Amazon News: ChatGPT Privacy Leak, India Tariffs & AI Team Shakeup

Amazon News: ChatGPT Privacy Leak, India Tariffs & AI Team Shakeup 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: Hello, hello, everybody. Another week, another Friday already. The weeks go faster and faster, it seems like to me, but especially during the summer. I've been super busy as Noah and I were talking about running all over the country for trainings and stuff like that. And I'm actually out building a fence today, this morning. So trying to get our fence around our house done and working with the neighbor to build theirs as well. So having fun. How's your summer going so far, Noah? Speaker 1: I would say it's great that you are very correct in the whole, it's flying by very, very quickly, but that's okay. I'm a big fall guy myself, so I'm gearing up already for fall. I got my Halloween decorations over here that I'm going to be throwing up here in next month or so. Either way, great summer though. Speaker 2: Yeah, yeah, already back to school. So anybody in the back the school space, I'm sure are booming already last month, they were probably booming. And now it's really going crazy for all the last minute shoppers. I know some of my relatives here in South Carolina, they go back to school next week. So I know every place is different, but they're back to school on Monday. And that is booming right now for sure. So this is your Prime Day and Black Friday all rolled into one if you're in that space. But yeah, we've got lots of different spaces on Amazon and today we're diving into a lot of AI related stuff. So if you're interested in AI related stuff, definitely stay tuned because we've got a lot of really cool stories. And we're gonna be diving into some prompts and stuff as well that I found and go into them. So if you're out there watching, if you have any questions or comments about anything related to the stories or just anything in general, throw them in the comments and we will try our best to bring them into the show and dive in. We always love answering the questions for everybody out there. So please do that. Do not be shy. Throw your comments in there and we're happy to answer them. But with that, let's go ahead and jump onto our first story here. Your public ChatGPT queries are getting indexed by Google and other search engines. An OpenAI experiment temporarily allowed shared ChatGPT conversations to be indexed by search engines like Google and Bing, exposing some surprisingly personal content. Although users had to manually opt in to share a chat, many were unaware that these links could show up in public search results. This raises privacy concerns for those who use identifiable details, including job applications and resumes, without expecting broader visibility. OpenAI has since removed the feature, citing accidental oversharing risks, but the incident highlights the importance of caution when sharing AI conversations, even for Amazon sellers using ChatGPT for listing creation or business help. So yeah, definitely a cautionary tale. They already fixed this, thankfully. But it's a an important lesson that when you're putting stuff into chat GPT into grok into what's Google's called I forget. Speaker 1: Gemini, Gemini. Speaker 2: It's going into somebody else's servers, somebody else's computers. And so there could be leaks, that stuff could get exposed. So be careful what you're putting out there and make sure whatever you're putting out there, you would be okay if it accidentally got exposed to the public. Speaker 1: Yeah, I think this is one of those things that as AI is progressing and people kind of forget the aspect that, yeah, it's been touted as like a personal assistant, right? And so people get very personal. I mean, I've even seen people who are using like ChatGPT and like Claude as like personal therapists, which I'm like, you should not be doing that. It's very Black Mirror-esque. But I do think this This is something that we just we don't know this stuff is going on until it's going on. Right. And, you know, we have to remember that realistically, all of these AI companies are I mean, they're just data centers. Right. Same thing with Google. Google is able to do everything it does because it is just a data mining warehouse. Right. And you're seeing, I think, sites start to get I think it's a lot wiser to a lot of these AI. So like, even with ChatGPT's new agent model, they've already, like Amazon, for instance, recently made it so that you can't use things like ChatGPT's agent to even actually access Amazon listings directly. And so it's one of those things where, yeah, it seems really cool in the moment, but The companies out there, Amazon, Google, all of them, wouldn't be making these restrictions if it wasn't for the fact that they're like, yeah, this is a problem for us. So if it's a problem for them, it's definitely a problem for us. Speaker 2: Yeah. Yeah. Well, with that, and we talked about this last week a little bit, which of course you weren't on the show last week, but We talked about how it's kind of reminiscent of the days, the battle between the search engines and the browsers and the operating systems, right? Everybody wants to have their walled garden and they want you to play inside of it. And so as long as possible, they're going to battle each other and block each other and try to stop each other. From getting access to each other's data, until some point you have, you know, Google Chrome and Google search win. And then everybody's like, just take all of our data, we want to be in your search engine, right? So you're going to have the same kind of thing, I think, with with AI. And it's going to be a bloody battle, I think, because this is the future of everything. Online and offline, potentially, with robots and everything else. So whatever company, one or two companies ends up at the top, stands to gain a lot of money and a lot of market. Speaker 1: Yeah, well, and I think as well with that, you know, you have the aspect of like ChatGPT where People are using it basically as just like a newer gen search engine too, right? Like that's a big thing where it's like, you know, you have someone like Google who is a search engine. Google's getting into the AI game with Gemini and whatnot. But I think part of the reason they're doing that is because they see that ChatGPT isn't just, you know, this language model. It also, and all these other ones, all these other AI, it's also something that people are essentially coming into that search engine realm. We are gonna see that I think over the next couple of years. I don't think it's been, you know, I know ChatGPT