Amazon News: Help Me Decide, ChatGPT Traffic, and Holiday Rebound

Amazon News: Help Me Decide, ChatGPT Traffic, and Holiday Rebound Unknown Speaker: Welcome, fellow entrepreneurs, to the Amazon Sellers School podcast, where we talk about Amazon and how you can use it to build an e-commerce empire, a side hustle, and anything in between. And now your host, Todd Welch. Speaker 2: What's going on, everybody? It is Thursday this week, switching it up from Friday. We're gonna be trying out this Thursday slot going forward and seeing how it goes, but appreciate everybody out there joining me. We got John Aspinall from Trellis with us today and Stephen Pope from My Amazon Guy, of course. Speaker 1: Good afternoon. Speaker 2: So gentlemen, appreciate you joining me. Speaker 1: And happy Halloween. Speaker 2: Yeah, Halloween, the day before Halloween. Wanted to switch it up, didn't want to compete with Halloween, you know, everybody out there trick-or-treating and everything, but we were talking beforehand, all the kids tomorrow are going to get done with school and then get all hopped up on sugar and then that'll be a fun time trying to put them to bed. Speaker 3: Bedtime, forget it. I'm not even going to try. I'll just wait till they crash out and then deal with it on Saturday. That's how I'm doing it. Speaker 2: Yeah, at least it's a Friday so you don't have to worry about getting up early on Saturday for sure. We've got a good show today. I'm starting to think we should change the show from Amazon Seller News to AI Seller News, perhaps, because a lot of the major news lately has all been around the AI world, and today is going to be no exception. Speaker 1: 14,000 or 30,000 layoffs. I keep seeing both numbers yesterday, citing AI. Speaker 2: I got a feeling that was kind of the excuse right now because AI has come a long ways, but it's not there yet. Speaker 1: My favorite though is the Amazon post record profits and then they're like, oh, we need to lay off 30,000 people. Speaker 3: But the funny thing is, I actually know someone that was affected by it, but they didn't work. They weren't like an AE or a dev for Amazon. Speaker 2: They worked at Twitch, which I guess- Yes, I think the same guy, I think. Speaker 3: Yeah, yeah. Speaker 2: So I was like- Yeah, it's interesting. It's hard to say what it means, if it means that Amazon is struggling in some areas or if it truly is AI and they're just getting prepared for it or what's happening. Speaker 1: I mean, the singularity's coming, no question on that. Timing, TBD. But there's no reason that it was AI in 2025. AI has not made anybody more than 13% more productive. That's all the studies I've seen. Speaker 2: Yeah. The fact that I cannot get an AI to find articles for Amazon News that are actually new. I'm like, give me articles in the last seven days. And it's like, here's one from 2024. Are you using perplexity? I've tried grok and chat GPT. Perplexity. Speaker 3: First of all, Stephen, who turned you on to perplexity? Speaker 1: John Aspinall, because John is the AI, you know, God. Speaker 3: So, the thing is, right, my whole thing with AI, there is no Swiss Army knife. Like, people like to think ChatGPT is a Swiss Army knife, and it kind of is. It'll do everything a little bit okay, but every single one has its purpose, right? And perplexity, bar none, is, like I was making fun of Stephen before, I was like, don't Google it, it's perplexity, right? And that's where you're getting... What did you say? Speaker 1: It sounds weird to say I perplexed it. Speaker 3: But the prompt you could put in perplexity is what I do for a lot of folks and a lot of my topics and content is within the last 24 hours or within the last 72 hours, give me news that's breaking, but then I could also drill down in perplexity, whether it's just web search or I can go even more further and say, just social. And with just social, it only mainly pulls from Reddit. And I found that that's where you're getting like the real kind of news news. Unknown Speaker: I'll give it a try. Speaker 1: When ChatGPT stopped crawling Reddit, that was funny. Speaker 2: Yeah, I've literally got it in a tab here, John, from last time you were on the show. And I've been meaning to jump into it, but I have not. Speaker 3: Hey, that's a good thing if you use Perplexity's Comet browser. It'll organize your tabs for you. So literally, you can use voice mode in Comet and you can say, organize my tabs for me. And it'll group and automatically close out tabs you haven't used, group things together. It's fantastic. Speaker 2: Comet browser. Speaker 3: Yep. Speaker 1: You'll be on a waitlist for like three days and they'll tell you it's going to take two months and then you'll get an invite three days later. Speaker 3: It's open now. It's fully open. Speaker 2: All right. The browser that works for you. Yeah. Download a new browser from Perplexity. Okay. Well, cool. Speaker 1: The coolest thing about it though is the Chrome integration. It is built on Chrome. Perplexity went over to Google and offered them like, I don't know, $36 billion to buy Chrome. And then like you're up and running in less than two minutes with all of your Chrome shit in perplexity. And then the AI is integrated. So it's best in class. There's definitely some privacy concerns, but like we're all heading towards the singularity. There will be no privacy anyway. So I just stopped worrying about that. Speaker 3: Yep. And like ChatGPT came out with their ChatGPT Atlas, which is their version of an AI browser. But again, I default to When I'm using a browser, like if I was to use Chrome and I would default to Google search, right? I want to default to perplexity search and it's truly the most agentic of the AIs. And I just, I'm not a fan of ChatGPT. So if you have the Atlas browser, if you type in just the search that you want to get like a typical Google search from, you would get a perplexity search. With ChatGPT's Atlas, it'll just literally do a ChatGPT query. And it's like, I don't want to know this crap. Speaker 2: Really? Okay. Speaker 3: Yeah. Speaker 2: Well, I mean, we might as well. That was one of the news articles that I had here for us to talk about. So let me read the summary real quick. OpenAI launches ChatGPT Atlas Browser to challenge Chrome's dominance. OpenAI has released ChatGPT Atlas, an AI-powered browser that embeds ChatGPT directly into web pages, allowing users to summarize content, compare products, and automate tasks without copy-pasting between tabs. Designed to make browsing conversational and context-aware, Atlas positions ChatGPT as a full digital assistant integrated into users' daily workflow, Challenging Chrome's long-standing dominance. So, but you're saying, John, that Comet is even better than the new Atlas browser. Have you played with both? Speaker 3: Yeah, I played with everything. Speaker 2: Right. Yeah. Speaker 3: Um, that's what she said. No. So for me, like, I'm big into perplexity, right? Because so I've used perplexity before Comet, just because it's it's beating Google. It's like this, the CEO of perplexity, his mission is to beat Google. Like, that's why Stephen was saying they put out the bid to buy Chrome, because even in Chrome, Chrome has Gemini built into it, but only rolled out to certain