Unidentified Host or Guest (2:51)

Do you see any kind of future where Amazon partners with OpenAI in a way that is like letting OpenAI scrape Amazon's kind of marketplace data in a way that then you can shop products through ChatGPT. Do you see that ever happening?

Patrick Miller (3:06)

So Amazon's cracking down much more on scraping than they, you know, have, you know, historically, because they know the value of all that data. And so let me phrase it a different way. What do you think? You know, Amazon and Walmart's biggest regret is that a single search happens anywhere other than their sites. And so like, you know, they, the fact that they are, you know, some of Google's largest customers, you know, and when they're not pausing ad spend, they don't want another, like, they don't want another one of those. And so this is a frenemy, you know, sort of relationship. And so if they want, you know, if any of them, you know, want the growth, it's like, how does, you know, OpenAI or any of these other LLMs support the growth that they want? And one is that retail growth versus their cloud business growth. And I think that's, you know, one of the real keys in sort of pulling apart the relationship.

Emma Erwin (3:49)

Got it.

Unidentified Host or Guest (3:49)

I knew that that was the answer. I was just buffering on how to say it back to you from your deck and I was like, okay, okay, okay. Oh my God, you're on the spot. You have to actually say something beautiful. Okay. Can you actually, I'm curious for anyone that isn't super like well versed in the cloud business, can you talk a little bit more about like why Amazon, like how, why they need AI partnerships to grow that business and why Jassy's in that mindset just to, for anyone curious.

Patrick Miller (4:14)

So like, if you think about like the Amazon from like, like an investor perspective, it's a sort of a sum of the parts business in the sense that you have aws, you have an ads business and you have a core retail business. And so while it started as a bookstore, they then developed aws. You know, they didn't disclose it, you know, for the long period of time because it was high growth and high profit. When they did it was like, oh my God, I cannot believe this thing. And you know, and I think at its peak it was hosting like a third of the Internet or something like that. And so Google and Microsoft saw that and they said I want one of those. And so they had their cloud businesses and so Amazon is, you know, they saw a dominant share. But Azure and GCP have also started to take off. And so whenever you sort of look at their financial reporting, it's like, okay, like I look at like how's retail doing? What's happening sort of on a, on a unit spaces there? What's happening sort of growth in North America versus international. Then look at the ads business, look at that sort of, you know, overall growth. What do they disclose on search versus sort of upper funnel? What is that from a tacos perspective? When is it sort of going to decouple from the overall retail business? And then looking at the business right now, the least about which is sort of aws, but it's just sort of like what are the, what is the growth, what is the margin? And then compare that to GCP and then to Azure. And so even though it has dominant share, it has, you know, it has moderated versus those competitors. And but it's hard to keep up that pace when you are already the biggest. So they have to come up with ways to accelerate that business. And so if they end up sort of doing some of the compute, you know, hosting, you know, some of these LLMs, you know, that's a way to drive growth. If they figure out ways, you know, it's like Google like I guess two weeks ago announced that like all your Google Docs, you know, are then integrated into Gemini. And so like and it shows like all the incumbents will figure out how to featurize this as opposed to prioritize it. And then that will then you know, sort of figure out ways to drive growth. And so many companies use AWS already to host all their data. Hey, how do they then layer the models on top of it to drive the growth within aws? And it's, it's principally a compute growth, not a, not a storage growth.

Unidentified Host or Guest (6:09)

Got it. Last question on this story from before our brand friends that are listening for what should they be thinking about right now with everything like Based on maybe Jassy shifting perspectives or maybe that doesn't actually impact anything. But like, how should brands be thinking about what's happening specifically in that age?

Patrick Miller (6:25)

So there's an interesting article I read over the weekend and it was like, some, like Google, like SEO guru and I don't understand really how Google works, but like, but one of the things that they were pointing out was like, when you query the LLMs, they're oftentimes firing off a query to Google or to Bing that then populates how they then build out the response to your query. And so that's a long way of saying it's sort of same as it ever was. And so like, there's all like these like magic geo companies that are out there and it's just like when at the end of the day it's like, man, just like get your catalog right and it's like, yeah, there's some hacks and. But like, those things are not, those are not durable advantages. And so like, what did, like, you know, what's the infrastructure that you can build, you know, that creates a durable advantage? And ultimately that is, you know, things like, you know, having a clean catalog, having all that data there, driving up what's feeding into these PDPs across the various retailers. That's the biggest thing that you can do. As boring as that is, it is the right approach.