basically has been like the, I think, most common one that people have utilized and it has the most use out of like all of the different ones. But I think we will see as new things start coming out, more and more people will switch to different platforms and, you know, you'll see Anthropic and all these other companies just grow a little bit more and it will kind of diversify as time comes. Speaker 2: Yeah, definitely. AI is eventually going to eat all of Google Search's lunch, a large majority of it. And a perfect example, as I said, we're installing a wooden fence out here, right? So we're digging the holes, putting the posts in. And I jumped on Grok and I said, hey, we're we're putting these wooden posts in for a six foot fence. How deep do we need to dig them? And I said, dig them two feet, you know, calculated all of it. And then I said, OK. For that, we're putting four by fours inside these two foot deep holes. How much cement, how many bags of cement do we need to fill up the holes? And it calculated everything and it said, you know, two 80 pound bags, which it ended up being an 80 pound and a 60 pound is ideal. So it was a little bit over. But those are all questions that I normally probably would have jumped on Google, watched a YouTube video, wasted 20, 30 minutes trying to figure out the actual answer. And instead, I jumped over to Grok and had the answer in a couple minutes. Speaker 1: That's one of those things that kind of scares me about AI, because it makes me feel like people aren't actually going to be learning as much in the future, right? Because people are just going to rely on this platform to give them like an instant answer. And they're just going to take that answer and run with it, right? Whereas like, there is merit still, I think, in like learning certain things. You know, even if you can get The instant gratification of an answer by typing into Google and getting Gemini to give it to you or ChatGPT or Grok, right? I think there is still merit in the aspect of like, yeah, learning how to do something and like, and having that knowledge available, I think is worth something still comparatively to us getting like that instant gratification answer. Speaker 2: Yeah, yeah, 100%. It's, it's super important to remember that AI can definitely get it wrong. There was recently a study And it'd be cool if I had that pulled up, but I don't. But I heard about it that they did a study where essentially they were trying to help AI get smarter. And they had kids, like young kids, like six, seven, eight year olds. And they would put a prompt into ChatGPT and then look at the answer and determine if it was right or wrong. And the 5, 6, 7 year old kids, 8 year old kids were able to point out mistakes. That ChatGPT and the other AIs were making because, you know, AI, if it doesn't know the answer, it doesn't necessarily want to tell you that. So just very convincing. Speaker 1: It's something that people I think forget is that majority of these AI are language models still. So they are essentially doing nothing more than like researching and getting that information and then parsing that information into a readable format. Like that's all A language model is at the end of the day. Now, there are ones like Claude, like Anthropic, right, where they are, you know, it is actually like an intelligent model where it's learning things. And GBT does have some of the other ones that are more intelligent, but like the average just search functions are, yeah, it's just a language model. And it's one of those that I am a little, I am a little worried about. I mean, I even see it nowadays in the Amazon space with Recently, I had someone book me for a coaching call, and they were somebody who started their own agency and their own brand and whatnot, and they were working on a couple of different brands. I was just asking some questions, trying to understand what their history is with Amazon, because they were asking some pretty basic stuff for someone who's managing multiple brands. It turns out they have no Amazon background. Basically, everything that they have been learning comes from GBT and they had mentioned that in the call where they're like, well, GBT said this. I'm like, I don't care. It's wrong. I have a post I put up on LinkedIn a couple of months ago where It was exactly that where I asked ChatGPT a question and it told me, no, that isn't a thing anymore. And then I linked it to the actual Amazon article that says, yes, it is still active. And ChatGPT goes, sorry, you're right, I lied. And I was like, what do you mean? So it will lie to you very frequently. Speaker 2: Yeah, it's an important thing to remember. And I'm getting I'm 43 now. So definitely getting up there a little bit. And I got my first computer when I was 14 years old. And I feel like there's a major benefit to knowing life before the internet and after the internet, before cell phones and after cell phones, before AI and after because it really helps you understand You know that they that they do get things wrong and know how to do things without the use of those tools. And that's going to be a struggle for younger people. Anybody in their 20s doesn't know life before the Internet, for sure. Maybe not life before cell phones. And if you're younger than 20, you definitely don't know. And the kids are just coming up now are not going to know life before AI. It'll be really important to teach your employees, your kids, anyone you're working with, that you have to not only rely on these tools. They're a great addition to whatever you can do, but don't fully rely on them like the guy you were talking about. Don't just think you can be an Amazon account manager by typing everything into ChatGPT. Speaker 1: Well, and I think it's so interesting that In my mind, there's like a 20 year area, I would say of like what I would say probably like 25 to 45, right? And that like 20 year area is basically people who Came up in a time in which you basically had to learn and like troubleshoot computer issues, internet issues, all these different things. And you had to learn a lot of that. And like as technology was developing, right, you basically had to teach yourself how to solve these problems with technology. Right. And you would think as we've moved more into like a technology forward area, everything that the younger generation would like be super tech savvy and like technological. And it's like, No, they're not. I look at my nephews and nieces and it's like, they've had an iPad in their hand or a computer or a cell phone in their hand since they're three, four years old. And it's like, you'd think that would make them super