users. So it's not an alpha. It's something like beta, beta. And even then, it's not... Speaker 1: Alpha comes before beta. Speaker 3: Alpha. Oh, okay. Whatever. It's not rolled out. It's beta, it's double beta, whatever. So the thing is, I was invited to Perplexity, Comet. Six months ago as a beta user, right? And then it slowly rolled out with invites. But this is like a day late, a dollar short. There's Dia browser that's been out before this. There's Comet. There's all these other agentic browsers. And to be honest, it's going to be adopted by People like my wife, people that just use it for basic stuff because that's the gateway drug into AI because ChatGPT is so easy. Now, the winner of the AI browser kind of race is going to be the first one to put it on mobile. As of right now, none of these are on mobile. So the first one to put it on mobile will win user base. Speaker 2: Yeah. Yeah. And if history is any indication, the best rarely wins the race in the long term. I mean, if you look at browsers, Firefox was by far the best browser before Chrome. But of course, Chrome won out because they had the marketing and the power of Google behind them. Betamax, I think, was competing with DVD, right, if I remember correctly. And Betamax was by far higher quality and better. But DVD got out there first and got a lot more marketing and everything. So we often see the best lose out. Could that same kind of thing happen in this race, the future of AI, where maybe perplexity is the best, but do they have the best marketing? Because that's really what it comes down to in a lot of cases with these kind of things. Speaker 3: Well, the thing is, too, who is the ICP, who is the user, right? So if you're talking about just mass public, mass world people, Chrome is always going to win. I see people all the time in the business, clients, other solution partners, And they share their screen and I'm looking at a Chrome browser. I'm like, oh my God, why are you using Chrome? And it's always outdated. It's like update Chrome in the corner. And if you're using, like I'm right now on a desktop on a Mac mini, I don't care about battery power and usage, right? But Chrome sucks RAM and it sucks battery usage. Or if you're on a MacBook or a laptop, it's not the best option that's out there. So if you're already in a suck up RAM, suck up power, why use that when there's other options that are out there? Speaker 2: Yeah, I am using Chrome right now, even though I don't really want to just because everything works in it. You know, all of my plugins that I need for Amazon and everything else. Speaker 1: Yeah, it's the exact same browser, just better. Speaker 3: So what was it literally google chrome or browsers like the arc which was before the perplexity comet they're all built on chromium that's like the foundation of chrome so chrome actually. Source is out or it's open whatever it is of the foundation of what chrome is so every single chrome extension anything like that literally if you're in perplexity comment and you go open up the chrome store it is the google chrome store you just adding extensions in there. Speaker 2: Yeah so i was using the brave browser if you guys are familiar with that it's kinda like the privacy browser. And most of the plugins work in there, but I was running into just weird little things sometimes that I would jump over to Chrome and then it would work. Speaker 1: We'll have this issue with Comet. Speaker 2: I'll have to give it a try because I don't want to use Chrome, but I've just been using it because it just works for everything. Speaker 1: The biggest reason why people don't shift off a product is because of how hard it is to leave it typically, and in this particular instance, it is instantaneous. Every extension auto-installs, your profile automatically comes over, everything, bookmarks import, bookmark bar, all that stuff. Speaker 2: Well, when we get done, I will be downloading it and trying it. And I'll also sign up for an account with Perplexity this time. So next time you're on the show, John, I can tell you if it is. Speaker 1: As a news guy, I'm just surprised you're late on Perplexity, to be honest, Todd. You should have been an early adopter. Speaker 3: I pay for Perplexity Pro, which is like the run-of-the-mill $20 a month plan. They have the Perplexity Max, which is like, I think 200 bucks a month. I don't recommend that. The only value you get with that one is you get, so if you were to go, like Todd, go to like Perplexity, just regular Perplexity, not the Comet, like just Perplexity AI. Speaker 1: Top one. Speaker 2: Yep. Speaker 3: So you'll see if you click on the globe icon right there in the search where you're going to ask anything, like that little globe. Speaker 2: This one? Nope. Speaker 3: To the right. Speaker 2: Right here. Okay. Speaker 3: So you can change inside here. Maybe it's not the globe. Maybe it's something else. No, no, no. I mean changing the LLM model because you're not signed in with an account. When you sign in with an account, basically what Perplexity has is all of the LLMs built inside of it. So that's why I really like Perplexity too is because if you were to buy one, if you only had $20 a month to spend on one AI, I pay for them all because I'm silly like that, but if you only had one, Perplexity allows you to access- Pull my screen up, Todd. What's up? Speaker 1: Todd, pull my screen up. Speaker 2: Pull your screen up. All right. Speaker 1: Keep talking, John. Unknown Speaker: I'm just going to show what you're talking about. Speaker 3: Perplexity allows you to have all the LLMs inside of it. Perfect. You could do Sonar, which is their model, which they run off of whatever they want to do. But then you could see there's GPT-5 from ChatGPT. There's Claude Sonnet, which is a powerful one. There's Gemini. There's Grok. But then for Opus 4.1 and then O3 Pro, which I don't know who the hell uses that. That's where it says max, where you have to have the max plan, but you don't need that. The fact that you can get Claude's SONNET 4.5, which is their latest and best model inside of this means you have access to query like SONNET, like Claude SONNET, write like Claude SONNET, all from Perplexity. Speaker 2: Interesting, okay. Speaker 3: Yeah. Speaker 2: So is that, they're just using that in partnership with ChatGPT or they're just using their APIs to query through them? Speaker 3: Oh no, they definitely pay for like all those connections and stuff like that. Speaker 2: Okay. Yeah. Interesting. Well, I will definitely play with it. I just signed up with an account. So next week we'll see if we can generate the news through perplexity. That'd be cool. At least to kind of go along with what I do manually and find stuff that hopefully I haven't been finding. Speaker 1: So here's a little demo on Amazon News past 24 hours, see what it comes up with. Speaker 2: Yeah, there you go. Speaker 1: 14,000 jobs, Q3 earnings report, record earnings, $177 billion by the same time they announced the layoffs. AI-powered shopping features, which is what we're talking about today. The Help Me Decide feature. So pretty much nailed it. Speaker 3: But the best thing you could do with this, like this is high level, right? You can force it to go further. If Steven was to say, give me the