You May Also Enjoy: RetailPlaybook: Optimizing Amazon/Alexa, Walmart/Sparky and other Retail Shopper Agents w/ Andrew Bell
Greetings Jason and Scot show listeners! We hope you're enjoying the new pods and we have another Jason and Scot show on the way, but until then I wanted to share a new podcast we're launching called RetailPlaybook. Andrew Bell is the top expert in optimizing your product catalog and PDP for maximum sales on Alexa for Shopping (Rufus), Sparky and Target Shopping Assistant.
In his new podcast Andrew will be discussing all the news and best practices. If you are a 1P/3P brand or seller that cares about what Alexa, Sparky or TSA say about YOUR products, this pod is for you. There's also an accompanying substack newsletters at retailplaybook.ai. Subscribe to the pod and newsletter today and be sure to forward it to your team!
0:00 Greetings Jason and Scot Show listeners. This is your co-host, Scot Wingo. We hope you enjoyed episode 339, our special Don't Say Agentic Commerce episode, which was a big struggle for me. And good news, we already have episode 340 recorded and it is in post-production, so it's gonna be coming out soon. While you wait on episode 340, we thought you may also enjoy this new podcast called Retail Playbook by my friend Andrew Bell. Andrew is a top expert in agentic commerce optimization for AI-based retail shopper agents like Alexa for shopping, the artist formerly known as Rufus, Sparky on Walmart, and the Target Shopping Assistant, unfortunately nicknamed TSA for Target.
0:43 On this new podcast, he's going to be discussing the latest news, trends, and best practice playbooks for one P and three P brands and sellers on Amazon, Walmart, Target, and more. Give it a listen and subscribe on your favorite podcast platform. If you are a brand, uh, or seller that lists on any of these one P, three P platforms and this topic is of interest to you, you should also check out retailplaybook.ai to subscribe to the Retail Playbook Substack newsletter where Andrew will be publishing strategic playbooks and the tactics you need to optimize your strategy on Alexa, Sparky, TSA so that they recommend your products first.
1:24 Thanks, we'll be back soon and until then, enjoy this podcast. You know, good ACO, good agentic commerce optimization is playbook specific, right? Good ACO on Amazon means optimizing for Alexa for shopping while also keeping like the basic mechanics of like SEO and such. So it's like you're optimizing for two different layers, two distinct layers, but two… that have, uh… are in symbiotic union with each other. Uh, so there's just a lot of different dimensions, um, each with a different playbook. Same with Target, um, even going into like Shopify and stuff. Um, but I'd say the main ones of Amazon, Walmart, and Target.
2:05 I wanna pre- help prepare brands and give playbook specific practical strategies, uh, to equip them to be the best they possibly can on those platforms. Welcome to Retail Playbook provided by RefiBuy. Hosted by top retail agent researcher Andrew Bell, Retail Playbook is your go-to resource for cutting edge research, original content, and future-proof strategies to make sure your business is staying ahead of the competition Welcome to this episode of Retailgentic, and the first episode of Retail Playbook Today's guest is top Amazon agentic commerce voice Andrew Bell.
2:47 Uh, if our studio looks a little different to you, we are not Zoom, we're live here, uh, in person. It's actually the first time we've met each other IRL, so this has been a fun day. Andrew's down here in North Carolina visiting the team and sharing his Amazon knowledge. Um, we just announced him recently as our new VP of Research, uh, and Andrew is joining RefiBuy to help us get extremely smart about everything going on at the intersection of Amazon and AI, which used to be called Rufus and is now Alexa for Shopping. Um, also today we are announcing the launch of a new Substack and podcast. We call that a channel. So we have one channel that you know already, which is Retail Agentic, and now we have a second channel, which is Retail Playbook, and Andrew will be your host for that new podcast.
