The RetailWire Podcast

RetailWire Live at NRF 2024

• RetailWire

Share episode

Unlock the secrets of AI's sweeping revolution in the retail world with Ricardo Belmar from Microsoft and Jamie Tenser of VSN Strategies. As we wrap up NRF 2024's electrifying discussions, we share ground-breaking insights about how AI is no longer just futuristic buzz but a practical powerhouse reshaping retail. Imagine store associates liberated from mundane tasks, thanks to AI's magic, and the game-changing evolution of machine learning to sophisticated generative models, all dissected in this episode.

Jamie Tenser & Ricardo Belmar, delve into whether smaller networks can stand strong against the titans of trade. We analyze the seismic shift from traditional advertising to the dynamic world of retail media and its redefining influence on the industry's future. Get ready for a strategic exploration of how retailers balance their merchandising finesse with the burgeoning demand for promotional prowess in the age of retail media networks.

RetailWire is the retail industry's premier source for news, analysis, and discussion. With a focus on the latest trends, technology, and consumer behavior, RetailWire provides a platform for industry experts and thought leaders to share their insights and perspectives. Whether you're a retailer, supplier, or service provider, RetailWire is your go-to destination for staying informed and ahead of the curve.

Be sure to leave us a comment and let us know what you think. You might even hear your comment read on the next episode!

To learn more, or to join our Daily Discussions, visit RetailWire.com.

Speaker 1:

Welcome everyone to another Retail Wire Live. We are here live at the final day at NRF 2024. I'm Ricardo Balmar. I lead partner marketing for retail and consumer goods in the Americas at Microsoft and I am here with good friend Jamie Tensor.

Speaker 2:

Greetings all. I'm glad to be back here. At NRF Jamie Tensor, I run a little consultancy called VSN Strategies. It's my pleasure once again to be with Ricardo I think sometimes as a partner in crime trying to bring to you a condensed summary of what the things we thought were important from this show. There are many, many and we're going to try really hard to boil it down to the richest syrup of meaning that we possibly can. So bear with us. There's a lot to say here, but I think if you haven't been at the show, or even if you have, there's a good chance. You didn't get to see everything. I certainly didn't.

Speaker 1:

Yeah, that's true. We probably should start off with the one sort of obvious thing that everyone knows was inevitable to be talked about at the show, and that's all things related to AI. It was all over the agenda.

Speaker 2:

AI is everywhere, without a doubt.

Speaker 1:

Without a doubt.

Speaker 1:

I think maybe some of the things that I think we were talking earlier we agreed.

Speaker 1:

I feel like, despite the hype and there's no avoiding the fact that there's a lot of hype like that there always is at the latest technology, at every NRF I came away thinking that there was a bit of a balance or maybe a counterpoint to that, that there were plenty of examples of tangible things people have already done with AI over the past years and a willingness to talk about it and give examples. You know there were some sessions. Even I saw retailers talking about something they did, for example, with AI to augment the capabilities of their store teams, where they weren't looking to replace people but they were looking to make the work easier and trying to eliminate some of the mundane tasks, and retailers had examples of that. I think I even heard some retailers say they had KPIs that they were targeting. One return to session that an AI session I was at said they were trying to save every store, associate an hour of time every week with the use of these AI tools at a minimum and they felt that they were achieving those targets.

Speaker 2:

Well, I agree with you.

Speaker 2:

It feels to me like we've gone fairly swiftly from the science fiction stories of AI's taking over the world and jobs disappearing in thousands to a practical conversation about well, here's a technology that actually is coming of age and that we're able to start to build it into some of the practices and solutions that we use to run our stores every day, make decision making better, eliminate I like to call it junk work, but let's say tasks that are repetitive, dreary and that maybe we don't want our smartest people spending their time on, or even our store people, for that matter that can easily be automated.

Speaker 2:

The other thing that comes across here, as you said, in certain I don't want to call them point solutions, but in specific areas of practice, there are machine intelligence components that have been in play for a long time, and we've seen examples in things like, you know, inventory, forecasting and supply chain. Now, these are I'll call them special purpose machine learning tools. They are AI, if you want to use the technical definition, but what they are not is they are not large language models, the chat, gpt type of generative AIs, which I think we've crossed the threshold on in the last year and a half or so. Yeah, it's all artificial intelligence, but it's a different nature, maybe a different animal. Now, that's where the hype cycle seems to have come from lately, and I'm kind of glad that it seems like we're moving through that, at least in our industry, a little bit, a little bit briskly, and getting back to practical conversation.