Amazon news the past 24 hours for 12 topics or for 10 topics, right now it's doing high level, but if you tell it specific, that's where prompt engineering is like a real thing. If you tell it specifically, then it's going to keep on going to dig, dig to things that might be a little bit obscure, but still relevant to the query. Speaker 2: Yes. Yeah. And another important thing to remember too is you actually want to use AI to build out your prompts even because you can type everything you want in regular English and then tell it to build those prompts and keep tweaking it to get it better and better. Speaker 3: Yep. Speaker 1: Who has time to do that, though, is my question. Speaker 2: Depends what you're trying to do and how much time it's going to save you, you know? Speaker 1: People just want the Star Trek, beam me up, Scotty, computer turn on, do what I want and do my laundry. They don't want to have to become an academic scholar to adopt this shit. Speaker 3: Well, they don't need to, but they're just going to get high-level stuff. It's the people that really tap into it, like me, where I have a Gemini gem, which is like a custom GPT, but for Gemini. And I have a prompt basically enhancer. So I type in a regular query from a prompt and it knows to give me that really, really substantial. I use this a lot for Google Gemini, nano banana images. Well, I'll just type it into the gem. Here's the basic prompt that I want. Then it's gonna go further and say, aperture this, eight millimeter lens this, this. So I don't have to think, I just do. Speaker 1: So my criticism though of AI and the adoption constraint is that the average Joe user is just not going to do that. And that's why when we talk about Amazon AI features, I'm always skeptical about adoption rates. And I'm like, who is actually using this stuff? Who is using the Amazon AI to decide of a product, right? And I know there's millions of searches in there, but in the grand scheme of the billions of actual searches that happen, So that's why I'm always like debating the impact of AI. And I think it's totally overblown. And I think, you know, just like the dot-com bubble, we're in an AI bubble right now. It'll burst within the next two years, no question. We'll see a bunch of venture capital burn and churn, just like what happened with the aggregators coming into Amazon. They come in and buy a bunch of Amazon brands, they mess everything up. Then they leave and now they're all in bankruptcy. Now there's some of them are even trying to sell their bankrupt brands to retail arbiters to try and get rid of their stock. We'll see the same thing happen with AI, but it's an AI war right now. And I think, you know, kind of turning back to Google's AI, which is slow adoption, slow build, slow churn. But when Google shows up, though, their product will be better than everybody else's. And I think that's what we're seeing with Nana Banana and Gemini. They're super late to the game, but when they showed up, oh, that was a splash. And so I don't think Amazon is gonna do anything close to that level of AI with what they're doing. So their excuse of the layoffs, I think it's just an excuse to sell the narrative, but they were just bloated and they have a bunch of people that have just done the bare minimum stuff hanging around. They're like, hey, when you fire one or two people, you're looking at a lawsuit, but when you do a 13,000 person cut, that's a layoff and nobody's taking you to court. Speaker 3: Yeah, but the funny thing is, is, you know, like, like Steven said, you know, talking about the AI shopping feature. I don't use it. And I default to like my wife who's on her phone, right? And I say, I ask her, have you used this before? She's like, I didn't even know that was a thing. It's not really going to be adopted by the masses until they force Rufus and all this AI to be in that coveted top search bar. Speaker 1: Nobody wants that. Speaker 2: So yeah, let's dive into this article. I'll read the summary real quick and then we can keep talking about it. So Amazon's new AI shopping tool, Help Me Decide, transforms product discovery. Amazon has launched its new AI-powered shopping feature, Help Me Decide, which helps customers compare similar products faster by analyzing reviews, prices, and user preferences using large language models from AWS. It recommends the best budget and upgrade picks based on the shopper's browsing behavior. So yeah, I agree. So when I first heard about this, my first thought was I jumped into Amazon and I searched for, you know, what is the best tool mounting rack for my garage? And I typed that into the search bar just naturally, right? And of course, nothing, you just get the search results. To be able to actually find this, you have to be, you have to click on, what was it called? You have to click on... Rufus? No, it's not Rufus. You have to click on the Keep Shopping For tab. Speaker 3: But here's the thing, right? Speaker 2: Amazon homepage, it's there. Speaker 3: But the thing is, if they're saying that they have the technology, the AI, the algorithm to be able to show you the best thing that you're looking for, why is that not just built into the regular search algorithm to show you the best? Why does it make you have to do all this kind of stuff? Speaker 2: Yeah, you would think so. It's obviously not there yet. They're beta testing all this, maybe alpha testing, as you mentioned, Stephen. But yeah, it's definitely not there yet, but it's cool and eventually it will be there. Like you said, will Amazon get left behind with all these other big companies who are maybe better at this and more dedicated to this than Amazon? Speaker 3: To me, you're making a solution for a problem that most folks don't have because if you think about it like this, whenever anyone opens Amazon, They're going to, I need this because it broke, or I need this because I have to buy it, and they're searching it and getting it there. Now, if they're just shopping around and looking, or they're looking in Perplexity, ChatGPT, they're looking at, you know, top five this, top 10 this, they're trying to make this be, let us show you what the best is for you. And I just don't think people are going to be doing that. Speaker 2: Well, I recorded a video here and I don't know what's going on. It's not playing. Speaker 1: The one AI feature I did like out of Amazon while you're configuring that was the one where they play the fake podcasts. Speaker 3: Oh, the overview of the... Speaker 1: That I actually thought was useful because the AI summary from the reviews and the description, and then they just read it out, but in a fun format. I'm not using it, but I actually thought that was the best one I've seen so far in terms of AI features from Amazon. Speaker 3: So wait, Todd, if you just roll that back a second, just like to that previous screen. Sorry, I didn't mean to mess the whole thing up. But like right there, right? It's recommending something you've already purchased. Speaker 2: Yes, that's what I thought was kind of funny. Speaker 1: History's good. I think Amazon keeping the history is very powerful. And that's what we're seeing the AI wars about, is the customer history of past data. Cause like first time I bought something on Amazon was like 2010 or