tech literate. I think it's done the opposite. I think that all these companies, they've made technology or optimized it so much to where it's so user-friendly that you don't have to have that base level knowledge. And I think AI is going to exacerbate that even more. Speaker 2: Yeah, and they've never had to learn how to use an iPad or a computer because they got it young enough that it just became automatic, right? Speaker 1: And it's so intuitive. Speaker 2: You don't have to learn how to speak English if it's your first language. You just know how to do it. Now, if you want to learn Spanish, it's a whole other process and you're figuring it out a lot differently than someone who speaks it natively. Same kind of thing with computers and iPads and AI and everything. You kind of lose that ability to understand how it works if you just grew up with it. For sure, so it's something to think about. If you're younger out there, keep that in mind. And remember that it changes and it gets things wrong. And if you're older and you're hiring younger people, keep it in mind as well that you need to teach them these kind of things that you're not just relying on ChatGPT to do all your work for you. Don't just tell ChatGPT to make you a PPC campaign and trust that it's correct. Speaker 1: Ain't that the truth. Speaker 2: Gotta learn these things on your own. There may be a day where we can do that, but it's not quite here yet. But all right, let's go ahead and jump on to our next post here. Let me pull up and again, anybody out there watching, I see a bunch of you out there watching. If you have any questions, comments, throw them in the chat and we'll bring them into the show here. All right, so some more fun tariff stuff. I'm sure everybody out there is probably sick and tired of hearing about tariffs by now. I am for sure, but it's part of our world at the moment. So Trump's 50% tariff on Indian goods could reshape Amazon's apparel supply chain. President Trump has imposed a steep 50% tariff on Indian textiles and apparel, which could spike costs by up to 35% and slash U.S. bound orders by as much as 50%. This move is already causing US retailers to cancel shipments and renegotiate supplier relationships. For Amazon sellers in the fashion or private label space, this disruption may impact pricing, availability and sourcing, especially if you're importing from India or competing with brands that do. Sellers should monitor supplier costs closely and consider diversifying sources to countries like Vietnam or Bangladesh, which are still at lower tariff levels. I kind of disagree with a little bit of the summary there. I'm not sure I would start looking for alternative sourcing quite yet. I mean, you should regardless, so I shouldn't say that. But I don't think there's any panic around this as of yet because this new tariff is strictly around the Ukraine war and trying to hurt Russia more than anything. So we'll see how that plays out. But interestingly, so far, most people have been Wrong regarding tariffs to some extent. A lot of people thought it would be the end of the world with tariffs. It hasn't been the end of the world. Everybody thought prices would go up a million percent. They haven't really. I just read recently that only about 4% of products on Amazon are currently at their highest price ever. So it hasn't exploded prices as much as we thought. But this one kind of hurts a little bit, I think, because India was one of those places where everybody was talking about moving to. Vietnam and India and Mexico were kind of the big three to get out of China. So this one hurts a little bit. What do you think, Noah? Speaker 1: Yeah, I mean, knowing especially the fact that like majority of the goods coming from India right now are apparel, right? Apparel, in my opinion, has always been something that is one of the easiest barrier to entry categories for literally anyone, right? And it's because it is so easy to manufacture things, is so quick, so fast, and mostly cheap. I don't think this is going to have that big of an impact though, on the aspect of like cost, because again, it's already so cheap as it is. The big thing I think here as well, I mean, Especially with tariffs, right, the idea at the end of the day is to try and bring manufacturing into the U.S. Honestly, I don't even think 50% of apparel would bring it into the U.S. I was just recently, I'm going to be at Accelerate next month. One of the things I'm going to be bringing there is a bunch of hats, right? So we were getting hats ordered and made and everything. And we were looking at U.S. versus other places to have them made. It was still cheaper to get it from somewhere else and ship it into the country than it was to have like a U.S. manufacturer. And so we ended up just doing the U.S. side just because it was going to be easier and faster. But it was one of those that I was looking at the pricing. I'm like, I don't think realistically this is going to have that large of an impact in the apparel market. And it even still the apparel market in general is so saturated that I think competition itself will still keep prices fairly stagnant. Speaker 2: Yeah, I agree. And like I said, this is strictly designed to try to get India to stop buying oil from Russia basically is what they're trying to do. Odds are it's not going to stay because it was 25% and then it went up to 50% when now that Trump is trying to get Russia to end the Ukraine war. So how long it's going to go on, you never know. I don't think it'll last that long, but it's impossible to say when it comes to war and stuff like that, how that's going to play out. But yeah, it's definitely not high enough to stop people from importing apparel, but it's high enough that it's gonna hurt a little bit and maybe put pressure on companies in India, which then will put that pressure back on the government of India to figure something out. And maybe work with the US or however it goes down, it's hard to say. Never good to see this kind of thing, you know, for sellers to get trapped in the middle of this. When you're in the middle of a war, which essentially, this is, you know, Russia is attacking Ukraine, but in reality, it's Russia versus the US and the EU, essentially is what it is. It's a proxy war, and China involved in that a little bit as well. You know, it's the big gorillas duking it out a little bit and the little guys get stepped on until they figure it out here. Speaker 1: It was like Godzilla versus Kong type of thing. Yes, exactly. All that's happening while everyone, all the tiny little people in the city below are just kind of getting crushed. Unknown Speaker: Yes, yes. Speaker 2: That's basically what's going on. If you're importing a lot from India and I have a new client that is coming on and they're in the soap industry and this actually does affect them because they import some stuff for, I believe it's for the bottles from India. So they were talking about watching to see how this affects the pricing for their soap and stuff like that. So we'll see how that plays out. So I may be able to give you guys a little bit of an update on that down the road here once I work with them a little bit longer and see how that affects pricing on Amazon. But we'll see. They didn't seem too worried about it, but they did mention it. 