Speaker 1:

That's right, and even then, I think some of the that, jen, you're right, that's where the probably the majority of the hype has been, because it's the most impressive when you see a demo of technologies right. It's truly creating things that are in a manner of moments that would otherwise take you hours, not longer, to create, and even though, as I've seen in the example, I think Tractor Supply has been here, I was talking to folks from there as well and they've leveraged those tools again and something that they've handed to their store associates where they can leverage those language models to just ask natural questions when they need to ask.

Speaker 1:

So there was another retailer who talked, in fact, interestingly enough, that in order to put those tools in front of or in the hands of their store associates, they needed to retrain them, because everyone now instinctively knows how to search by keywords, but when you're working with these models, that actually doesn't help you.

Speaker 2:

It's not the same know-how.

Speaker 1:

You need to really use a natural question. People wanted to do a product search in a store because they have a customer asking where is this particular power tool? You don't just search for the name of the power tool as a keyword. You would ask the AI where do I find power tool ABC in the store? And then you would get an answer and that required some retrain.

Speaker 2:

Well it does.

Speaker 2:

I remember when we all started to use search for the first time back in the previous millennium, we had to learn how to do that with some effect, or we couldn't find the results.

Speaker 2:

I think today, sometimes, when we interact with chatbots, we don't phrase our questions right, we get frustrated, we can't get the right answers and as the more powerful AI is coming to play, it's a skill, it's a learnable skill. I really think that for store associates and people in management, it's not going to be all that hard to learn this once a little bit of consciousness is raised. And remember, the machines learn back, so they're going to learn our limitations too and become better and better at understanding what we're trying to get at. So, when it comes to answering a question, whether it's about merchandising a store, whether it's about, perhaps, a merchandise plan that an executive is trying to work out for the next selling period in each of those instances there are going to be better ways to get at the answers, and that's something to look forward to. I think I'm a big antagonist for anything that is junk work.

Speaker 2:

Right, and you can automate it and you can trust the automation. Can we set that aside? Can we set that 80% of test aside and then free our minds, free our hands, if necessary, to do things that really add value and, ultimately, more satisfying, make it a better job?

Speaker 1:

Yeah, I agree, and I think that's true for again. So many examples I don't know that I could list them all, but, like you mentioned earlier, some examples are related to supply chain and forecasting, which aren't necessarily new applications of data and seed, but I think that the difference I see is more open conversation about it. For retailers, more discussion about the benefits and what any individual retailer is gaining, and that as an example to why others should be using it and that's being promoted a lot.

Speaker 2:

I think it's also important to distinguish between those large language model, the stuff that's gotten the kind of general media hype, the open platforms for artificial intelligence that essentially that protect nothing, that scoop up everything For a business application. That's not the place to necessarily set up your artificial intelligence tool. There's lots of reasons to keep a walled garden, so to speak, around your own instance, and if you have an AI platform that you're using to drive whether it's decision making or whether it's merchandising tasks or any number of other applications inventory, forecasting, et cetera that needs to stay in house. A short while ago, that was beyond comprehension that we could possibly afford and enable that within organizations. Now it doesn't seem far-fetched at all. In fact, it's been happening in small ways for a while. Now it's happening in larger and more exciting ways.

Speaker 1:

And I think there were a good number of maybe not as many, but a good number of consumer facing examples and applications. There was one session where some from Walmart was talking about how they added it, and again it was sort of a search enhancement, so adding into the Walmart app the ability to ask a natural question about something that you were planning, not a product-specific question, but the example they used on stage is what if there's your a parent trying to plan a birthday party for your daughter and your daughter likes unicorns? So instead of individually searching for party favors that have unicorns, like we would normally do, you would just ask the Walmart app what do I need to plan a unicorn-themed party for my daughter? And then it would return back search results to show you all of those elements. Another example was I'm planning a Super Bowl party, what do I need? And so the search results came back from anything ranging from grocery items to a new big screen TV.