Unidentified Host or Guest (7:25)

Got it the same as it ever does. It reminds me like, when I first started at Flywheel, my job was keep the damn catalog clean. It is, it is true. And it prevents a lot of problems. Okay, for the sake of time, we're moving on to story number two, all about Walmart. So you just came back from presenting at Walmart Connect's annual sales conference, which means that you got to talk to the people who are working to keep WMC in the lead when it comes to where brands invest in retail media. They must think you are really cool. I guess. But I have your deck. Are there any kind of key points made that you want to share with the broader audience here? Like, what did you focus on? Did you drop any hot takes? Any like crazy mic drop moments with the crowd?

Patrick Miller (8:02)

So first, like, wonderful folks to work with. Like, I showed up with crayon, like, you know, like Ms. Paint and like, they're like awesome designers and like made like it's the prettiest PowerPoint that's like ever had my name on it and like, it's a low bar, but like, they were just awesome to work with. So, you know, it was Fascinating. And just, you know, it's just like one, this, the scale of that team and how quickly they're growing was, you know, just like, wow. Two, the number of sort of industry veterans that like I got to see. And it was like every time I turned it like I think, you know, like I would know that many people and I was like, oh wow, like all these people that like I hadn't seen in forever. So like that was like really enjoyable for me. So I guess like, as I think about it, like a couple of things. One, you know, the durable advantages that Walmart has. Price, selection, convenience, you know, is, is, you know, is, is very true to, you know, their, you know, how much they're leaning into video and the, you know, the Vizio acquisition, which is absolutely the right strategy. And so, you know, it's sort of like, you know, how do you sort of pivot away from you know, the garbage that is so often the open web and even it's like there's a reason there are $2 cpms like and so getting brands to break up, you know, with this idea of chasing cheap inventory and flight to lots of conversation around sort of gen AI and LLMs.

Unidentified Host or Guest (9:16)

Got it.

Patrick Miller (9:17)

Boom.

Unidentified Host or Guest (9:18)

Now I have to think of the. I was like, oh, he's going to

Patrick Miller (9:20)

talk for like you were going to ask me a follow up question around LLMs and that's why I didn't get into the LLMs as much.

Unidentified Host or Guest (9:25)

Okay, yes. But I was like, okay, that's going to come in like 120 seconds at least. Okay. On, on AI and LLMs. We recently saw the news of like Walmart and Gemini partnering and essentially you can through Gemini and Google's Universal Commerce Protocol, link up your Walmart or Sam's Club and shop through Gemini and it actually has data on you and that makes the shopping experience much better. It's not just like random product outputs, but as we've talked about before, like Walmart is partnering and building when it comes to AI and this is another partnership. Is this the right move for Walmart? Thoughts on and then also like is it also the right move for Google?

Patrick Miller (10:02)

So it's a game of traffic first and foremost. So if I'm Walmart, I want to take all the traffic I can get. And so you know, I'm going to play nice with OpenAI and play nice with Gemini, et cetera. But I'd also like if I'm them, like I prefer the queries to start and the conversations to start at Walmart. And so, so how do you Differentiate there. And so, you know, it's sort of like, what information do they give to the hyperscalers in order to get good traffic? What do they hold back so that it is, you know, a better experience on the app? And so, you know, it's like I was doing some experimenting with like, using OpenAI or Gemini to, like, do research for my wife's 40th birthday. And then I was going to a Walmart app and doing sort of the same stuff. And, you know, and so it's like, queries were fine, you know, on the LLMs. And it was like, hey, you can go get this. But it was, it was clunky, it wasn't great, the integration wasn't perfect, but like, Walmart knew, like, my purchase history and Walmart, like, you know, has all that data at their fingertips. And so in certain ways it's like the, the LLMs are a feature of shopping as opposed to a sort of a product in and of itself. And so when I'm on the Walmart app and interacting with Sparky, and it's not just like, it's more than just the chatbot. And I think people are too obsessed with the stupid chatbot. And it's like, yes, that is fine, that is a thing. But it's also all the other surfaces across these apps that are leveraging, you know, the LLMs. And so like, whether it's, whether it's refinement, helping you decide, you know, et cetera, those become sort of featurized, prompting you of, like, what to hit. And like, think about this query. Like, all of those different surfaces are, you know, are then being developed, you know, by the retailers to make it a better and better experience, not just the chatbot as an end state.

Unidentified Host or Guest (11:37)

Got it. And that case should, in general, like, is the answer for retailers to really focus on their own AI services, not necessarily chatbots, but like, everything be owned. And then do you actually expect shoppers to adopt, using these individualized kind of retailers, specific surfaces over time?