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. Speaker 1: Guess we'll have to see. Speaker 2: Absolutely. All right. Let's jump on to some more AI stuff and pull it up over here. We go. All right. So. So Corey replaced half his team supposedly with AI and they thanked him. So I definitely have some questions about that, if they really thanked him for firing them or how exactly that worked. But it's pretty cool, some pretty cool tools that he's got here. So that's why I wanted to bring it on the show here. So Corey, a $13 million Amazon seller and coach, shared how he replaced half his team using seven AI tools, freeing up his team for higher value work. He uses tools like Claude for writing, ChatGPT for research, and Make to automate repetitive tasks, saving over 20 hours weekly. Let me just go here. Claude is what he's using for writing, ChatGPT for researching, Whisper for brain dumps at speaking speed, which is, I'll have to check that one out because I haven't used it. Speaker 1: I just tried Whisper for the first time two days ago. Speaker 2: What'd you think? Speaker 1: I love Whisper. It's really, really cool. It beats every other like transcription, even like Gemini transcription out of the water in my and, you know, from what I kind of saw, because it's just instantaneous comparatively to everything. Speaker 2: All right, I'll have to play with it. Notion for one sentence prompts building entire databases. I have not used Notion for databases. I'll have to play with that. I use Grok a lot for writing app scripts to manipulate data in Google Spreadsheets and such, and that works really well, but I'll have to play with Notion. Make automates repetitive tasks without code. Manus handles grunt work, slides, lists, meeting prep, 24-7 assistant. Otter captures random thoughts, summarizes into actionable notes. So for me to add onto this, Grok, like I said, definitely for writing code. If you know how to do any kind of coding and stuff like that, even basics, you can use Grok to write app scripts, PHP, whatever kind of code you need. I've had it create a, I used it to create a transparency code database for my warehouse so they can get transparency codes and print them out. I'm the label printer for the products that we have that have that and that worked really well. I use, what was it? Otter AI, I believe it is, for meetings, giving transcripts, or not transcripts, but summary of meetings and then important to-dos from the meetings and things like that, and that works really well for me also. Do you have any AI tools that you guys like to use over there at my Amazon guy, Noah? Speaker 1: Um, yeah, I mean, we use so meetings, we use Fireflies. So Fireflies has its own AI. Then on sales side, we use Gong, which has an AI that's built for like sales teams, which is great. Claude, we use for writing pretty often. I use Claude every day. GBT, I mean, we're using it for, you know, random tasks here and there. We've been using or trying out Repl.it, which I don't think many people know about, but it's basically a coding tool. It's only for coding. That one's pretty nice. I like that. Yeah, I mean, I think the big thing, and this kind of circles back to our first kind of thing that we were talking about with all the AI in the realm of There's so many of them. Everything is AI now and like there are a lot of crap ones where it's like it's just a GBT wrapper, right? And like there's a lot of just like different things that are calling themselves AI and it's like, well, how much of it is actually just AI versus like, you know, an algorithm that's programmed in and things like that. And so, you know, things like that were like, I think of like Fireflies and I think of like Otter and I'm like, Is it really AI, you know, or is it just like an algorithm with transcription and the ability to understand, you know, asks and tasks and like a chatbot basically. And so, I don't know. I haven't used Manus before. I don't know that one actually. That's the only one from this list I don't know, I should say. Make, I've seen before. They actually, Make was wanting to, we're a big Zapier, no-code, low-code kind of company. And Make was wanting to take over all of our Zaps at one point in time. So I know about them, but yeah, the only one on here I haven't used or tried before is Manus, which I guess I'll have to give it a little shot and see what it's about. Speaker 2: Yeah, yeah, similar for me. WhisperFlow and Manus I have not used. Not really used Make before. I've heard of Make. I use Zapier for different things, so it's probably similar to that, but I'll have to check those three out for sure. And I was wrong. It wasn't Otter that I use. It's Fathom. Fathom is what I use. It does really good for, I'm very impressed when I get off of a call and it's like, it's got all of my to-dos in a checklist and I can check them and it crosses them off and everything and it's really nice for making sure I remember everything from a call and I don't have to sit here with a notepad open writing down notes as I'm trying to talk with clients and stuff like that. So it works really well. Speaker 1: There we go. Yeah. Well, and I think that kind of is similar with all these different ones, but I always wonder with those, if it's like, do they actually have an AI or are they basically just really prompt engineering, like GBT in the background? Unknown Speaker: I guarantee it. Speaker 1: Yeah. Nine times out of ten, it's like, yeah, I can almost promise that's what it is. That's where it's like a lot of these, you know, a lot of these different AIs and whatnot. I'm like, yes, it requires you to have like the knowledge to be able to correctly prompt, you know, GBT in the way that will make it do, you know, the specific types of tasks very frequently in the exact same way every time. But it's still a bit iffy on me to be like, yeah, we have this AI company that does this. It's like, okay, but yeah. Speaker 2: No, there's no way all these companies have the