Speaker 2:

Even I know you need avocados, exactly so which is? We're going to see this constantly. We're going to learn as end users, we're going to learn to interact with it better and they're going to get smarter at understanding the way we interact with them. So one of the applications where machine intelligence has been in play for some time has been in inventory accuracy, inventory forecasting. So can I make a clever segue a little bit into one of the other points? That is a conversation that we've heard more of from more directions, I think, certainly from the vendor side of the question here.

Speaker 2:

I think inventory accuracy has been certainly a discussion for some time. It's a source of loss Both over and under inventories in perishable products, anything that's perishable or that can expire or go out of fashion. You don't want too much of it, but you certainly don't. You do want to have enough that you can satisfy the demand, and it's no longer good enough to rely on instinct to do that. So the fast-moving consumer goods world has been using various forms of inventory management and forecasting and ordering tools for quite a while it could be 17, 18 years, if not longer and they've seen the benefits from it. They, I think, have found new ways to capture information, including the shelf status and other essential parts of the equation, and so they can get more and more accurate In the soft lines, which, I have to confess, because I really am a FMCG type of guy I still find it mysterious and awe-inspiring that bets can be made on apparel orders 18 months out, sometimes from a factory very far away, in anticipation of a certain amount of sell-through that will happen during a holiday season.

Speaker 2:

It's a marvelous mystery. It does look like a place maybe where a little machine intelligence might help narrow those error bars and the end of the day, if you can make the mark down quantity is lower, then that's a smashing win. So it seems like that's discussion. It's more often in the discussion. A few years ago we didn't hear that that story, inventory conversation quite as much. Now this seems to be a really great emphasis on it and I think it's connected to the interest in fulfilling store orders that are received online as well. So that feels like it's maturing a little. It's becoming part of the regular vernacular for the industry and it definitely references to it all over the exhibit floor.

Speaker 1:

I think that maybe takes us also to into discussions that have been happening around returns when returns?

Speaker 1:

are sealed in a very top theme that we're hearing more and more of. It kind of ties into both what you're pointing about the inventory management and I think also what we were saying earlier about AI, and is there an application that makes sense for AI? I know there have been some solutions I saw there where the focus was on AI helping you. I describe it as redefine how you position the product, because the idea then is to prevent the need for a return, because once the items purchased right and what do you do really to control the consumer they're going to return it. They're going to return it.

Speaker 2:

It sets up a stream of events that you just can't stop.

Speaker 1:

You can control it up until that moment of purchase and to make sure that it's the right purchase decision without the need for doing things where you're doing a. I don't know, maybe it's the right thing.

Speaker 2:

I'll buy it anyway because I know I can return a kind of situation which you're trying to avoid.

Speaker 1:

It's the retailer and maybe there's an application that AI tools can help there.

Speaker 2:

Well, with garments, getting the fit right is a key element.

Speaker 2:

And there are lots of attempts and I think some work better than others, but they're trying to incorporate tools that the shopper can use to get a better idea about how they will look in a given purchase. We see it for eyeglass frames, we see it for you know, there's a company I've been looking at where you can put in a bunch of metrics and get a custom made suit for not that much money. This kind of thing depends on our artificial intelligence to drive that, and the hope, of course, is that there'll be much less likely to be returned, because each time a return is authorized, a whole cascade of events has to happen that the retailer would rather not have to bother with and pay for. It's worth noting that one of the major sponsors here this year is a company called GoTRG. Their logo is all over the show. And what do they do? They help retailers manage their returns flow. So it is a big business.

Speaker 2:

Every time you shave that down by a percentage point, money is dropping to the bottom line, but it's also a reflection about getting it right for the shopper in the first place. So there are, I think, a host of different places where process and experience improvements. They're trying to shave a little bit here, improve a little bit there, and collectively it adds up. But returns cost a lot of money and there's a question of what do you do with the product when it's received. And there are some products where there's no point in even authorizing the return. Yeah, and I've had that experience I don't know if you have where Amazon has said don't bother sending back that part you ordered by mistake. It was an honest error.