Patrick Miller (11:54)

Oh, they already are. And I. But, but here's the thing. It's like, okay, if you go to a website, do you care if, like the backend is rds or redshift?

Unidentified Host or Guest (12:03)

No, I don't even know what that means exactly.

Patrick Miller (12:05)

Cause it doesn't matter. Like, you just want this stupid thing to work. And so it's like that LLMs are a. It. It's just another tool in the shed. It's a powerful tool, but it allows you to create sort of new features across the shopping experience. And so one of those features is a chatbot. It is not the only one. And so OpenAI is obsessed with sort of their weekly, you know, average users. Doesn't mean that is the end state. And so they, it is a product in and of itself. But there are also all these other features that the incumbents will use because they do not want, you know, OpenAI to take their cheese.

Unidentified Host or Guest (12:39)

Got it. Makes sense. Okay, last question for you on this story. If, if Walmart or Walmart connect both of them together, we're consulting you on a like forward looking strategy to how they like, like really, I know they're already at the top of all of the, all of the things but like really drive growth going forward. Like how would you build with, how would you build, what advice would you give them, especially with like supplier growth in mind?

Patrick Miller (13:03)

So one of the things I was like most like thrilled to hear about, you know, was, you know, one, talking to some of their colleagues from China and two, some of their emerging leaders, you know, about brands that start on, you know, Walmart.com and are then being picked up, up in the stores. And so the more that that and I was like, I was like, like, why aren't you all talking about this more? Like this is what everybody wants. And so, you know, it's the fact that they're doing it and, and delivering and getting these insurgent brands, you know, you know, within, you know, there are thousands of stores. That's the killer app. And so, you know, it's an and statement, it's online and brick and mortar and, and I think they're, you know, they're doing an excellent job but they're humble, quiet people. You know, they're not chest thumpers. And that's, that's a lovely part about their culture. And you know, they just get, they just do the work. And so, you know, I think sometimes though that they do themselves a disservice by not necessarily letting everybody know what they're doing on behalf of these insurgent brands.

Unidentified Host or Guest (13:57)

And I think especially living in Arkansas for a while, like seeing those insurgent brands come onto the shelves was. So there were ones where like I discovered the brand from Instagram or TikTok and then I see it on the Walmart shelf and I'm like, yes, this is also how you bring in. Nobody ever talks about like the young shoppers or Gen Z with Walmart, but I'm like, this is how you bring us in and keep us here, start keeping that kind of cycle going. Because I think Walmart, we always talk about like families and people that have to feed kids and rural shoppers, et cetera. But I'm like, no, the youth. You're doing something good here.

Patrick Miller (14:26)

It's simple. It's price selection and convenience. Selection matters.

Unidentified Host or Guest (14:31)

Yes. As it is. Add it to my tattoo list. Okay, Story number three. Amazon is shutting down Amazon Go and Fresh stores. They're expanding Whole Foods Market and focusing on grocery delivery innovation. Let me guess. You probably are not surprised that they are shutting down their Amazon Fresh and Go stores.

Patrick Miller (14:47)

Let me ask you a different question. Have you ever been into an Amazon Fresh?

Unidentified Host or Guest (14:50)

I did once when I lived in Baltimore. I drove, like, what took an hour and a half to wherever that one.

Patrick Miller (14:55)

There were, like, two, but D.C. suburbs. Yeah. I think Go is great, like, from a technology perspective. But, like, I can never figure out, like, how much of that was technology versus magic elves in the background, like, making sure I didn't, like, try to pocket candy bar or something. And so. But clearly it wasn't scalable. So I think that's why they. And it's just like, what's the reason to believe Leave there? And how is it additive to the customer experience? So, you know, so Go is, I think, a technological challenge. If they could have made it work at scale, I think it would have been amazing. It was a really cool experience. But if they can't, you know, whatever, kick the can down the road, Fresh, I don't know why it ever existed. It just, like, you know, it just. It just sucks as a grocery store now. That said, ordering on Amazon.com grocery now is amazing. And, like, I'm, like, addicted to it. Like, and, like. And I love Instacart, but, like, it's prying me away because, like, I'm going in and I'm like, like, I'm already, like, I'm already ordering diapers, right? And so then I'm like. And then it's like, oh, you can go get, like, asparagus, too. And I'm like, okay, Amazon, like, get me asparagus. A little bit of yogurt, and it just, like, shows up. And, like, I don't, like, have to deal with, like, the quirks of, like, the personality of a shopper. Like, Amazon just does it. And what I think is really interesting with, like, what they're doing is, like, they're making money two ways. And so, number one, they're saying, like, if I'm like, hey, I need cough syrup, and I need it right now. Like, I'll pay an extra 4 bucks so I can get it right now. I don't have to drive across town and go cvs. But if I'm also like, like, all right, yeah, I don't really have a cough. Like, I can wait like a week because I still have some more cough syrup and I can get these other things and then I can build an asynchronous basket and then I get 6% back on like my prime Visa card. They make more money because they bundle it into one trip. So they're, they're doing the magic trick that like nobody thought they could do, which is they can, they're breaking into the grocery category without it being dilutive to the overall retail business, which is absolutely stunning. And if I'm another retailer, I should be scared to death.