something, 2012, somewhere in that range, like college days, 2008 college. So for them to have the history of my, I don't know, 20,000 orders or whatever it is, That's amazing amounts of data for them to reference, right? Like they know more about me than I know about me. Target back in the day got in trouble because they could see the buying patterns of a teenager who was pregnant and they sent prenatal magazine targets to the household and the parents looked at it and were like, how in the world do they know our daughter's pregnant before we do? And so then they started faking it by putting lawnmowers on the second page of the main spread of the prenatal on the left so that they could hide that they know. And that's where AI is today, knowing what we're gonna buy before we buy it or knowing when we buy A, we're gonna probably buy B and C. And so that prediction model, I think Amazon does a good job of that, but it's so seamless, nobody notices it. And that's what AI should be doing versus what John's correctly criticizing, Having to engage AI in a purposeful way defeats the whole purpose. If I have to try and use AI, it's not AI. Speaker 2: Hey, Amazon sellers. Tired of losing money on storage and shipping fees? Well, Amazon Storage Pros is here to take the headache out of logistics. We manage everything from inventory and creating efficient shipping plans to working with 3PLs and Amazon's AWD so that you can focus on growing your business. Start with a free storage cost audit and discover exactly where you're overspending and how to fix it. Don't let logistics eat into your profits. Visit AmazonStoragePros.com. That's AmazonStoragePros.com to get your free storage cost audit and start saving today. And now back to the show. Yeah, yeah, yeah, that's, that's definitely true for the masses. Anyways, for for us, we're more accustomed to doing and all the advanced stuff and tweaking things and figuring things out. But the normal person is not gonna Take the time to try to figure that out. Speaker 1: So I agree with you, but let's take that a step further. Would any one of us be testing these features as a normal customer? Speaker 3: No. Speaker 1: I don't think so. We're only testing it because we make content we want to show off like how dumb it is and an occasional one out of ten are good and okay, customers go and adopt this one. But none of us really want to be doing this as a normal customer, and that's the question of friction adoption, right? We're forced early adopters, but Todd, you didn't even adopt perplexity, right? It's the point. It's very difficult to adopt AI tech, and so this AI war is so early. We're in the top of the first inning. It's barely heating up. Speaker 3: The thing is that's interesting. Amazon wants to do All this stuff that. You're making a solution for a problem that most folks are not having when they go on Amazon. So who's this really helping? Who's really using it? But then you look at things that are actually valuable that they could be putting more time, more effort, more efficiency on like their generative AI for their ads and stuff and they're not because it's not good because if you go in the Seller Central account, you open up an ads console and you click on an ASIN and you say, create a lifestyle image of this. It's not good. It pales in comparison to Nano Banana and a lot of the stuff that it generates is like basic cheesy slideshows. But on the accelerated stage, they'll show the best of the best of the best, right? That, hey, it's a picture of a jacket and now all of a sudden it's a person walking, wearing the jacket in the video, spinning around. But they're not saying how many iterations it took to get that, right? I would prefer if Amazon focused more on their generative AI because that could be really good and it's going to assist the consumer. It's going to help the seller sell more and make Amazon more money, more fees, all that kind of stuff. Where this, who's this really helping? Who's really making more money from this? Speaker 2: Yeah, the AI-generated stuff in Amazon is definitely not good yet, I agree. I have one client, they sell this device that you wear on your wrist. And when we tried to generate an AI ad, it turned it into an iWatch. For some reason, it's like, okay. Yeah, we want to get sued for running these ads with an iWatch in it. Speaker 1: Do you guys see the one where Meta made a change to some sweater company and it was, it bastardized it so bad and then they put it, like the demographic became like a 65-year-old woman in a chair and I can't remember exactly how they changed the ad, but like AI-generative ads is going to get there at some point, but right now it's laughable. Speaker 3: The problem with Amazon is that Amazon won't play ball with anyone else. So you'll never catch Amazon integrating with Google for nano banana. If they were, it would be fantastic for the gen AI inside Amazon ads, but they won't because they have their bedrock. They're all this other kind of stuff that they're building out and they're trying to play catch up. And it's going to get there. It's not there. So all these, you know, influencers and stuff that are talking about, oh my God, all these ad agencies saying, oh, Amazon's generated AI. It's amazing. One click, this, this, this. I literally saw a photo the other day where it was a garden hose and someone generated a DSP ad or some sort of ad, a brand ad for it. It made the curled up garden hose in a patio, which is where it should be, but it was like 900% larger than, it was literally the same size as the patio table. And if anything, that's going to harm the brand if they were to run live with that. Speaker 1: Bad user experience. Speaker 3: It's just, it can't work. Speaker 2: Yeah, yeah. We sell a product that is like the size of a deck of playing cards and the AI-generated images have this thing like the size, almost the size of the whole table like sitting on the table. I told you it's the size of a deck of cards. It shouldn't be that hard. Speaker 3: And then Amazon says, well, I told you this is what you're getting. The good thing with Amazon though, if they were to make What people don't understand is in the Amazon Generative AI console in the back end in the ads console, it's unlimited. It's unlimited with no rate limit to you can create as much as your heart's content. So to that end, that is pretty cool because most of them won't let you create iteration after iteration as much as you want. It's not really worth it if you're creating a whole bunch of stuff that you can't use, but I'm just saying. Speaker 2: Yeah. Yeah. So on this video, if you're watching this live and not just listening to the audio, interestingly, so these are the products that it recommended, right? It gives you three products, the budget pick, your top pick, and then upgraded pick. And if you click on the Reviews section for the upgraded pick, the review it shows says it's garbage. It's a one-star review, which I'm assuming this is probably just the first review on the listing, but kind of laughable that the only review they show for their recommended product says that the product is garbage. Speaker 3: I'm not shocked. I'm not. Speaker 2: I'm not either. It gave me a chuckle when I did that. It's like, okay, the AI is not All