Speaker 2:

It's cheaper for us if you just keep it. You shouldn't do that too often and there are probably people who abuse that. But that conversation, I mean that's going to be going on for a long time. But it does connect to another kind of process which is a little more sinister, and that is what happens when somebody stole the item that they're trying to return. And is that a good segue into the loss prevention question? Returns are maybe a form of loss from the retailer, but it's benign in its intention usually. And there's a whole another sinister thing that happens when an entire store is emptied by a group of people who next day are attempting to bring those items back for return elsewhere or send them back. That's ugly and it's gotten a lot of headlines in the past year. So shoplifting it makes good TV images, I'll say that. But it's not the only form of loss, that's the thing. In fact it's not even the largest.

Speaker 1:

I actually think that, given the recent things that you have to interact, there was maybe a lightning. I think a bit of a tone of how severe that is. On the show of Florida it was still being discussed and still being a real problem. Maybe it was a little overemphasized in the past year and so it's felt a little more subdued, I think, in the discussion.

Speaker 2:

Well, more measured it ain't nothing, that's clear about that.

Speaker 2:

It matters, stores in certain locations are forced to take remedial steps to reduce theft. It actually undermines the experience of the shopper, which is which is pretty sad. And I think about certain drug stores in Manhattan where everything worth buying is behind loose site doors and you can't purchase it without somebody. If you can find them to come with a key, that's. That's a real Disincentive for purchasing. So it's not just costing those retailers the the cost of those remedial Activities, it's costing them actual sales and you kind of wonder what's the arithmetic here In terms of what we save. But it also sends a message that I was discussing it through someone today. It makes me as a shopper walking to that store and think this is not a safe place. Right, that's a terrible thing to do to the reputation of the retailer and and, by the way, the news stories that we've seen it's kind of the same thing. You know I'm right, I don't want to walk into a jewelry store where any minute somebody's gonna back in SUV into the front and Try and steal, you know, break into the counter. I know that's, that's hyperbole, but it has happened. No, so the loss prevention discussion is big. We know there have been separate events about it.

Speaker 2:

One conversation I had was about whether RFID tagging, which is used principally these days for inventory management, could also help with with that problem of theft and what happens to products after they're stolen, and Might even help the returns process too. They are interconnected, but RFID seems to be Moving faster to that tipping point where it's starting to make sense for many not all, but many types of products. The tags are less expensive. There are, there are fabric tags that can be sewn into a garment invisibly. There are, and there are there are Detectors that can be used for checkout. For a check out, what someone I know described the bins in uniclose stores where you just throw a bunch of garments in. It reads the tags they get, they get pushed out the back into a bag and you walk out the door. Yeah, I, this is. It's a nice convenience for the shopper and also they have a very precise Set of data about what's being sold in those instances.

Speaker 1:

I do think there's a little bit of an RFID moment happening.

Speaker 2:

It is, you still can't put them on soup cans, but but maybe, maybe that's not what we care about. Okay, right, that's okay, because that's one of the ones that need where.

Speaker 1:

So yeah, I do. I do think there's definitely a Little bit more happening in that space than there was, so here's a long that technology.

Speaker 2:

So here's a notion we can posit about this. What happens if we, if we collect that data and it's it's shared among retailers so that An item, so that patterns to be detected in theft and resale and use that to actually help support enforcement? And also to decline to accept a return for something that's already known to be on a List of items that were stolen? Right, I, could that be enough disincentive for certain behaviors that we, we really would like to see Decline in the population. Yeah, so we will. We'll circle back to that one day, but I thought I thought it was a, I thought it was a thoughtful suggestion.

Speaker 1:

Yeah, very well, I think we can't. They can't be a retail show without a discussion in our next year around Retail media Straight. You can't avoid that one. So there was a whole, there's a whole day. For the first time, I see nrf had a pre-day dedicated to retail media with a number of sessions. I wasn't able to attend those, but they were also at least one other session that I think we both yeah, we did, we did attend.

Speaker 2:

I wanted to send my clone on Saturday. I just I just couldn't, I couldn't muster that up. It was, it was an entire day, it was eight or nine sessions and but there were several here that mentioned that and one night. I think that stood out that we both looked at. Yeah, it was kind of a summary of retail media. They think it included them Well on analyst Andrew Lipsman, representatives from both Walmart's platform and wall greens. So the great walls are both there talking about about some of the developments and I kind of came away.

