd like to welcome Evan Hovorka from the Albertsons Media Collective.

All right. Testing. Okay.

We're good. We're good. All right. So Evan, I think it would be great just to start quickly with your background, your role and a little bit about the Albertsons Media Collective.

Absolutely, yes. My name is Evan Hovorka, VP of Product and Innovation for the Albertsons Media Collective. I've been in retail for 18 years, 16 of those at Target where I built out a lot of the brand marketing channels and then eventually was led to Roundel, which was an in-house kind of evolution of the retail media network at Target. So I had a lot of years in retail media, a lot of years in just retail marketing in general. And kind of the one consistent piece of that journey has been the need for innovation and the dependency on data. Fast data in the hands of the right practitioners. So excited for today's panel.

Excellent. So I think Evan might be understating his -- how well known he is because one of the things he's done is taken his innovation garage and taken it on the road, which I think if anybody was at [indiscernible] last year, you might have seen that, a really innovative idea as well. So -- I actually have a picture of your garage here, I think.

There we go.

We'd love to have you talk a little bit about what the idea is and how it's been executed?

Yes, absolutely. Growing up small farm town, kid in the garage a lot just fixing everything that broke, motorbikes, tractors, you name it, and I kind of had a love of taking things apart, learning how they work, getting educated, breaking things maybe worse than they were before, but then learning from that experience. And I think I carried that through to adulthood. And the houses I've lived in, I've always had a garage space where tools, innovation, usually some -- maybe some scotch or whiskeys in there, too, for some fun, but I always love to [tank] or build prototype. So I love the 3D printer space when that was big, figuring out ways to use technology to solve problems, usually from a personal and selfish level. But as COVID evolved and workspaces changed when I move to Albertsons, I still live in Minneapolis, I used the garage as really a meeting space to host innovative companies to help us solve problems. And I think we'll get into what some of those problems were. But even companies like El Toro, join.me, Pinterest, Google, folks that are really also trying to figure out what is next for retail media, how do we solve it together? There's not a lot of off-the-shelf products that you can just plug and play. And so having that collaborative space where it's not a face-to-face meeting room, it's not a sales conversation where you're going to buy this off-the-shelf product, we're just going to sit down, hold hands and put all our Lego blocks on the table and talk about how we can build better models together by understanding each other's pros and cons.

It's a great idea, and I love it. I've never seen a garage live, so one day, hopefully. So we'll get into the retail media network of the media collective in a second, but maybe you could start off a little bit with the data journey that you've been on?

Yes, absolutely. So this is kind of the core Snowflake conversation. I mean, data is the core of any retailer long before retail media came along. And that data has been quickly evolving to service things like websites and apps and everybody here can appreciate what item data looks like, images, price points, coupons. That all seems pretty simple, but when you're talking about 1 million-plus SKUs across 2,200 stores, all with varying price ranges and availabilities, sometimes sourced from different distribution centers, that data collection while it's the core of what our consumers are buying, is really complicated. And so then trying to centralize it, standardize it, make sure everything is available to put media on top of it, is the core of how retail media gets started. You got to have the items because we are very capable at selling products. So those product data points are critical. That's kind of like one-on-one. I think the audience would go to sleep if we got too much deeper into that topic. So then what happens next is how do you tie that to the CRM system. So now we've got those items in the baskets of our customers. Fortunately, at Albertsons, our loyalty program is really well adopted. Our transactions are identified at about a 93% plus rate. So we can show which consumers are buying which products, which brands, which ones are sensitive to price, which ones redeem coupons. But all that now means we're manifesting data up at a higher level. And not only that, then we want to use that for marketing purposes, analytics, research and really partner with our CPGs that want to run media through us. So that data has got to take shape into identity graphs, audience for targeting, measurement solutions that can show the value of that media investment. And so you start to introduce things like clean room or fast-moving data in transit where it's encrypted, but still in service of the consumer. So those are now getting into 201, 301 types of applications of the data, but it all starts with having accurate fast-moving data at the core and then layering on top of that with ideally as few as systems as possible and then making the media shine on the back of that data.