resources, number one, and the server farms to have their own AI. I mean, you know, Amazon has theirs, ChatGPT has theirs, Grok has theirs, Google has theirs. And these are big companies. We're not talking about, you know, million dollar companies. These are all billion dollar companies that We are running these meta has theirs, of course, as well. So every one of these little companies that pop up, they're not building out their own server farms and have hosting their own training their own AIs. They're using one of them as a back end and they're creating the front end. It's almost like going to be an operating system in the future, right? Everything's going to be built on top of A handful of these, one or two of them was probably going to win out. And that's where I was talking about the arms race. I wasn't so much talking about all these little tools. The arms race is really between the two. Speaker 1: Which one is going to become Windows? Yeah, which one's going to be Windows? Which one's going to be iOS, basically? Speaker 2: Which one's going to be Windows? Which one's going to be iOS? Which one's going to be Android? You're going to have one or two that are going to win out. The rest of them are going to go into the background. They'll probably still exist. IBM still exists. They still make their operating systems for servers. And they make plenty of money. But for the most part, unless you're in a niche that they targeted, you don't really hear about them anymore. And that'll be the same thing in the AI world. Speaker 1: Well, and that's actually one of the interesting things where I think being within like our niche of like e-commerce and Amazon, like these are all the tools that are exceptionally useful for us, right? Because they are doing that. Whereas I was talking actually to my mother the other day and she works within health care, within like general health care and stuff. And she had received like a presentation from a company that was a note taker, but it's specifically for Like healthcare and for therapy and things like that. And that one is like something where, yeah, they have to develop their own AI and they can't work off of any of these systems and all that type stuff because like HIPAA, right? And it would be like violation to like send this data to different companies and things like that. So they actually do have to like create. And that I think will happen more and more as the as things go on where The AI companies will, yes, be around, but they'll be very industry-specific, whereas a lot of these major ones that we're talking about are kind of just like the general ones that everyone knows about and can be used pretty much anywhere for just generalized things, whereas the more niche, industry-specific stuff, it'll come out at some point. There was another one a couple of months back I was looking at that basically It was for construction, where it completely automated the entire bidding process of projects for construction, taking what normally would be a three-month process for bidding, boiling it down to a week. That kind of stuff. Can't really do that with any of these. We'll see it unravel as time goes on, I suppose. Speaker 2: Yeah, there's going to be a million tools, AI tools built on these different platforms and they'll get specialized and some will die, some will grow, others will niche down and find a specialty where they can make enough money to keep existing. And you just got to find the ones that work for the tasks. That's going to be an issue. I think a lot of people I'm just used to going to ChatGPT and they go to ChatGPT for everything. And while ChatGPT is a really good general use tool, it can do a lot. If you use some of these tools that are more specialized and niched down, they're going to have better results because not that you couldn't do it in ChatGPT if you figured out the right prompts, but what these tools are doing is they're doing all the work behind the scenes to figure out the prompts. And you get to take advantage of that without having to figure them out yourself. And I'm sure they're changing them all the time as AI changes and at least if they're a good team anyway behind the product. Speaker 1: Yep. Agree. Speaker 2: Absolutely. All right, we got one question that popped in. Let's see here from Hira. Is using ChatGPT Pro version for Amazon PPC keyword research or CO listing valid? Depending on that data, will they give 100% credibility? Well, I would say real quick and I'll let you go Noah, they're definitely not going to give you 100% credibility, but they're really good starting point. You can use them to get a lot of really good ideas and then build on top of that. Speaker 1: Yeah, I don't like using GPT for the actual keyword research part of it. We've done a lot of prompt engineering on our end and internally trying to create essentially a really, really good GPT that can do competitive research and things like that. And I've even worked with some people who are way more in the AI space than me, like John Aspinall and stuff. I'm creating that stuff and doing the research and seeing how we can get that keyword. I still just don't think it's great. I mean, it's definitely if you have like a zero base level knowledge, and again, I think this comes back to original thing of, you know, ChatGPT cannot be your Amazon account manager. But if you have a zero base level knowledge, I think it's okay. We more often than not, what we will do is do the keyword research on our end, and then we'll utilize those keywords by giving that and feeding that into ChatGPT to help with copy optimizations, right? And that's what I think it's a better use case for than trying to say, you know, hey, give me keywords to put into my listing. Right. And yeah, I think that's how it's better utilized, to be honest, rather than the actual research aspects, because I don't think it's not built for that. Right. It doesn't know SEO like it doesn't know Amazon's algorithm. It doesn't know ranking again is a language model. Speaker 2: Yeah, well, it knows of all that stuff, but it may not have all the data that we have access to. So if you can use tools like Jungle Scout, Helium 10, SmartScout, to get the keywords and then kind of narrow it down from your own knowledge. And then also get the data from Amazon, from Brand Analytics and stuff like that. Feed that into ChatGPT or Grok or whatever, and then have it create the listings for you. And then still go through and validate everything because, and this is where it's really important to know your niche that you're in and know something about the products that you're working on. Even if you're just