Speaker 1:

I think there were some key things. I mean, a lot of the themes are recurring. I think we can go back to probably things we both saw and said at the grocery shop that were similar. But To me, this is one of these areas where the the momentum is not declining, it's still growing, that's still accumulating and although there is, I think, a very legitimate debate as to, you know, people want to ask the question Well, just how many successful retail media networks can there be? Not every retailer can have a media network and I View that, as you know, there's some logical the things that have to be true, right, for a retailer that is a specialty retailer, where they're fully Vertically integrated and the only products they sell are their own, it doesn't really make sense, right, and you're not going to see as much of a strong need, I believe.

Speaker 2:

Well, to be successful and not going to pay themselves to advertise our own product. So who's going to?

Speaker 1:

be to advertise them. They would and that's one of the things that we we've not seen in store for example, advertising in these networks for Something that's not products sold there, right? So you're not going to. You know what? Would you see? An automotive company, for example, pay a retail media network for an ad placement, right?

Speaker 2:

but I know, but I know, but I have, I have seen potential, I have seen a manufacturer, porsche, comes to mind, in fact where they develop a, a Content network in order to push out the most exciting content to their dealership network, but but they're not paying advertising fees to do it.

Speaker 2:

They were invested mightily in the platform itself. So I think you're right. If you're 100% private label, then the only justification for having a quote unquote retail media network is because it's going to drive greater sales of your own products. Not a bad thing, but there are several ways to accomplish that. Most of the discussion we've seen has been about incremental revenue for the retailer and for endemic products, that is, products sold in the store that are provided by brands. This is a way to ask, or try to garner additional spending above and beyond the traditional promotional spend and allowance support that have become traditional in the business. And there's good reason to do this, I think, from what we can see of the numbers, the growth and I don't know if you wrote down Andrew's forecast, but it was at 70 billion.

Speaker 1:

It's for? I think it's for this year. What I found that were the 60 billion and revenue and end spend for retail media networks, but that is equivalent this year to media end spend on linear TV, which I think the eye-opening thing that he presented is that that means that by next year retail media will overtake linear TV and that by 2027, it will be more than double linear TV end spend.

Speaker 2:

And this is not just because the retailers wish it to be so. It's also because the brands have enough feedback from their activity so far to believe that there is a payback, and so they're steering their dollars there. But they're also making investment in their know-how and personnel, because managing that much spending is a complex undertaking. So that conversation you mentioned earlier about can more smaller networks exist it comes down to how much bandwidth do the brands have because they wanna reach those audiences, if they can possibly do it. The issue is right now it's not yet manageable, but there are platforms in the future. I think that'll help make those decisions better, and maybe it's not about buying the network this is my thought but about buying the audience and then using a set of retail media tools that help you optimize that reach across a variety of channels.

Speaker 1:

Yeah, I do think I didn't hear it talked about here but I haven't other places. But I think there is this potential that when you start to look at these smaller networks, you know, when you move beyond a Walgreens or a Walmart or a Kroger network, what if those smaller retail networks sort of connect you know, gather together right and form their own sort of retail media aggregate network right, and found ways to work to and present it that way? So then you could have 10 smaller retailers in total add up to a Kroger or a Walmart.

Speaker 2:

Exactly right, we have in this country. Certainly we have wonderful regional grocers who cannot, who have trouble competing, and they were setting up their networks and the distribution of product through those channels is very important to retailers in each of those geographies and in aggregate, equal if not more important than the very largest retailers. But there's still progress to be made. There's still process to be created to make that efficient and effective.

Speaker 1:

Yeah, so it certainly worked to be done. I think this was talked about again here in measurement right measurements, and Andrew Lipson made the point that measurement makes markets that once the measurement tools are in place to your point, jamie right so that the brands have an easier time to manage all of this, then that kind of opens the floodgates right it opens the ability for them to invest more because the audience is there.

Speaker 1:

I mean he pointed out that Walmart's in store audience. If you compare their foot traffic right to Walmart, that's double the major broadcast networks in TC in terms of audience size, so that has to be appealing to the brand?