Great. So step one is let's get all of our data in the system and make sure that we've got our ecosystem of data providers there as well. And then you can only then activate that for your -- for whatever the purpose is. And so obviously, you mentioned the media network, and that's the core of your business that you manage. So maybe you could talk a little more -- I think a lot of people in the room are curious. I know we've got valuable data. I don't necessarily want to monetize it all. Some of it is our secret sauce, but some, I think we have real value, but I don't really know how to make that happen, right? So once you've got your data in a data platform, how do you then take that to the next level and really activate a true retail media network?

Yes, there's a couple of simple things. So we'll touch on it and we'll get into a couple of complicated topics on top of that. Usually, the retailer will have an enterprise marketing team, and they're there to drive foot traffic into the store, promote own brands, promote a new store launch, and they're really looking for category lift, store lift. When you get to the RMN level, sometimes there's a layer of granularity in those retailer systems that doesn't support brand-specific requests. So Pepsi, for example, is not interested in Albertsons' category growth in soft drinks. They want to know what did my shoppers buy? Are they buying incrementally more of that? Am I gaining category share and share of voice. And so the specificity of a system that would measure soft drink sales for a retailer has to double-click into those specific SKUs. And then especially when you're getting into the marketing side where they're only wanting to talk about the diet gluten-free protein shake and it's just these -- literally these 7 SKUs, those are the ones we have to market and put the media behind and then only show sales of those 7. So it puts an extra strain and tax on a retailer's traditional system, which just means faster data, faster servers, maybe more specific analytic skills to help show the brand, not the category, what the performance is. But I think where we're going with this is, all of that needs to manifest within the retailer's owned and operated property, so website app. At Albertson's, we're fortunate to have 12 big brands: Safeway, Jewel-Osco, Albertsons and a bunch more. And you can imagine putting media on those platforms and apps, I'll say it's relatively easy, although it is a lot of work, but we own the end-to-end ecosystem. And that's a wonderful place to promote a new protein drink for a soft drink company. The growth, though, if we stayed on platform is very limited. So I would say within the first year, RMNs of any size are looking to move off platform and take that audience, that targeting, that measurement into the platforms where those shoppers are spending more of their digital time. So Pinterest, TikTok, Meta, Google, Trade Desk, [OpenWeb], and being able to do that then requires another layer, maybe we're getting into 301, 401 of encryption, data distribution, data onboarding, LiveRamp is coming on here in a little bit. They're a great vendor for that. Snowflake is a great enabler of that. But you have to get the data at that granular level with the consumer background behind it safely into these other platforms with efficiency, right? So we're talking thousands of campaigns happening concurrently. And now we're at a level of really ad publisher and ad tech that a retailer traditionally isn't staffed to support. So now you're looking at different engineering or different partnerships to make sure that, that new capability is pretty scalable.

And maybe even in Snowflake, obviously, we're interested in hearing more about that. So tell me a little bit about how Snowflake helps you enable some of those 301, 401 capabilities you were talking about?

Right. I mean, the clean room concept is probably the greatest, but there is core RMN support that Snowflake offers. But getting it in -- selfishly for my needs, getting it into the hands of our media practitioners safely, securely with as few as hops as possible is at the core of any modern retail media network. I mean, gone are the days of bundling up audiences and shipping them out over 2 or 3 days, decay of that data is occurring, control of that data is occurring. Being able to do that in a centralized clean room on an identity that's anonymized with the reduced hops and as close to live data as possible, just creates a much better platform to enable things like AI, fast metrics, like trying to do that on audiences that have been copied and pasted 2 or 3 times across different ecosystems. You've created this kind of codecs and translation layer that's going to hurt performance. So centralizing it makes perfect sense for privacy, and it makes great sense for media performance because we just have such faster access to the truth.

Excellent. And so obviously, I think you mentioned you were at CES this past week, and now you're at NRF. You've been to a lot of kind of retail-focused conferences. What are you excited about? Like what's coming up in 2024, either for Albertsons Media or just in general for the retail space that you're excited about?