working on it for someone else because I sell a lot of fishing stuff. And there's a big difference between stuff that you use to catch trout versus bass versus freshwater versus saltwater versus marlin or sharks or whatever. It's completely different products. But if you just rely on ChatGPT to write you a listing and you know about the topic and you look at it, you can very easily see now that doesn't have anything to do with trout fishing That's saltwater, need to change that, et cetera. And so that's important. Try doing these kind of things. If you don't believe us, try having ChatGPT write something about a topic that you're super knowledgeable about and you'll see the little issues that it comes up with. It's kind of like when you're watching the news or reading a newspaper or something like that and something comes up. About a topic that you're really knowledgeable about and you're like, no, that's not right, you know. Speaker 1: The age-old thing of being in a conversation with someone or anything like that and they bring up the one topic that you're like the ultimate expert on. I actually do love it in the Amazon space. Whenever like talking with just some random people, I don't know, like out at a restaurant or just, you know, I'm hanging out with people and somebody brings up a topic about, you know, Amazon e-commerce all day. I'm like, You've, you know, hit my ultimate. This is the one thing that you can't argue with me about. Speaker 2: Yes. Yeah. And being in the computer, my background is computer programming, computer networking. So anytime I hear them talking about, you know, computers on the news, I'm just like shaking my head. Like that's wrong. That's wrong. So then I'm like, how wrong are they about everything else? If they're wrong about computer stuff all the time, But yeah, you'll get the same thing in AI. So just keep that in mind. You gotta validate all this stuff regardless. But do the research yourself and then dump that research into AI to give you the output that you're looking for and then validate it from there. Speaker 1: Bingo. Speaker 2: All right, cool. So let's go ahead and jump on to our last one. Some more AI stuff here. Scroll back to the top. So some competitor spying prompts for Amazon sellers. I seen this and I'm like, that is pretty cool because they've got some use cases in here that I think are really helpful for Amazon sellers out there. And you probably see all these extra tabs up here. I actually did some of these prompts myself. So they've got some really good, quote, competitor spying prompts. Really, they're just looking at listings and stuff on Amazon. But I am pretty surprised at what AI has been able to do to pull because we were talking about earlier about how Amazon and stuff is blocking AI. At least at the shopping, but they're obviously not blocking AI from accessing their pages based on these prompts. So some of the prompts that we're gonna look at here, and these first prompts I ran on this door handle. And I just thought of this because we're installing a barn door in our house right now, a sliding barn door. So I seen the handle pop up and I thought it'd be a good one. So I asked Grok and ChatGPT a couple of these. And so this was the first prompt in that article. Analyze the Amazon listing of ASIN. Provide an overview of title, bullets, A plus content, reviews and pricing strategy. Identify what seems to be working and any weaknesses. And so just off of the ASIN, it was able to jump out there and grab the title. And we can verify that is the correct title, I believe. It's a little bit different that it did add the brand name at the front, which they don't have here. But otherwise, it's pretty close to the same. It changed it a little bit. It only gave me one bullet there and it said it doesn't have A+, so that's exciting if this is my competition that they don't have A+. I'm like, okay, I can compete there. Does Gibnick actually not have A+, though? Unknown Speaker: Good question. Speaker 2: Let's see. Yeah, it does. Got it wrong. See, there you go. Speaker 1: So this is actually one of the issues that I was finding when I was trying to do competitor research or create like a competitor research. The thing about GPT like agent, right, is that I think it can't actually from my research, it can't actually view Amazon listing pages. So it always will pop up with the dog. If like you use the ChatGPT agent mode, you'll actually see that when it pops open its window and it tries to go to an Amazon link, it shows with the dog. Instead, what it's doing is it's just pulling all of the cached metadata from the Google search algorithm. So that's why the title probably is a little different because that's a cached thing. And then like from the A+, I'd be curious if they have alt text on this A+, because maybe if they don't have alt text, it wouldn't show up. Again, let's see. Speaker 2: It doesn't look like it. Speaker 1: Yeah, so there you go. That's why it thinks it doesn't have... Speaker 2: Well, yeah, it does right here. Speaker 1: Oh, so it does. Well, I don't know then. Yeah, but this is where I'm a low-grain assault, I suppose, when it comes to these kind of competitor research things. Speaker 2: Yeah, you definitely got to be careful with anything AI as we've already said. It did a little bit of a review of the reviews. Positive, sturdy, easy install, quality look, durable, negative, scuffs easily, outdoor longevity concern. So that's kind of cool. It does say it's getting these directly from Amazon, supposedly. 1797, no discount, list price or subscriptions noted, affordable for premium hardware. So it gave us a little insight in the market there. High ratings for durability, install ease, versatile, no A plus content. So it got that wrong again. And then I just told it to expand and it did give me a little bit extra, but not a lot. Speaker 1: Now, wait, actually, if you go back up to the A plus section, it says right there, not detailed in sources, likely minimal or absent. So I'm wondering if that's exactly what it is. It's basically pulling that data because A plus content isn't cached by Google. Speaker 2: Yep. Yeah, it could be. It doesn't because it doesn't give a link for that. Everything else it gives a link. That's where it's getting that information. It doesn't for that. Now, I do have my grok set to give me minimal responses, which I'm thinking I need to change because it's actually giving me two minimalist responses on some things because ChatGPT's response was quite a bit more detailed. So this is for the same product. It's given me the product, snapshot, the ratings, the price range, and then a competitive