3:33 Andrew, welcome to Retail Agentic, welcome to RefiBuy, and congrats on the launch of your new channel. Yeah, it's good. Thanks for having me, Scott. I appreciate that. You've had a… You've got a lot going on. I do. You having fun so far? Yeah, I love it. Good. Uh, yeah, it's good. It's good stuff. Uh, uh, love, love the team, uh, everything we're doing here, and we're on the cutting edge of, um, everything. Um, I'm, you know, a big fan of agentic commerce optimization, um, and your, you know, philosophy around it, and I wouldn't be here if I didn't think so. Let's start by, since we have people's attention, I'm kinda mixing it up now that, now that we're into it. Um, let's talk about, you know, what your vision is for Retail Playbook. How do you- Yeah … um, what's this new channel?
4:14 How's it different than Retail Agentic, and what are you gonna focus on? Well, it's playbook specific, you know, to different marketplaces like Amazon because, you know, good ACO, good agentic commerce optimization, is playbook specific, right? Good ACO on Amazon means optimizing for Alexa for Shopping, while also keeping, like, the basic mechanics of, like, SEO and such because, uh, Amazon operates, for example, on what's called the A9 algorithm, which is their keyword search-based algorithm, and then it also works on Alexa for Shopping as well. So it's like you're optimizing for two different layers, two distinct layers, but who… that have, uh… are in symbiotic union with each other. Uh, so I think, like, that's one aspect.
4:56 Same goes with Walmart Sparky. They have their own algorithm when it comes to search, and they also have their own algorithm when it comes to, uh, Sparky and recommending products. Uh, and I imagine too, Sparky, you know, uh, joining forces with, uh, you know, when you're browsing through Walmart physically and, and Sparky being right there on your phone being able to, to talk and show you where products are, things like that. And of course, with Amazon's play, I think Alexa for Shopping's gonna help with the same thing once they, you know, build their super centers that we, we hear about. Uh, so there's just a lot of different dimensions, um, each with a different playbook. Same with Target. Um, even going into, like, Shopify and stuff. Um, but I'd say the main ones of Amazon, Walmart- And Target.
5:38 I wanna pre- help prepare brands and give playbook-specific practical strategies, uh, to equip them to be the best they possibly can on those platforms. Yeah, we call this strategy, um, agentic commerce optimization everywhere, or ACO everywhere. Yeah. A lot of people start, um, with the, uh, the exciting new thing, which is the answer engines, right? Which is what we focus on here at Retail Agentic and at re5 Buy. Um, but then what we've discovered is the, the retailer-hosted, uh, AI systems, uh, are very powerful because they're just sitting in that huge transaction flow. Uh, Amazon's like, what, 400 billion a year? Um- Yeah … that's not even counting the, the GMV that's like, that's actually smaller than it actually is 'cause they only count their revenue of the GMV.
6:23 That's their revenue number. So it's probably actually more like 600 billion or something. They're just a massive economic engine, and they've had some pretty big stats. Um, you probably know these off the top of your head, though. So Andy Jassy has been really promoting, uh, the artist previously known as Rufus pretty heavily. Um, and so what was it they have… I saw they have 100 million folks going through there. They just announced that. Yeah. Um, what else have you seen, um, as far as some of the stats they throw around about how big this is? I mean, it's, it's, it's giant. I mean, we're seeing conversion rates go up 60%, uh, overall on product pages- Yeah … in particular. Um, I mean, like you said, hundreds of millions of users now, now using that.
7:03 Um, ironically enough, I kinda knew that was, was the case because, uh, when I went to one of the engineer's LinkedIns, you go in, you look under their product, the description of their job, and it kinda shows some of their accomplishments, and one of them said, "Tens of millions of users responsible for that." So kind of an incidental detail, um, you know, that, that I kinda found out. But, you know, I'm, I'm not surprised that this is happening, and I see Alexa for shopping as a very, you know, contextually rich as far as behavioral data goes for buying. Um, whereas like ChatGPT, for example, doesn't have, simply doesn't have that kind of data. And so having that kind of vertical agent I think is very important and, and f- and bringing in Alexa too really brings in the horizontal side of it.