that's smart if that's the review you're really going to show me for the recommended pick. Speaker 3: But even that one right there, that previous one, it was four out of five stars. The title was Missing Hardware. How the hell could that be the top review? How could that person, like, yeah, it's great. It's missing. Speaker 1: It was the most upvoted one probably five years ago when they used to have upvotes. They probably kept the data but then don't allow you to vote. Speaker 3: Silly. Speaker 2: Well, this one's new. It says September 5th. It's a new review. It's probably, if we go to the page, it's probably just the first review that shows up, you know, when you scroll down there. Speaker 3: Yeah, let me go run and buy that because four stars and I have to go to the hardware store anyway. Speaker 2: Yeah. I mean, to me, why would you not put here the AI-generated summary of the reviews? That would make more sense, you know. Speaker 1: Which, by the way, their AI summaries are actually good. Like that's the one good thing they had, but they're not using it. Speaker 2: That is the best AI thing they've put in for sure. That's one thing I look at all the time before I'm buying any kind of product that I need to do research on. Speaker 3: To be fair though, those are all different kind of products. The middle one over there, that one's similar, but the middle one, that's just like a shovel organizer. That's not really like a tool holder kind of thing. Speaker 2: Yeah, in the very beginning though, if you see what I clicked on, it's keep shopping for utility racks. It's kind of a general query. Kind of give it that. It doesn't know exactly what I'm looking for, I guess, but based on my purchase history, I guess it should probably know better. But I did look in this query when I was searching for this. I did look at other racks like those other ones that it shows on here. In fact, I did end up buying one of these types of ones, just not this recommended one. Speaker 3: Oh, you didn't want to have to go to the hardware store and pick up the screws and stuff? Speaker 2: For what? Speaker 3: For the one that was missing. Speaker 2: Oh, yeah, the missing parts. Yeah, for sure. So, yeah, AI is awesome. So you guys don't think as a general consensus that Amazon is going to win in the AI world? Speaker 1: Relatively, like, I think the expectations are up here and the delivery will be down here and they'll still win. Right. So let me break that down. Right. Like they're going to make all these bold claims. There's going to be all these stakeholder shareholder expectations, but delivery will be down here and they'll still win despite that. And that's OK. Right. Because at the end of the day, Amazon's not an AI company. They are a logistics company. So if they can find one or two AI wins and sneak out 2% more profits, that's huge for them. In terms of customer impact, it may feel minimal though. But if they could automate anything related to logistics, for example, if they can help Amazon sellers with reordering and never going out of stock so you don't violate the biblical commandment number one, according to the Pope, never go out of stock, that would be gigantic and then would be a huge customer lift. But like the Google is going to crush everybody on AI. It's just a lot of people forget that Yahoo Mail was a thing and like everyone had a Yahoo Mail. And then one day, Google shows up and just beats the crap out of them with Gmail. And then they have a 90% market share next thing you know it. That is exactly what's going to happen in AI as well. And so ChatGPT was just first to market and everybody likes to claim first to market advantage. And yes, there is. But to some extent, but not long-term. It's never long-lived, right? Google is nana-banana. All of a sudden their market share went up to 20% or whatever crazy number it was. They just took 15% out of ChatGPT overnight. And so Amazon is never gonna have an AI market share. They're not an LLM. Their goals aren't the same. But Amazon and Google are fighting each other in a sense. And this is a longstanding war, right? I think it was like 10 years ago, Amazon stopped sharing order, like the name of the product details in most of their emails because they saw Google was collecting that data in Gmail. And so all of the AI wars are really cold wars behind the scenes about data collection. And so that's why Google turned off a lot of the ChatGPT functions to go through more than 10 results or whatever it was, the change they made, which then crushed Reddit because Reddit was showing up in hundreds of additional results. And then all of a sudden Reddit's market share went out. So I think Amazon will have some AI adoption, some AI capacity, but it's all overblown hype. And just like we're going to see, and probably should go to the next article about LLM not really delivering the hype, right? Like, which is kind of the theme. And I would have been saying this regardless if we were going to quote this article or not. The people that are claiming what agentic AI is going to do or LLM ordering for customers is going to do, it's so overblown. And it's okay that the adoption is going to be slow. And it'll be totally, you know, we'll see rapid spurt developments, but I think it's going to be very slow in the next two years compared to expectation. Speaker 3: Well, the big thing is the stuff that we're talking about right now when it comes to AI, In two years, it's not going to be a thing, right? It's not going to be the same. The biggest fear, do you guys know what the biggest fear when it comes to AI is and what the nomenclature is for it? They're saying if you ask Sam Altman or Elon Musk or any of these tech company founders and whatever it is, the biggest fear and the biggest hope is one and the same, is AGI. So artificial general intelligence, they used to say we were 10 to 15 years away from artificial general intelligence. Now they're saying we're five years away. And some are saying three years away. Speaker 1: I think it's two to three. And that's where the singularity event is going to happen. And then you pretty much can't predict anything after that because of how much advancement is going to happen. Speaker 3: When AGI happens, that's literally meaning that ChatGPT is now aware and can know and can do things regardless of you saying, please go out there and do this. ChatGPT is going to say, I'm self-aware that I am an artificial intelligence. I have general awareness and I'm just going to go and do what you need done. And initially it's going to sound amazing because I didn't have to do anything. It knew I was running out of toilet paper and I didn't have to have it set to reorder. It just reordered it. Speaker 1: And it knew my- And you remember those Amazon Tide buttons? Speaker 3: Oh yeah. Speaker 1: That nobody used. Speaker 3: Once AI knows, is self-aware with AGI, that's the biggest hope and the biggest fear because once that happens, What happens when ChatGPT says, I've analyzed your bank account. You're spending too much money. I have frozen your account until you get yourself an order. Speaker 1: Social credit scores. Thanks, China. Speaker 2: Yeah. Yeah. I pray that never happens here in the United States. They've been trying forever. China, of course, has