Yes, there's 3 key pillars that we're excited about. One is this evolution of media and merchandising. The reason that brand is talking to the retailer is because they're looking to move units off a shelf, maybe grow their shelf space. And they really have leveraged that core merchant relationship long before I ever showed up. And so that's the core of why the retailer exists, and that's the core of why the shopper comes to the stores, is to get that product and that quality and that price point. Bringing that relationship closer to the media side really is just an honor of what the brand is trying to do. They want to move units. We've required them to invest in media and start talking about things like impressions and clicks. That's not what they're here to push, right? Pepsi doesn't cash checks on clicks and impressions. They need to see product leaving shelves and ideally in an incremental way. And so merchandising plus media allows the retailer to show up with our best foot forward in that total brand relationship. It does require a little higher-level conversation, we call them joint business plans, but bringing in the top level execs from both sides so you can plan out the whole year versus sometimes media is on the short run cycle every 6 weeks. The merchants run on 12-month, 18-month cycles. So a lot more strategic, lot more thoughtful in their year-long plan. The media should just plug right into that and support any of the merchant decisions. So that's Pillar 1. Pillar 2 is channel expansion. Like I said, online versus off-platform is important. But off-platform means many, many things. And so social is easy to understand, but there's a lot of social players. So we're going to grow into some of the social platforms that we're not in today. We'll grow into CTV. CTV has been a difficult channel, I think, for retailers that sell primarily grocery. The margins are tough, but we've got some creative inspirational ideas that are going to help push that forward. And then in-store would be the other channel expansion topic. The stores themselves are great places to have conversations with the shopper. If it's incremental and not disruptive, which we've probably all seen some of those experiences that feel a little distracting, our strategy has been one of, let's be thoughtful with how we treat our customers in our stores, they're our guests. So we should be only helping them or keeping things neutral. So things like, the avocados, they're on sale, great. Here's a video of maybe some different recipe options for Guacamole that you haven't thought of. You can engage in it or ignore it. If you want to engage, maybe there's a QR code, gets it into your basket. So channel expansion. And then the third one is really around data collaboration. So this concept of clean room has been around forever as well, but you've got to have it in secure, affordable, fast-moving systems so that we can partner better with our social platforms for closed-loop measurement. We were the alpha partner with Pinterest on closed-loop for social, a very much in-demand product for brands who want to see their incrementality. And we're continuing to push the envelope with Trade Desk, how we allow autonomous access to audience and measurement through that very popular platform. And then our big partnership with Meta and Google is still evolving around things like clean room, identity, safe and secure identity mapping, all in service of what's coming in 2024, the cookie apocalypse. We think we'll shine brighter than those reliant on cookies, most retail media networks will because it's all dependent on first-party consented CRM systems tied to a reason for being, right? There's a lot of companies that have a full path on the United States consumer base. But if it's not tied to a system where they're converting, our customers shop 2.6x a week, then what does it tie to? Like a home address, a survey, a credit score. This is tons and tons of records of SKUs tied to a human, tied to contact information in a fully anonymized secure system. And they're going to keep coming back, right? Because the reason for being is, oh, you're a grocery store or you're a convenience store or a gas station, whatever the retailer is, that's the magic, right? It's not just the identity and the ability to activate on the identity. It's the connection back to the conversion, which is what the brand cares about. So any hop between your identity and the conversion, the reason for being is going to be a tax and the one that's going to be penalized because of cookie deprecation.

Well, great. Evan, thank you so much. I wish we had more time, and so much more I'd like to ask you. But if we can give a quick round of applause for Evan and his garage.

Thank you so much.

Thank you. I'd now like to welcome up Dirk van Renesse from NielsenIQ and -- I'm sorry, from RNDC and Troy Treangen from NielsenIQ. So let's just start with Troy. I'd love to just briefly introduce yourself and tell us a little bit about -- I mean, I think your title of CPO kind of explains itself and maybe a little bit about your scope of control and kind of what you're focused on within NIQ?

Sounds good. So I'm Troy Treangan. I'm the Chief Product Officer at NielsenIQ. As you can probably imagine or you probably know, I know, we do -- we operate in about 90 markets all over the world, measurement analytics, price promotions, et cetera, et cetera. I'm accountable and my team is accountable for all the products that we make across the board. So we're going to talk about some of those here in a little bit.

Great. So maybe we can start with the power of data sharing and what the value has been to NIQ around data sharing?