price range. Speaker 1: I just noticed it's pulling from Amazon Saudi Arabia there for someone. Speaker 2: Yes. Speaker 1: That's kind of interesting. Speaker 2: Yeah. I wonder if it was just able to access that page. It's got the right star rating and 1.5 thousand, right? Yeah. 4.8, 1.5. And that's probably showing up on the Saudi Arabia page. Let's look. Yeah, it's still showing up there. Speaker 1: Oh, you also notice on the Saudi Arabia page, the brand name is in the title there. Speaker 2: It is, so it's pulling it from there and I bet you there's no A+. Yeah, no A+. Speaker 1: So that's where it's getting it wrong. So Grok wasn't wrong, it's just pulling, and this might be the thing where the better way to do this would be to have it only pull the data from like the US marketplace, right? If that's something you're trying to do. Speaker 2: Yeah, so what we probably want to do is edit this and analyze the Amazon.com listing. Probably something as easy as that could get us a better This is the response from it. Now let's hopefully this doesn't take forever. Grok has been going much slower than ChatGPT lately, so we'll see here. But yeah, now it's analyzing.com, searching the web. So we'll see. Oh, there we go. I wasn't able to pull the full Amazon page directly due to technical limitations, but I gathered enough info to give you a solid overview through secondary sources. So hey, ChatGPT was at least telling us that it wasn't able to accomplish. It didn't just try to lie. We'll see what Grok does. It's still analyzing. Speaker 1: Yeah, so you saw there it said the page seems restricted on the Grok side of things. And that's the thing. I think that... Speaker 2: Maybe it's a Bose product. It's kind of funny to watch it think sometimes. Speaker 1: Yeah. Speaker 2: It's really funny when you have a conversation going, you ask a question and it's like, oh, it asked me to expand. I'm not sure what it wants to expand, but let me see. And then it's like, oh, it's probably from the previous conversation they want me to expand on. It's interesting to watch it thinking sometimes. All right, so I found it there, but let's look at the ChatGPT and see if it got any better. So it did pull the US, that looks like the, yeah, easy install. So that's the US one it got now. The bullets, let's see, high quality construction. No, that's not the same bullets that it has here. So it wasn't able to pull those specifically. So that it got from India and Australia. Speaker 1: It says from similar Amazon regional listings. Bullets highlight. So it's not even pulling the actual bullets. It's like the highlights of similar listings. And again, the no direct data is available about the A plus content. Speaker 2: Yep. Okay. And I wasn't able to access review counts, ratings or comments directly. That's interesting. Because ChatGPT or Grok got the reviews. Speaker 1: I wonder if it's all the reviews initially from the Saudi Arabia marketplace and if Saudi Arabia... Speaker 2: Probably. Speaker 1: Yeah, Saudi Arabia might not be locked down or restricted compared to Amazon US. Speaker 2: Yeah. Yeah, so it gave us some decent information. But again, you know, it's guessing at things if it doesn't know. At least ChatGPT is politely telling us that it's guessing. So that is a good improvement that it's telling you that, hey, these might not be completely correct. But it's giving you some recommendations and stuff like that. Take everything with a grain of salt. Grok is still thinking, so there it goes. So we'll come back. It's got it now. It's spitting it out. Grok has definitely gotten slow lately, I've noticed. Sections on multifunctional sturdy construction. So now it was able to determine that it has a plus. Multifunctional application, sturdy construction. Let's see if that's... Speaker 1: I think it says that there. Speaker 2: Sturdy and elegant. Great look, solid feel, comfortable, high quality. Hardware can be used on. So it's got multifunctional. Three different sizes, quality with a touch of style. Let's see. Versatile, multi-use, easy install with hardware, comfortable grip, stylish black design, specs. So it picked up probably the alt text. I imagine it was able to crawl somewhere because it looked at .com and CA. Oh, sorry, I was looking at the wrong thing. Sections on multifunctional applications, which is correct. Sturdy construction, stylish design, affordability, detailed specs, emphasizes rustic charm and indoor-outdoor durability. So it actually was able to get the information from those photos, either the alt text or the actual photos. Now the reviews are wrong, though. 4.7 implied from snippet 100 plus mentions across aspects like sturdiness. It looks like it was able to get some information on them, but their star rating is wrong. Speaker 1: That's one of those things with all this. I'm wondering how it pulls. It would be nice to see, actually, if we could get a breakdown of how did the data get pulled, where exactly was it pulled from, things like that. Speaker 2: I think you can, actually. If you look at the thoughts for Grok, You can see a little bit more, you can scroll back here and just see its entire thinking process on what it was doing. So it does give you some things there in Grok. I don't know if it does that in ChatGPT. Speaker 1: I think ChatGPT only does it for the agent mode or the research modes. Speaker 2: Yeah. So not bad. Let's jump on to the next one here that I wanted to look at and then we'll probably be out of time here. But this next prompt, I did a different product. So we could see a couple different ones. And this one is, summarize the most common complaints in one to three star reviews of the ASIN. Group by theme, suggest how a new product could address these gaps. So, supposedly Grok was able to go out and get some information. It says, blades dull quickly after multiple uses, motor fails or stops working. Performance, saw tip ineffective for thicker branches, called a joke, jams on debris or thick twigs, shipping, missing parts, damaged packaging or broken screws, new product suggestions, use harder steel blades for longevity, enhanced saw with sharper powered reciprocating mechanism, improve packaging with secure inserts. So assuming that it was actually able to look at real reviews. Speaker 1: Well, that's what the thing I noticed there, though, you know, it's 49 pages. You know, and so I'm like, looking at that, like, why is it looking at 49 different pages, you know? Speaker 2: I'm wondering, so it's looking at individual review pages, it looks like it works very well for trimming brushes. Speaker 1: It's pulling from Wired. Speaker 2: Yeah. So this is an actual