7:46 So now you have a horizontal vertical agent, right? You have that domain specificity and, you know, don't, you know… That, that's gonna happen too with them for music as well, you know, for grocery and things like that. Uh, but I think, you know, Alexa in different custom-built models, right? So, like Alexa for shopping's built on Claude. It's built on its own custom model. It's built on Nova, their own proprietary- And I think with each of these models, like, having a, um… The most important thing is having, like, you know, is your product being reasoned through, right? How is it being reasoned through, um, I think is, is very important.
8:28 What's your vision for the channel? Obviously, you talked about playbooks, but let's unpack that. So you've got a Substack. Um, in your mind, what- why are playbooks so effective at helping people understand and, um, and, you know, actually do things? Yeah. Well, why does, why did the Patriots always win the Super Bowl, right? You know- I guess you're a Patriots fan. Well, Tom Brady fan, so. Uh, but yeah, you know, having a specific playbook, you know, for l- let's say, defense against a really good offensive team, it's gonna look much different than a playbook you might have against, you know, someone that's better at running the ball or someone that's better at passing the ball.
9:08 Or I like the example of, like, think of the whole Jordan/LeBron thing, you know. Jordan just lived in a different era, right, where you had to shoot two-pointers all the time, basically. Whereas now it's more spread, and people are shooting three-pointers. So how you play defense is different. Before, it was man-on-man. Now, it's a lot of zone defense. So kinda in the same way, think of every marketplace as like, um, an era, right, uh, of basketball, where you, you're adapting a playbook for that specific time- Yeah … um, and, and for that specific place. What I like about your playbooks is, um, they have a strategic piece to them, but then they have tactical. Yeah. You know, it's kinda like, here's the big pieces of the playbook, and then here's, like, the individual things you need to do to be successful at this.
9:51 So the, so it's, it's hard to write content that's kind of both strategic and tactical, but, but these playbooks you write achieve both. As you know, like, I can be very dense in a lot of my, my topics. Uh, I go through all the theoretical aspects, but then I'm able to put that in practical strategy from all my background that I've had, uh, at different brands, looking at different ASINs and being able to optimize those, um- Uh, per, per platform, uh, whether it be Amazon, Walmart, uh, Target And on the podcast, what's your vision for that? I think being able to bring people on that have hands-on knowledge like I did, uh, hands-on, uh, practical knowledge of, um, optimizing for products at different levels.
10:35 Um, and so bringing on people in the Amazon world that have this domain expertise that I know, right, uh, to help inform our guests of, you know, what they can do, what's out there, how the Amazon world is vastly different from, um, from like, you know, showing up on ChatGPT, for example, or showing up in Claude. Um, 'cause even in ChatGPT and Claude, I mean, there's distinct… You would say, okay, yeah, there's distinct playbooks. But, um, going into, uh, you know, going into Alexa for shopping, you know, there's an even more specific playbook because unlike like with Google, Google for the longest time, you know, started, you know, going more semantic before Amazon did, right?
11:16 You know this as well as I do. Yeah. Um, but Amazon has stayed relatively keyword-based in a lot of ways And so being able to optimize your products in such a way that's not irresponsible. Um, and so, like, on Amazon, you wanna be able to balance SEO, you know, what's it called, GEO and AEO, really, which honestly, those three come together and form agentic commerce optimization. Mm-hmm. Good ACO on Amazon is good SEO. So we're doing this kind of Christopher Nolan style now. Let's go. You obviously know a lot about this, so let's kind of figure out how you, how you got here. Um, tell us a little bit about yourself, um, and then how on earth do you know so much about, uh, how you optimize things for, for Amazon, Walmart, and the like?
11:57 I didn't even major in business, Scott. I, I studied ancient Greek and preaching at a Bible school. Um, but if there's one thing I have, it's a, a self-taught spirit. Um, I spent my waking days in Barnes & Noble, uh, six months, and would read every book off the shelf in business, and then I discovered online PDFs, and that kind of changed the game for me in terms of education. Learned everything I could about the business world. Ended up joining a company called Touch of Class, and, you know, was able to take them from zero, uh, $100,000 per month to, uh, 4 million in a, in, in annual, uh, revenue. Boy. Yeah. What category were they in? Luxury home decor. Okay. Unique luxury home decor.