Speaker 2:

Well, it is, and they're beginning to actually have in place the screens and the channels to start to put messages out there. The brands who make those investments are going to want not just a head count right but also Direct, connected evidence that their investments in each of these messaging channels Actually it results in measurable improvement in sales and brand equity and certainly in their, their total relationship of the retailer, right, I think. For their part, walmart, I think that's a pretty good example and they're not the only one. I think they're determined to provide that evidence. It really are.

Speaker 2:

Yeah, but you know, it's been kind of the money first and the conversation is quite complicated, because now we see and hear of of meetings in which the merchandising team has their hand out, asking for the, the, the allotments of, of Promotional dollars that they've expected to always have. The. The network team is asking for investments in, in retail media and the brand is saying well, it is a finite number I'm working with, how about we take some from here and move it to there? Well, I can assure you retailers are not investing in retail media networks just to see money move from one pocket to another, but they are, I think, reluctantly, or maybe now just Pragmatically accepting well, some of it should move to places that are ultimately more productive in total for all of us Better margin and I and they're definitely hints in the discussions today that that's, that that kind of practical thinking is taking hold.

Speaker 2:

I mean, the payoff is so great that maybe people want to adjust their minds A little bit around it right and I think it's also why you know there was.

Speaker 1:

We saw evidence of all these other partnerships outside of the retail space that these networks are making right, whether it's with trade Desk, whether it's with Pinterest or social media networks or connected TV platforms and streaming TV. When you take that, you know what matters the brands at the end of the day is the consumer data that they're getting access to for these ends and the ability to track. You know, did my campaign work? That I have more consumers by my product because I invested in these ad networks? And the more the Retailers can establish that connective tissue to all these other mediums, then that makes them the preferred source.

Speaker 2:

Well, I come. They comes back to the data and maybe this is a good way to sort of Wine this down. But retelling more and more it's about stuff, it's about people and it's about the data, about the stuff in the people and the, the audience information. We've talked about advances in that area for years and years. It's enabled this thing where maybe the brands are, I think they're mostly Positive about. We can get more for our spending this way and perhaps also speak more Specifically and uniquely to individual shopper segments that in ways that really serve them well. So it should be a win all around, and I'm sure they'll be hiccups along the way, and I also think that this is we're still kind of in the nascent stages.

Speaker 1:

Yeah, because it is right, we ain't nothing yet. Yeah, it's still developing and I think it's only got at this point. I don't think we're in a place where we should be talking about is there a bubble that's gonna burst? It's still in that growth mode and it's still in that early, early stage.

Speaker 2:

Yeah, well, yeah well, so as usual, we could go on.

Speaker 2:

We could go on, but we won't yeah, but it's to be merciful to the audience, we're gonna try and wrap it up and and there'll be other opportunities that other shows. Ricardo, I've done this for a while and it's always really enjoyable. I can't imagine being with a smarter a Smarter, a companion, on this kind of, this kind of task, and it's a privilege to speak to all of you in the audience. So nrf, with its 40,000 people and thousand exhibits and it was about 150, 160 presentations yeah, it's, it's, it's overwhelmingly big and we've just been able to give you a little, a little in a little insight into a corner or a few corners of it. I hope that those of you are here will Benefit from it and those of you weren't here will at least get a sense of the flavor of this thing.

Speaker 1:

It is, it is, it's a phenomenon this year, as it is every year and I think we can kind of close close here and Let everyone know that you know if you did miss out on nrf, there definitely was a sense of excitement in buzz. The crowds were here, so anyone who was wondering would, would people attend? Is there still you know what desire to come to a show like this? I think the answer is a resounding yes. We're back to those pre-pandemic levels where the world is moving on and we're everyone's ready for it.

Speaker 2:

Surely the energy is great, yeah, so so thank you for listening to us. I hope that we enlightened you a little bit. Thank you, ricardo, for making me look good, and and and.

Speaker 1:

I can't wait to do it again. That's right. So we'll sign up here and we hope that everyone Remember to join us for online discussions at retail wirecom. Thanks for joining us, everybody today.

People on this episode

Head Shot

Brian Crum

Host