review of that product, it looks like, that it was able to find. Yeah, then it pulled... Oh, what do your Amazon star ratings really mean? Sorry? Speaker 1: Can you go back to that one review? It looked like it was a four star, not a one to three. Speaker 2: Which one was it? This one? Speaker 1: I think so. Speaker 2: Yeah, it was a four star that it looked at. It doesn't necessarily mean that it pulled data from that, but that it's using that as reference. That's one of the pages that it browsed. So Wired, what do star ratings, Quora. How does a product star rating work? So I was just looking up. Speaker 1: It's learning. Speaker 2: I need to know what a star rating is. But this is one thing that I do like about Grok as opposed to ChatGPT. It gives you a little more depth and you can see there's a lot. Home Depot is in there I've seen. So Home Depot probably had the same product. Yeah. So I looked up reviews on Home Depot as well, which is kind of cool that it's pulling from various different places. And I said expand on that. So then it gave me not much more info. But let's see how chat GPT did. Underwhelming saw blade performance, blade length or cutting reach limitations, lightweight handle, okay, but maybe too light. And then some product redesign recommendations. So It seems relatively accurate. Somehow it is actually getting access to the reviews. Now let's see if ChatGPT, can you click on this and see? Speaker 1: I wonder if it's because review pages aren't blocked like listing pages are, and that's why it's actually going to the review pages versus the listing page. Speaker 2: Yeah, that's what it looks like. Well, these are all product pages. It looks like different products though. So it may or may not be getting everything from the same product. That's not even Black & Decker, I don't think. That's a slob. Something from China. Speaker 1: Whatever the heck that is. Speaker 2: Yes. But as you can see, it's kind of cool. It is. Use it as a tool to figure things out if you're just looking at your competitors and stuff like that. But don't take it as gospel, right? It's not, as we've been saying the whole time, you can't just assume that it's correct. It was incorrect in several cases. Speaker 1: This is where that knowledge aspect comes into play. Speaker 2: Yes. Yes, so you got to know something about the products that you're selling. If you're a private label person, you're creating a product. This is where it involves creating a product around something that you know. I always tell people focus on products that you have an unfair advantage around. I love fishing. So I know everything about fishing. Well, not everything. I'm not that good, but I know a lot about fishing more than most people, right? And so when I'm looking at a fishing product, I can tell you a lot about it just from my own knowledge. And I can look at the copy and say, that doesn't make sense. Where someone who doesn't know anything about fishing, Like my graphic designer, for example, they create a picture that has nothing to do with the actual product or it's a different fish, I should say, that you wouldn't catch with that product. And so that's where you can get an advantage when you're focused on a passion that you have or something that sets you apart. I remember talking with someone He was retired from the New York Police Department. So I'm like, maybe look at something around tactical equipment or something like that, because you have an unfair advantage. I don't know anything about that world. So me trying to compete with you, it's not a fair fight. You have an unfair advantage. So it's important to look at those and use those kind of things. And it can be crazy things that you might not even think about. You know, you've got an unnecessary knowledge base of Pokemon. And maybe you can figure out something around that, that most people have no clue about. And you can create products around that and sell it and have an unfair advantage over others. Speaker 1: There you go. Speaker 2: Yes. Anything to add to that, Noah? Anything you guys recommend for people over there at my Amazon guy if they're looking at creating products? Speaker 1: The biggest thing that I'm always going to recommend is, I mean, yeah, I think using tools like this for doing product research and trying to understand different products Yeah, makes sense, but my advice is always going to stay the same and it's exactly like yours. Make a product that you actually know about and that you would use. If you would not be willing to use your product or even someone you know everything, if you have a wife, daughter, son, whatever, if they wouldn't use your product or you wouldn't want them using the product that you're selling, What's the point? You know, unless you can know your audience really, really, really well, there's not really a point in trying to get into something. So you'll never know it well enough compared to the person who does know and use that product. And your competition is just going to eat you alive always. Speaker 2: Yes, for sure. Especially if you're in the US or the EU or something like that. And you're competing against, you know, Chinese sellers that are just Throwing everything and anything against the wall to see what sticks. You've got to be building a product that's really good, building a brand, building an email list and everything that goes along with that. And if you're not at least interested in whatever you're building, You're probably not gonna wanna do all that and do what it takes to build a real business because Amazon, the days are past where you can just go on Alibaba and grab a product and throw it up and hope it sells. You've gotta be a real business, you've gotta be building a real brand and building high quality products that people are gonna wanna buy and pay extra for because if you're not in China, Your product's probably gonna be more expensive than the ones coming straight out of China. Speaker 1: Ain't that the truth. Speaker 2: Yep, for sure. All right. So it doesn't look like we have any other questions or anything like that. So I appreciate everybody out there watching. Been awesome having you. We do this every Friday at 1 p.m. Eastern. So please join us in the future. Jump on the live and get your questions answered, whatever the case may be. But we go over the top news and top insights from the Amazon world. So Noah, I appreciate you coming on the show. We'll definitely see you again in the future. Speaker 1: Absolutely. Always a pleasure, Todd. Always a pleasure. Speaker 2: Have a great weekend, everybody. Unknown Speaker: 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.