12:38 Covered about 30 different categories within home decor, and personally, let's say by hand, optimized 1,500 product detail pages. And then I was able to take that philosophy of product detail page optimization, put them into GBTs, or let's say at the time it was prompts, and when the GBT store came out, I was one of the first ones to put my GBTs out there, and all of a sudden, thousands and thousands of sellers started using. Um, even tens of thousands of sellers started using my GBTs. Uh, but then I began to see a picture of, like, building something so much more robust and something that has a lot of data in it as well.
13:19 Um, so I, I think there's that aspect, and it gave me all the experience I needed, too, for, like, you know, creating ad campaigns. I've created, like, 1,000 ad campaigns across sponsored products, brands, display. Uh, so basically, I'm, I'm going through this phase of building up my domain expertise within Amazon, and within Walmart, too. Mm-hmm. Right? But Walmart has always lagged, been a laggard behind Amazon in terms of the tools that sellers are given, vendors are given, right, to, to be able to, um, properly market the product. Was Touch of Class 1P, 3P, or both? 3P. 3P. Got it. Yeah, 3P. Yeah, a 3P company. Yeah. We should say, everything, um, that we'll be covering here on Retail Playbook will be applicable to 1P and 3P.
14:02 Yeah. Um, and so the audience is merchants of all size, so, and shape. So big brands that don't sell direct, that they're just wholesale through, uh, a retailer, all the way to smaller sellers that are selling 3P. Hopefully there'll be a little bit of something for everybody. Absolutely. I, I think if we have to pick one, it's gonna be those brands, um, 'cause that's where we're getting the most, uh, questions, uh, from people of- Yeah … they're, they're really leaning in and trying to figure out… Um, sometimes they have an acute problem, like we've seen… Um, I've seen one where, uh, I can't remember which engine it was, but it was basically saying some chicken nuggets were vegan, for example. And obviously not right. Um, or there was one that said, um, that a, a drink was kosher, and it's not.
14:46 Um, so, so sometimes the engines do pick up some, you know, weird- They do ideas about what's going on in the products, and that's the kind of thing… That's part of optimization, is making sure it's not saying wacky stuff, but there, there's much more to that as well. Yeah. Um, af- after Touch of Class, uh, then where'd you go? Uh, Touch of Class, I was a senior account manager for a company where I managed a number of different brands, but then I got an offer from NFPA, is the National Fire Protection Association, and they're a, you know, big brand looking to get, you know, sales on Amazon. And was, um, you know, they had a little bit of a more distinct problem with, uh, you know, unauthorized, uh, resellers. And so I was able to get, uh, build relationships with high-level Amazon contacts, and, um, essentially, uh, get it done and, and, and make sure that, you know, we had the proper content.
15:35 Uh, we had the, um… You know, we, we were of course in Amazon Brand Registry, and we decided that, like, we need to get rid of these unauthorized resellers. And, um, I can't go much more into that- Mm … uh, with, with, with this part, but I, I can tell you that it's led to an exponential amount in sales, and our buy box increasing threefold. Uh, which it, which is very essential, especially for, uh, 3P. But all the while this is happening, I have my own consulting gigs where I worked with 1P companies, uh, seven, eight, nine, and even a couple 10-figure grants. Yeah, and then so you started doing your GPTs, and then you've started, um, writing the playbooks, and you have a, um, you have a LinkedIn, um, newsletter called S- Amazon Science or Science of Amazon?
16:20 You know, it started Amazon Science Made Simple, but then it quickly became… I should've called it Amazon Patents Made Simple, 'cause really I unfold a lot of patents. But here's the thing, I- science papers also inform, uh, patents. Amazon science papers inform patents. Patents I consider a main source of truth because that's actual deployable technology from Amazon. Amazon science papers point to the future of Amazon, of where it's going, right? But you can catch onto cues. So, like, for a patent that recently came out, there was a paper done on it in 2023, right? That data supported what was in the patent today. So it comes to show you there's a two to three-year gap between a patent, um, and research done.