Unidentified Host or Guest (16:41)

Fair. My then follow up question is, why can't I order from like Amazon Grocery and Whole Foods at the same time? That drives me crazy.

Patrick Miller (16:48)

And give it time. But if you order Amazon Grocery right now and like something off of dot com, they come in two different trips. Yeah, but like, that's how they always do it. They'll lose, they'll gladly lose money, but like zip code by zip code they'll figure out like, hey, I can now bundle. And so this is the magic of the network.

Unidentified Host or Guest (17:06)

Got it? Yes. Okay, last question. I know if I'm shopping regular Amazon.com, and then I go to checkout, it will show me the screen that's like the grocery screen. And it's like, hey, do you want to buy more of the, like restock any of the grocery things? But if I start from scratch, I hate that like you then have to click into the grocery and then that builds a separate cart and it doesn't like connect back to Marketplace. And that kind of drives me crazy.

Patrick Miller (17:30)

It's clunky.

Unidentified Host or Guest (17:31)

And then, but then my follow up is like when you're in Amazon Grocery, I find the assortment of like brands to just in products to be kind of like bizarre and entirely random. And so my.

Patrick Miller (17:41)

Have you ever visited the marketplace? Like, it's bizarre and random.

Unidentified Host or Guest (17:44)

I know, but my groceries should feel less. I. I don't know. That's where I'm like, no, I want like the things you want, the creation.

Patrick Miller (17:51)

Well, but that's like the beauty of like good regional grocery. And so like, that's why like, like long term, I'm like, I ain't worried about Wegmans. They're gonna do just fine. You know, like, they have like, they have a customer base. They curate it. It's a nice experience. It's the ones I worry about is like a, like a Giant or like a Safeway where it's like, you know, sort of come on, like, kind of experience. And like, that's the stuff that, like, I think Amazon will really go after, you know, the boutiquey, nice regional will do just great. And same with the lower end. Like, I'm not worried about, like, Audi. Audi's gonna do fine. Trader Joe's is gonna do great. It's the mushy middle, you know, that is gonna get hollowed out.

Unidentified Host or Guest (18:27)

Actual last question, because I have to bring in something for brands. Thinking about this. Let's say you're like, standard CPG brand. How do you think about what Amazon's doing with grocery? Because I know not every brand that sells on Marketplace is also on fresh, but they could be like, how do you think? What do you do with this?

Patrick Miller (18:43)

So at first look at trade terms versus other classes of trade and like, where is it dilutive to the brand? Where is it not? It tends to be in grocery that it's more dilutive, which is, I think, which will stunt growth longer term. But if it's. But I would then compare it not necessarily to regional grocery, but to, you know, sort of Walmart and Costco when they're thinking through sort of their, their P and L and how much they want to drive. And then ultimately it is, you know, it is selection. And what are you, you know, what are you giving to Amazon? What is the form factor? How do you make sure that they get in stock? I mean, we've seen this. We've seen this play so many times before. You know, the laggards lose. And so if you want to bet against them, and that's fine, but, like, you know, good luck. And so, you know, I think especially in the, in the categories that have been and have, you know, they're late for penetration. So especially like anything fresh, anything frozen, if I'm those companies, I would double down in those categories because, like, it's never going to be. It's never going to be cheaper than it was yesterday.

Unidentified Host or Guest (19:40)

Got it.

Emma Erwin (19:40)

Boom.

Unidentified Host or Guest (19:41)

Thank you. Okay, with that, we can wrap up and we got. We got everything in there.

Patrick Miller (19:46)

See? Piece of cake.

Emma Erwin (19:49)

And that's it for this month's Commerce News. We'll be back next month covering everything you need to stay on top of it. I'm Emma Erwin, and we'll see you next time.