Yes. So, again, NielsenIQ, our biggest business is actually measurement, right? So you probably know we collect data in many markets. We can actuate in 2 ways. We have retail measurement data which is a roll-up of stores, which is probably the most common that everybody knows. The other part of it is we have consumer panels or omni-shopper panels in many markets as well. All of that data, we obviously collect, we create the products that we create and then we share that data with our various clients in multiple mechanisms. Obviously, our tools, so we have our own web interfaces to do those. That's one mechanism. The second mechanism is through our analytics capability. So sometimes somebody wants to do a -- like I said, the price promo analysis or a study or something along those lines. The last option though, which is transforming pretty significantly is our customers need the data in their data lakes or their environment so they can merge it with other data sets, do other things with it that maybe are not use cases that we focus on. And historically, we would do that by just extracting all kinds of data in a flat file and send it across. That is not the best way to do it anymore. I mean, the data is getting way too big, the data is changing way too fast and the cost to keep it up to date in a mechanism where you're actually creating a file and shipping it off is just very inefficient. So what data sharing brings, like in a Snowflake data share environment is, our clients can now access their assets that they license from us in real time in our environments directly. So instead of having to say, hey, Troy, give us an extract of that data and we'll process it and we'll do it a day later, 2 days later, they can actually change their processes or change their applications or analytical tools to connect right to our environment, access the Snowflake data share and then they can perform all of that, whatever their analytics or things they were doing way ahead of where they used to be able to do it time-wise. So that's where the value ultimately comes in with, again, Snowflake specifically and what we've developed with them.

And for those that are quite aware of how we do data sharing, the data never moves from your environment. It stays within the control of your 4 walls, stays within your governance and security model, but then it can be accessed by other Snowflake customers instantaneously. So you keep the control, but then you have the speed that you're looking for to make sure you can use that data effectively. And this leads me right into NIQ Connect, right? Maybe you could talk a little bit about the Connect product and we'll turn it to a mutual customer and talk more about it, but a little bit more about NIQ Connect in particular?

Yes. So we've been on a journey. We've acquired many companies over the last several years. We've also just recently merged with GfK, which is a tech and durables side of -- Nielsen IQ at its heart was FMCG based primarily, but then we just merged with GfK, which is tech and durables based for a lot of the same capabilities that NIQ has. We've been on a journey for the last 2 years to take our Connect platform is what we call it, which is our cloud-based capability and roll it out everywhere around the world. So now we're at a point where one platform is working everywhere. Now when I say everywhere, there's 2 caveats. One of those is China, and one of those is Russia, but there's other limitations and issues of why that's the case, but I still say the phrase everywhere just -- because it sounds better. But in reality, we've rolled all that out and one of the backbones that we have as far as our scaling capability, our data warehouse is Snowflake. So Snowflake is something that we do license to actually make that engine work. If you think about how much data we actually process, you guys would probably be pretty surprised. We collect, like I mentioned earlier, in 84 markets is where we collect retail measurement data. A lot of times, we get transaction log data for many of those retailers. And we process, on average, 16.7 trillion transactions a week to be able to globally in order to get outputs or analytical outputs to the client that we support. Now that number might sound maybe not that big to some, but just for reference. Like all of the 3 largest credit card companies process 6 billion transactions a year combined. We're processing 16.7 trillion transactions on a weekly basis to make this work. So scaling is a very, very important part of our ecosystem. And given that we moved all of our stuff to the cloud now and one platform everywhere in the world, we had to find an engine that could help deliver that, and that's where Snowflake is what we've chosen to do that execution and scaling. And that's what is -- again, it's one of our -- it's our core backbone of our product at the end of the day.

And building the product and scoping it in that value is one aspect of it. But the next question is really when we actually put that into effect, what does that mean? So I'll turn it to Dirk and maybe you could quickly introduce yourself and talk a little bit about your organization, let's -- I'm sure everybody is aware of the RNDC?