17:04 Yeah. And- So research first- Yes, research first … uh, and then patent. Does research always turn into a patent, or is there a- Not always. No. Especially if it failed. You know, most, most science papers, I have a, you know, levels one, two, three, four, five. Mm-hmm. Uh, five is completely deployable, uh, it- or it has been deployed, and it's showing data results from it. Uh, so I look at it from the perspective of is it level one, two, three, four, five? Most are level two and three. Most are level two and three. So when these patents come out, you're actually poking at the user interface to see if it's live or not. 'Cause I imagine some patents, they never… You don't have to implement a patent. It can just be an idea you wanna preserve and, and- Yeah … protect IP on. Right. Yeah, absolutely.
17:44 What percent of patents get implemented? Do you have like a rough estimate? Probably couldn't give you a rough estimate, 'cause I only look at the ones I know for sure implemented. I look through every single patent myself. Um, I don't read every single one of them. I have a method I go through, but I do look through all the figures that initially come up. Mm-hmm. Um, I have, I have a… I don't wanna… I have a… Between you and I, uh, if you cut this out, I don't know. But, uh, every Tuesday is, uh, patent Tuesday for me- Ooh, okay … where every accepted patent, actually granted patent, comes out, and then every non-granted patent comes out on a Thursday. And that's… I look at them between then, uh, Walmart, which is intensifying their patent production, whereas before they didn't have that, that level of production.
18:26 Um, but yeah, with, with Amazon, you know, what's, what's really important is, like they… there's nobody, no company under the planet, maybe Google, that has as much research and has as many patents going out. Um, and I think that predicts… Patents are Amazon's playbook of the future, right? Detailed explanations of really how engineers think through problems, too. Um, 'cause you have to get the wording right as well. And it… sometimes it can be broad, too, but, um, I think, you know, move it… We see ourselves moving toward use case-based language all over in product pages, right across the title, across bullets, across A+ content, across, uh, videos even.
19:07 Um, and even though videos aren't technically indexable, they are becoming that, and I tend to predict that. And, um, you know, one example of it's the playbook of the future is if, you know, a year ago, I actually predicted this, Scott, that, uh, you know, Alexa and Rufus would basically fuse together into something, right? Rufus' product intelligence with Alexa's, uh, superior personalization abilities as well I think it's the exact opposite It's the exact opposite I'd love to take that one down.
19:49 It's because it works. Yeah. There's a reason why… If anyone's tracked Rufus more than anybody, I can guarantee you that being on a, being on my phone 30 minutes every day looking and testing Rufus, um, sometimes two hours a day if I'm really doing heavy research, I can tell you that it's, it's increased its coverage everywhere. Yeah. Right? Instead of less better- Yeah … I would use it before. First it was on the, you know, uh, we had it in the product page, then it started coming on search results page, then it started coming on the home page, right? We're seeing it everywhere. In fact, there's three placements of Rufus on the product page itself. In the mobile app, it's got the star position in the bottom right where your thumb is. Yeah. Right. Exactly. So, so that's, that's the pole position. That's the place to be. Yeah. Yeah. And even now we, we see it triggering in, in search results too, uh, across many queries, usually, uh, more broad, nothing real specific yet, but I, I anticipate this happening and so looking at the, the digital shelf now, um, seeing it as, you know, uh, where, where you're ranking.
20:46 So like with Alexa for shopping, you know, yes, it pulls in search results from A9 algorithm, right? Products, you know, you have the normal ranking, but after it gets that pool of products in there, it has its own ranking criteria based on context of the prompt, context of the user, and personalization from Alexa, uh, to be able to put the proper products in there. So you might have a pool of 100 products from search results, right? But then it narrows it down to 30 or 20. But all we know for sure is it, it, it then produces the best eight. Um, about the best, the best eight. Uh, about five to eight products. I think that we may have a world where the, um, the search engine kind of melds… It, it kind of fuses, to use your word as well, where you no longer have, like, Rufus and the old search.