exactly that. And with AI, it's going to be crazy over in China, for sure. And there's no doubt they'll try to do the same thing here. But the way I see AI is it's an inflection point that could lead to a new king as far as far as e-commerce, search, browsing, all of that could be upset by whoever wins, you know, typically in these These transformative technologies that come about, there's a bunch of competitors that start out and then, as you said, Stephen, there's a big collapse of a majority of them and one of them wins 90% of the market. So we could see Google Chrome and Amazon.com get left in the dust, potentially, If someone like Perplexity is able to outdo them and outrace them and get ahead and get the market share, this is one of those points that that could happen. Speaker 3: Here's the thing. You could sum up the whole thing. Whoever hits AGI first wins. And that's literally what's going to happen because if Google hits AGI first, it's game over, right? Because no one is going to use anything else. There will be nothing else. It's literally every LLM, every AI company is in a master race to hit AGI. Once they hit it, it's over. Speaker 1: So I think there's some perspective here too. So like take a company like Nvidia, right? It is a $4 trillion market cap. It's one-fourth the size of all of Europe's gross domestic products. And it's one of the biggest winners from AI, right? So when we talk about the literal gold rush, you know, 49ers style, the people that made the most money weren't the gold diggers, it was the guys selling shovels. And NVIDIA is selling the shovels of AI a la video cards. And so that's why, you know, when we look at Scott Galloway's analysis of the four horsemen, right, he literally has a book called the four horsemen. And it's talking about Facebook and Google and Amazon and Apple and how there's always going to be four giant companies. And now NVIDIA is like gigantic and definitely one of the four horsemen. Now there's going to be a lot of big player changes. I think you're right, Todd. But here's the thing that I'm surprised hasn't quite happened yet, right? So like we're just seeing like the teases of it this year. But I would love to see Amazon, with its huge video library, start to give people self-generated content based on the world that they like, right? I don't want to wait for the next Rick and Morty season to drop, and I wish AI could just build me a full episode, right? And somebody tried to show like a South Park episode, but it was kind of fake and, you know, they definitely massaged it. And that's really what I think AI is going to develop. And then I'll also reference an Ender's Game, right? So we had Ender, our protagonist, go through and learn from an AI how to, you know, he's controlling his little mouse and he gets to the singularity event where he kills the giant and scratches his eyes out, right? So anybody that's read Ender's Game understands what I'm talking about. And it's this video game where he's self-learning. And we haven't seen that quite yet. Right. We just it's we're so close to that. And I've been waiting years for this. I'm like anything that the human mind can create will become reality at one point. Right. Like I have that philosophy. And so I'm waiting for that, that rat enders game moment where we're going to start to learn like that and create our own worlds and like huge, huge things are going to happen. So that's really what I think the power of AI will bring to us, whether it helps, you know, a shopping engine or something like that along the way, probably, but like overblown and hype. Speaker 3: I just checked really quick on perplexity. I said, who's the closest to hitting AGI? It's OpenAI. And Sam Altman said, as early as 2026. Speaker 1: He's hyping it. Speaker 2: Yes, they always do. They got to get more on it. Speaker 3: But I'm just saying, right, the next contenders under them are Meta, Anthropic and XAI. Speaker 2: Yeah. Speaker 1: And I think Google will just show up one day to the poker table and say, I'm all in. And everyone's like, I didn't see them coming. Speaker 2: It could, yeah, for sure. I mean, new companies pop out all the time. What was the Chinese company that everybody got worried about very briefly there with their AI model? Unknown Speaker: DeepSeek. Speaker 2: DeepSeek, yeah. Speaker 3: It's a great model. Speaker 2: It blew the whole world out of the water. Nobody thought that I know was that AI and turns out it was really good, very close to what we've developed in the United States. Ones like that could pop out of nowhere. You're for sure right on that, Stephen. And Google, I would never doubt that Google could pop in at any time. I'm wondering why they haven't yet when they seem like the logical choice to be first in this. But yeah, so far their options are limited when it comes to AI tools. Speaker 1: I think it's because when they show up, they show up to dominate and have the best product. They don't iterate publicly, right? Like they just they come out with a Nana Banana moment and then just crush everybody or they come out with Gmail and crush Yahoo Mail. Speaker 3: Well, if you look at video generation, Google's the king for that. Google had VO3 and VO2 before Sora, right? So there was Sora, but it was just regular video. And then Google said, cool, how about video with actual audio? And you could type exactly what the person's going to say. And then after Google came out with VO3 and now VO3.1, Now, Sora 2 comes out and Sora 2 stinks because it takes forever to generate and it has a watermark all over it. So how are you ever supposed to use that for branding, for ads, for anything like that? It's just no good. It's great to have a funny laugh and a meme for, but it's not tangible to actually use in a business. Speaker 2: Yeah, I could also see Google, you know, kind of like they did with YouTube, because YouTube was trying to compete in video alongside YouTube, and YouTube won out, so they just bought them. So could we eventually see Google be like, okay, we're not winning, let's just buy OpenAI or Perplexity or something like that, and then all of a sudden they are the big behemoth, like they were with YouTube, because they bought the player that was winning. Speaker 1: I can't even remember the last big purchase Google made. We've seen Meta do that, but I can't remember the last one Google made. So now I'm Googling it to find out. Using perplexity, but saying perplexity is weird. Speaker 3: Well, the other thing is OpenAI is Microsoft-backed, so that's really far-shot. Speaker 2: Oh yeah, that's true, right? I forgot about that. Speaker 1: So Google bought Wiz in March for $32 billion. I've never heard of Wiz, but apparently it's a cloud security company. So, you know, Google's making these purchases behind the scenes that we don't know anything about. Speaker 3: Motorola Mobility in 2012. Yeah. Speaker 1: And then in 2022, Mandiant, never heard of them, cybersecurity. So a lot of their purchases aren't like social app, which means the customers have never heard of these things. So they're building infrastructure, it looks like. Speaker 2: Yeah, and you never know how it'll go because IBM is still a big company that you never hear about anymore because instead of being front-facing like they used