Sure. Good morning. Glad to be here today. So I'm Dirk van Renesse. I work for a company called Republic National Distributing Company. We are one of the largest privately owned distributors of adult beverages in the country. So I got a great job. So sitting between these two guys, it's pretty fun to bring all of that together. My role in the company is, I head up all the category management processes. So I have a team of over 100 that we call on customers around the country, really consulting on their wine and spirits and beer portfolios in each of the shelf to the consumers. And a lot of that transaction that Troy talked about is we use that daily. So between the partnership with NIQ and Snowflake, it's all about speed. There's no question about it. We actually coined a phrase in our office. We don't care about insights anymore. We call it cross insights. Every UPC that we manage, we have over 16,000 brands and over 100,000 unique SKUs that we sell on a daily basis in over 40 states in the country today. When you think about that item with having over 450 unique attributes, there's a lot of fast things that happen with our data, and that's why we need the speed and the insights from that data. People ask me all the time, how do you really want to use that? And my job is really a business case. What's important for us and for those of you going on this journey, I highly recommend that you have a sponsor on the sales side. You need somebody to understand how the data is going to be used. We don't need another dashboard. We have thousands of those. We can't read pie charts anyways. So we need something that's executionable, something that we can get to the sales force, and something we can actually measure back on what happened on that. I can give you a great idea. I mean, I won't use New York because it's snowing here today, but I live in Atlanta. So let's say, Atlanta on Saturday, it's going to be beautiful. And let's say, it's going to be 60 degrees. You know what? I have over 500 restaurants that have patio bars. So we want to go after those patio bars. It's going to be nice outside. If age inventory on [indiscernible] and we want to get that out with a decent price, get it to the market, you get all the table tents made, have all this done by Friday and give it to the reps for execution and then be able to measure that on Monday morning on what we did. Then not only within those tools but actually understanding who those customers are and where they live and what they shop for, so we can actually tweak what our assortment looks like. On a high level, that's what we do. And we've actually found, at least for me, I think I've turned into more of a governor. I mean, I spent more time lobbying my IT team, teaching them about the business because they really need to understand how we're going to use this in the workplace. What exactly are we trying to get to? You have to be very specific about your approach. This can't go, yes, I need a chart or I need an Excel shape. No, I need -- this is how I got to use it and this is exactly how it's going to be presented and I have to be able to scorecard those results when they come back. And I find that through a sponsorship that we do about once a month, we literally present, like we're presenting to our customer through the head of our IT departments that are doing on this business. Because most of them, they're great, they're smart engineers. They're unlike me, I know [indiscernible], they don't. So we got to teach in how those brands work. But really in a very fast scope, that's what we do.

I love that. Actually, I love the idea, too, of making sure that your team that you're empowering to do the execution on the IT side, just thinking about the outcomes and not just the requirements they've been given because sometimes those 2 things don't necessarily line up, and there might be a better way to do something. So thinking about the actual outcome is a great approach, I love that. So maybe you can talk a little bit more about how the NIQ Connect solution comes in play with that model?

Yes. So we found for years, a great partnership with Nielsen and NIQ. They understand our business model. That's probably been the most important thing. They literally sit down over the last years and they talk about how we're going to be using the data. It's difficult in my world. We're in over 40 different markets. We're still based on 50 legacy systems because of our acquisitions, again, with the private family business. So we may buy another business. We just bought one in Fayetteville. Now we're on their platform, but they have to integrate all that to ours. So by doing that, Nielsen totally understands how we're using all the data from all of our customers around the country to integrate that into our platform. So not only do I need to know what's sold to the register, plus all the attributes that I give to it, I have to be able to scorecard that information back to the customer and say, hey, this is what we recommended, this is how we will consult with you. And by the way, what happened, right? Everybody sells something, nobody ever comes back and go, hey, how do we really do? We do a good job or do we not do a good job. So one thing that Nielsen built really can help us out on their platform and their reporting tools. We actually can customize all of our reporting tools using the same database for one version of the truth is specific to that customer. It's how the customer wants to look at their business. H-E-B is different from Wegmans, it's different from Kroger, different from CVS. They're all different. We have to make sure that we're pretending how they want to grow their business. Their OKRs, their object key results are all being generated the same without working in our business. And that's one of the big things about the platform that Nielsen offers.

Awesome. So maybe we can pivot a little bit and talk a little bit about what's next? Like what kind of innovations you're looking forward to in 2024? We'll start with you, Troy.