21:35 I think the old search doesn't get replaced, but increasingly Rufus handles- It handles a lot … a lot more of it. Yeah. Because that's the biggest problem. I mean, think of the mom, um… I use my mom, for example, who, you know, I have diverse interests, my, my, my brother-in-law has diverse interests, my sister does. My sister had, like I said, certain kind of makeup, so she has to search that brand on Amazon, this particular query. "Oh, uh, but I need it to be cruelty-free," let's say, right? And then me, I love Legos, but my mom knows I like Marvel Legos. And so instead of us, her having to type in all these search terms- Alexa for Shopping knows to do that search query plan, to go and perform the shopping mission- Mm-hmm right? Um, to, to do those search, uh, search queries, um, and to, to, to grab the product results from there.
22:20 Ruth, uh, Alexa for Shopping. It's gonna take me years- … to stop saying that. It's- I don't know how you do it. Uh, in some ways it's actually better at… So there's this set of people I argue with, and they say, "No, agentic commerce is you don't touch anything." And in some ways Amazon's closer to this full autonomy because they have that feature where, um, you can basically set a price trigger, and in there you can say- Yes … "When it hits this, buy it for me." Yep. Um, and there's no, like, six a- you know, "Are you sure? Are you really, really, really sure?" It just fires, uh- My wife is actually tired of all the notifications on our Alexa- Yeah in the bedroom because every notification is, "Hey, you have a buying, um, uh, alert," you know? S- something's on sale. Yeah. Yeah, 'cause you just signed up for so many of them.
23:02 Yeah. Yeah. Hopefully you didn't set the automatically buy on all of them. No, not yet. Um, and then, uh, you were one-- the first person I know that discovered Sparky was live on the web, on just… I think that was… Was that this week or last week? Oh, blur- It was … blur together. Um, so, uh, w- what are some interesting things that you've noticed about Sparky? Uh, Sparky you can… It's how it's not near as sophisticated in terms of the data it's bringing in. Um, it hallucinates a little bit more than, than Alexa for Shopping does. Um, and I think a lot of it has to do, again, you just… t- to look at the, the power of a company, you look at their patent architecture, it's just not near as sophisticated as, as Amazon.
23:43 And you can tell what they're doing compared to Amazon is, um, relatively primitive, um, in, in comparison, which makes sense 'cause they don't even have, like, a huge science division like Amazon does in terms of, like, search science, right? Um, but at the same time, it's increasing because of their partnerships with, you know, for example, with OpenAI. Um, and I think with their willingness to adopt the universal commerce protocol as well, uh, very clearly. I think that's, you know, gonna give them a, a slight advantage, and Sparky being implemented within there. And I imagine Sparky GPT, um, is being used behind Sparky. Um, the best evidence shows that GPT-5.1 at scale is the best shopping. I don't know, you know, it c- could be that still.
24:25 It was a month ago. Yeah. That's, that's browser-based kind of stuff, right? Now browser-based. Yeah, that's browser-based or action-based, yep. I have a theory about why Amazon, Amazon joined UCP, and I think our theories differ. What's your theory? Yeah. My, my theory, um, is so that, uh, products from Alexa, you know, would- It, it could be relative to, okay, Sparky now is gonna go into OpenAI. Well, now Alexa for shopping is gonna go into OpenAI, and that's, that's very possible. But if you saw the recent title change, I think this could be Amazon's play at UCP to… Because in the, uh, email, it said not just to be readable on mobile, but to be consistent with other online stores. So having that consistency is a big level, I believe, when it comes to UCP and providing the data they want.
25:10 So they, they don't have to, um, provide all the ingested data to those feeds- Right … like title, bullet points, A+ content. They can give them what they want. They can give them the title and the item highlights, which they deem, okay, that's enough for them to have. So you think… So you're seeing that as a signal Amazon's gonna publish this stuff out- Yeah basically. Yeah. Yeah. All right. So mine's, uh, this'll be fun. We'll see who's right, um, who's the better predictor. Uh, I have more years at this, but then you have more Amazon knowledge, so we'll see how it goes. The, uh… So I think, uh, there's this famous, uh, I call it the Death Star diagram, and it's the flywheel diagram where, um… And the, the industry lore goes that Jeff Bezos drew it on a napkin diagram. And, and this napkin actually exists somewhere at Amazon.