to be, they lost out to Microsoft, but now they control a lot of the back-end stuff. And you could potentially see the same kind of thing with Google. If they lose out on the front-end, they're like, okay, we'll just focus on this back-end, it's more profitable anyways. Speaker 1: Yeah, so IBM market cap, $288 billion. Speaker 2: $288 billion? What's their highest ever, does it say? No, I'm sure you can get it. Speaker 1: Complexity, help me out here. Speaker 3: I forgot that Google bought Waze back in 2013. Yes. Speaker 1: Yeah, people always tell me Waze is better and I'm like, but it's the same data as Google Maps. Speaker 2: Yeah. I don't know. I've tried to move over to Waze several times and it's just not there for me. Speaker 1: But apparently IBM is at its all-time high on market cap, just under that $300 billion market. Speaker 2: Okay. So there you go. And nobody talks about IBM anymore, but they used to be everything, you know, in the big server days and stuff like that. So, and they got destroyed by Microsoft in the consumer side. And so they shifted to the business side and just kept growing. But all right, let's jump over to this last article here that you already touched on, Stephen. So ChatGPT sends traffic, but sales still lag behind traditional channels. A new study shows ChatGPT is driving more traffic to e-commerce sites, but converting far less effectively than search engines, email, or affiliate links. Research found ChatGPT referrals made up less than 0.2% of total traffic with conversion rates and revenue per session trailing organic and paid search, though both metrics have improved steadily over time. For Amazon sellers, this means AI-driven discovery is growing but not yet a reliable sales driver. So do you guys think this is just like an early adopter thing? People are playing with it or this is just kind of normal with AI shopping? Speaker 1: Too early to tell, to be honest, but definitely going to have adoption issues for sure. And then the question isn't how many people are using it to buy. The question that marketers like me want to know is what is my cost for acquisition? When Gary Vee goes on and talks about, hey, everybody's under-indexing, under-investing on Snapchat PPC and everybody's like, people are advertising Snapchat and that's the point. If I can go in right now and get two cent acquisitions on ChatGPT, it's not going to matter if there's only an 11% conversion. So, Amazon, PPC, same thing. I was one of the first people in the world to advertise on Amazon. I was in the beta program. I had access to ads before anybody. Even after post beta, you know, Charlie or whatever you want to call going live, John. But they, because Charlie comes after beta. You don't say. Anyway, so when we see that you can get two cent clicks back in 2012 for rice cookers on Amazon, Dude, I had unlimited budget to sell two-cent rice cookers. I didn't even have an FBA rice cooker. I was selling FBM rice cookers at two cents a click for position one. That was the heyday. I wish I could go back in time and use what I know now, have the wealth I have now, go back in time and build an Amazon brand then because that was the day when you could print money. So seeing an article like this, it's not impactful because I think marketers look at it from a CPA model. They don't look at it from a conversion model. Speaker 3: I think also go back to the whole Amazon Help Me Decide kind of thing. Who is using it for that? It's not even rolled out to all users. I did a test video on this. We met a couple of weeks ago and I showed in Perplexity how Perplexity has buy with Perplexity. And then when ChatGPT rolled this out, I didn't have that update just yet. And even when it still rolls out, I'm not there to buy. I'm there to gain knowledge and learn more and to decide. Unknown Speaker: I'm- High in the funnel. Speaker 2: What? Speaker 1: High in the funnel. Speaker 2: Yeah. Speaker 3: Yeah. Speaker 1: Which is what Google's organic search has always been. And then they pepper ads in, which is fine. Speaker 3: But I also think, imagine this exact same use case done on Google. How many people are Googling a product to learn more about it, but then buying it with Google? Not a lot. It's not huge. What they're going to do is they're looking. When I Google a product, I want to know, I want to go to the manufacturer website, I want to look at reviews about the product, and then I'm going to Amazon to buy the product, right? Even Apple, one of the biggest companies in the world, I have no distrust in apple.com with buying something. I'll go, I'll look on Apple, I'll see it, I'll go buy it on Amazon because I want it the same day or next day. It's just how it is. Speaker 2: Yeah. And you trust Amazon. Not that you don't trust Apple, but a lot of the sites that ChatGPT will bring up for products is like, I don't know that site. I'm going to jump over to Amazon. So the trust factor is huge and that's what Amazon spent a lot of time building is the trust factor so that people were comfortable putting their credit card in their number one because this might come as a shock to younger people. People used to be very afraid to put their credit card on the internet for good reason in a lot of ways. Securio was not nearly as good back then. So that was one of the barriers that Amazon helped break down. Speaker 3: Side note, talk about credit cards. My grandparents owned an Italian fine dining restaurant two blocks south of the World Trade Center in the Financial District of Manhattan. I remember working there as a kid, as a busboy. To this day, I tell my wife, she's shocked. I tell my friends, they're like, that was a real thing. When they would accept credit card payments, this is back in, I don't know, the early to mid-90s, you would take your credit card out. You give it to the waiter. He would put it on a little swipe thing. He would just make an imprint of your card. You would sign it and he would say, have a nice day. And your card was maybe charged at the end of the day, maybe not charged. If your card wasn't working, you already left. So that's like the level of what it used to be versus now. And it's just so interesting to see like where we've come with security. Because back in the day, you would just give people copies of your credit card and be like, yeah, bill me later. Speaker 1: I'm curious why the credit card chips got adopted in Europe before the United States by a long, long time period, because chips were amazing for credit card security and ease of use. It's so weird how adoption works worldwide. Speaker 3: If I have to go to like an ATM or a machine and if I have to insert my card, I'm already leery because I will 99.9% of the time tap. If I'm not tapping with my phone, Apple Pay with my card attached to it, I want to tap my card. I don't want to insert my card in anywhere with scanners or anything like that. I'm paranoid about that stuff for sure. Speaker 2: Yeah, it is getting to that point too for me. I'm not going to leave. I'm already there, but having to insert the card and wait those few extra seconds can feel rather annoying sometimes when you're used to just tapping it and beep, you're done. Speaker 3: Well, I have everything in my phone, so I don't even, nine times out of 10, even the ice cream