25:52 I haven't seen it, but it's, uh, supposedly framed. Um, and in there it talks about, uh, it was the whole idea for the marketplace that was kind of born. They started working on it in '04. We, we had access to it at ChannelAdvisor early. Um, and then in '06, they really leaned into it. And, um, it basically says… Let's see if I can do it from memory. So, uh, in the middle is the customer, and then it has, you know, more s- more merchants bring more selection, which means… And then there's a spur that kind of like they compete for lower prices. Um, and then that, that improves customer satisfaction. Mm. So, so it's kind of around this idea that consumers love, uh, selection, convenience, uh, and that's kind of like Prime brings that from the side- Mm … uh, and low prices.
26:33 Um, and those, those are things that they never get tired of. Um, so embedded in the DNA in, of Amazon is they worry a lot about selection, and it always drives them crazy when a brand won't sell on Amazon. So, so kind of a famous one is Nike's been on and off, and that, that drives them crazy. Mm. Um, and then, uh, another big one is Birkenstock. Uh, they, they've always kind of had… And there's a knife company. There's some kind of famous ones. Mm. So I think that they're, they're worried that they see Gemini coalescing, Google coalescing, this kind of Battlestar Galactica against, or the rebels against their Death Star, for lack of a better- Yeah metaphor. Same, same thing. And they're starting to see, well, there's a big pool of inventory out there that we can't get to, 'cause there's, like, another one that's kind of famous is Lululemon has never sold on Amazon.
27:21 Yeah. Now, you can find it on there through occasional third-party sellers, but it's not like- It's not- the brand and an official- There's no brand registry or anything. Yeah, yeah. So I think, I think what they're doing is, I think Shop Direct is them defending against that, to your point about the playbook. Yeah. And I think they're saying, "How do we-" How do we pull that part of the Venn diagram that's not in our Vi- Venn diagram into our Venn diagram? And they have the Shop Direct, which is this new… We always had 1P and 3P and different flavors of FBA and Prime eligibility inside of those two programs. Well, now there's this third program, and it kind of doesn't make sense to some degree because there's a world where it's, like, actually the most attractive program, so every brand should be doing it.
28:04 Because you don't pay Amazon anything, it's free. Yeah. So that's pretty good. You can set the price, which a lot of vendors don't like in 1P. Um, uh, and, um, you can even have, uh, there's an logistic checkout. It's a small subset in the Buy For Me. But, but I think that's what they're trying to do, is surface inventory not on Amazon inside, and I think they'll have you buy that through UCP. So I think they're gonna-- they're- they'll have you be able to buy that in Amazon through UCP, but they're gonna in- they're gonna flip it from kind of what you're saying. Yeah. They're gonna pull UCP into Amazon. Into Amazon. Whereas you're kind of saying publish out. Now, the other option is they can do both, and we could both be right, so we'll, we'll see. I'd say, yeah, I see what you mean. So it's all consistent with online stores, and those, those are-- those show up too, so, uh, the difference is the same.
28:48 And it comes from-- And I think their other fear, and this kind of is my, my thinking around the Alexa thing, I think their other fear is being attacked from the sides because, um, ChatGPT just released some information that 20% of their conversations have, um, uh, a commerce intent in, in them at some point. Yeah. But when you look at this, uh, when you look at the data, they hardly have-- only, like, 2% start there, which is really interesting. So but what'll happen is someone will say they have-- people have conversations with ChatGPT they would never have with certainly, um, A- Amazon, um, and maybe even Alexa. Like, I don't use my Alexa for general knowledge. Um, you ma- You know, everyone's use differs. I've just kind of-- It was so bad at it, I've been trained not to. Alexa Plus is pretty good, but- It is, yeah.
29:30 But I think there-- I think the… But anyway, so, so- But mass adoption-wise- Yeah … I agree with you. It's kind of like music and shopping list- Yeah … and checking status orders is, like, what I think most people use it for. Um, maybe the weather or something like that. Um, and so, so I think they're worried about, uh- Not even having an attempt, a shot at… If someone's having a conversation about a home improvement project with, with ChatGPT, uh, they may, they will have the first shot at putting a product into that conversation should it go there. Whereas Amazon may never see that, and I think that's another thing they're worried about. So I think they're gonna start pushing Alexa further into those informational queries, and therefore having one common brand across all that, um, will actually attack them from a more horizontal attack.
30:17 Yeah, I think that's spread out, because they already got the vertical angle. I think they're superior when it comes to that. Uh, but as far as the d- now that depends what, uh, kind of data Shopify too is willing to share. I think that's one thing to consider as well with ChatGPT. Um, but I think overall Amazon has the best shopping domain-specific expert, right? Yeah. And Rufus or Alexa for shopping, that is. So you're… Let's go back to Retail Playbook. So, uh, you're having, uh, uh, uh, by the time folks see this, your first substack should be out. What, uh, give us a little taste of what you're gonna be writing in your first substack. Well, uh, Alexa for shopping. Uh, if you notice, I, I published the Alexa for Shopping, uh, Blueprint- Yep uh, part one, and, uh, this is gonna be part two of that, where I'm gonna go into actual practical strategies that you can implement there, and how ACO, uh, makes sense within Amazon framework.
31:11 Does 2.0 build on 1.0 or, or is it kind of like a, "Hey, here's the new… This is a replace-" Is it more of a replacement? Like- Oh, it builds on. It builds on. Okay. Yeah, it's way more practical. It's, you know… I was thinking about it coming out with certain custom GPTs as well to go with it. Um- Uh, you know, it's basically turning the philosophy of those, those papers into that. Yeah. Eh, so part two will focus heavily on strategy, um, with accompanying GBTs to go with it. Um, I'm gonna get into the topic of what's called query planning optimization. So we talked about how, you know, the burden of search reformulations where my mom's having to type in a bunch of different searches to get products, instead Alexa for Shopping is doing that.
31:53 So it's not so much about are you showing up on the search term, but from the intent and context of the prompt, are you showing up across all of these search terms combined together, right? Are you within that, that shopping mission? Cool. So we were super excited when you accepted the offer. Uh, it's been fun. I think we're three weeks in, and I've already learned a ton from you. Hopefully you're having fun as well. Um, I know you're, like, wound up like a snake ready to strike 'cause you've been wanting to get this channel live and you're so excited, so- Yeah uh, congrats on the launch of that. So that, that's awesome. We're, we're real excited to have you. Um, if folks wanna follow what you're doing, what should they do? Uh, well, subscribe to my, my Substack. Um, and then also you can follow me on, on LinkedIn.
32:35 Don't forget to subscribe, you know, to the newsletter there. I post regularly updates on Amazon Rufus, and if you, if you want an accurate answer on everything Alexa for Shopping, uh, you know, go to my, go to my DM, ask me a question. Uh, but yeah, Scott, super excited to be a part of, uh, RefiBuy, uh, which is the future of… Well, you coined the term, didn't you? Agentic commerce optimization. Yeah. Uh, so, you know, a lot of the, you know, part of the reason I'm here is 'cause of you. Um, and I, I believe in your, your philosophy and I, I believe it's the right one based on all the research and everything that I've done. So reminder everyone, let's, uh, help Andrew launch this channel with a lot of success, RetailPlaybook, uh, either .com or .ai.
33:16 Those are both live. Subscribe to the Substack, subscribe to the YouTube. Um, go subscribe to the LinkedIn. Uh, and, uh, we look forward to a successful launch and seeing what kind of awesome content you have. If, if optimizing Amazon, Walmart, Target, or any of these things are important to you as a brand or a third-party seller, you came to the right place. So sign up today. Uh, and thanks, Andrew. Look forward to seeing what you do with the channel. Thanks, Scott. Appreciate it.
Overcast
PocketCasts
Amazon Music
Audible
Youtube Music
RSS Feed