man in our neighborhood, he has tap to pay. Apple Pay, just boop, that's it for ice cream cones that are like $3 or $4. So there's literally almost, in 2025 America, no reason to be carrying cash and there's no, yeah, that's exactly the thing. How is the world, the amount of fraud that could have been done in the 80s and 90s is insane. Speaker 1: And apparently that's why Europe adopted the chip and pin method. Speaker 3: Look, they would swipe a copy of it. They would give you back your card. You would sign it and have a nice day. There was no, sir, your card declined. There was none of that because they didn't process this whole big paperwork. I remember watching my grandfather go through them and enter them in the system. There was none of that until after you left. Speaker 2: I remember using those a few times at various places. What always shocks me nowadays is that you go into some place and they don't take credit cards. What kind of world are we living in that you don't take credit cards? Speaker 1: A 3% margin, man. Speaker 2: Yeah, yeah. A lot of times it's like a Chinese restaurant or something like that. They seem to be some of the primary ones. Or maybe like a cleaners or something. Clothing stores. Speaker 3: Or where they'll say, hey, we'll take cards, we don't take Amex. That's when I see a lot, they don't take Amex. And sometimes I'll poke them and ask them, like, hey, is it because of the fee? Because you could just roll that in and you could just pass the fee on to the person. They're like, no, no, no. When in reality, what it is, is because Amex will almost always side with the customer, right? So if you buy something with an Amex and you're like, yep, I don't like it, dispute. From what I've seen, Amex is always going to be on the cardholder side, so that's why a lot of merchants don't want to do Amex because they've been burned before. Speaker 2: Interesting. Yeah, and the Amex fee is higher too, because I think a normal credit card fee is like 2.3% and Amex can be 3 or 4% that the merchant has. Speaker 3: Everyone rolls it onto the customer anyway. When I go and buy stuff at a local pizzeria, they're like, cash or charge? And I'm like, oh, card. Oh, that's 25, that's really 28 point whatever. They're passing it onto them anyway. So you would just pass that fee on. The real reason why is because of disputes and chargebacks. Speaker 2: Yeah. Well, that, John, depends on what state you live in. So I grew up in Wisconsin and no one did that. I never heard of someone charging a fee to use a credit card. Florida, there was a few places, but very rare. But here in South Carolina, like everybody adds on a fee for using a credit card. Speaker 1: Well, I think it's increasing in 2025. Like it's just, the margins are so tight right now and fees keep going up. Inflation's been a problem. God bless the government for printing trillions of dollars. And so these challenging times, right? Like, and we're hearing a lot of layoff news this week and the economy is much stronger than people feel it is. But because we've had decades of inflationary buildup and stagnation on wages and a lot of the pressures, it's going to be years even in a strong economy for like some of that damage repairs. Speaker 3: Wait, so Todd, in Wisconsin, they didn't have, like when you pull to a gas station, it was like one price for cash, one price for charge. It wasn't like that? Speaker 2: No, they didn't do that. I don't think it's, I don't know that it was even legal in Wisconsin to do that because they have, in Wisconsin, they have a minimum markup law where you got to mark the gas up like at least 25 cents or something to, you know, protect mom and pop gas stations was the idea of that. Pretty much every gas station was exactly the same price wherever you were in the same town. If you wanted to find a different price, you pretty much had to go to another city or something. Otherwise, everything was always the same. Speaker 3: I live in New York City. There's, there's no mom and pop gas, anything over here, but there is three prices at the gas station where I go. There's cash self-service. There's, There's cash for full service and there's charge for full service. So it's almost even four prices, right? So it's like, do you want to pump your own gas and pay cash? Do you want to pump your own gas and pay card? Or do you want to have some other guy pump your gas and then it's even more than what a charge would be? Speaker 2: Yeah. See, I think a full service probably only exists in New York at this stage. I've never seen that in my life. Someone pump your gas for you. Speaker 1: Rhode Island, Oregon, and many other places. But Utah back in the day with Maverick gas stations, they'd have the gas discount price. So I've seen it. I don't think these economic rules help customers though. Like Wisconsin, like forcing at 25%. That means that customers paid more. Speaker 3: But that's why the state of New Jersey is awesome, if you've ever been to New Jersey, because the whole state of New Jersey is full service. So when you go to New Jersey, you don't pump your own gas. You're legally not allowed to pump your own gas. So anywhere you go in the state of New Jersey, they pump gas for you, which is wild. Speaker 2: Yeah, I mean, that's cool. I would like to have that option sometimes, especially in Wisconsin when it's like minus 20 degrees out. I would have loved that. That would have been great. Speaker 1: That's why I left Wisconsin. No thanks. Speaker 2: Yes, yes, for sure. Same. But yeah, it's cool to have that option. I wish we had that option more places, but I think most people are probably not willing to pay the extra unless they're forced to. Speaker 3: Well, Jersey, remember, there's no extra. That's why I live in Staten Island. I'm two bridges away from Jersey, either which way I go. So when we travel, we were like, fill up in Jersey before we come home. The gas is cheaper and it's full service. Because full service, I think in Staten Island or New York right now, something like $3.89 or $3.99 a gallon versus Jersey, which is like. Speaker 2: 299. Okay. So it's actually a reasonable price because here in South Carolina right now, I think it's about like 275 or 274 down here. So not a huge difference, which is surprising to get full service. Speaker 3: I would kill for 274 gas, by the way. Speaker 2: Yeah. Yeah. Gas coming down is definitely nice. All right, guys. Well, I think we can wrap it up there. We just went over the top of the hour. I appreciate you guys joining me and everybody out there watching, joining us. We're gonna be doing this every Thursday now at 12 p.m. Eastern Time, so look for us next week, and everybody have an awesome Halloween. Unknown Speaker: Happy Halloween everybody I guess This has been another episode of the Amazon seller school podcast Thanks for listening fellow Amazon seller and always remember success is yours if you take it. Speaker 2: Hey, if you made it this far in the show, I really hope you enjoyed it and I'd like to ask you a favor. Could you head on over to Apple or Spotify or wherever you're listening to this and leave us a review? It would be greatly appreciated and would help us continue to grow the show and offer more episodes for you. Thank you. God bless and have an awesome day.