EP #305] [ENG] - How to leverage AI to scale your content game on Amazon - Daniela Bolzmann

EP #305] [ENG] - How to leverage AI to scale your content game on Amazon - Daniela Bolzmann Unknown Speaker: Welcome to The Ecommerce Lab By Ecomcy. This is the place for everything related to Amazon private label and e-commerce. Learn exactly what you need to start or scale your business. Get insights from the top industry experts who will discuss the latest trends and best practices in the world of Amazon. From choosing products and sourcing from a supplier to setting up your Amazon account and marketing your business, you will hear it here. Let's get started. Here is your host, Vincenzo Toscano. Speaker 2: Hello, guys, welcome to another episode of The Ecommerce Lab By Ecomcy, the place where everything related to Amazon, FBA, private label and e-commerce. My name is Vincenzo Toscano. Founder and CEO of Ecomcy. And today we bring in our special guest. Her name is Daniela Bolzmann. She's the CEO and founder of Mindful Goods, which I would say is one of the top solutions out there when it comes to really help you figure out how to do the branding and all the content around your products on Amazon correctly. And that's why I wanted to bring Daniela today, because I actually saw her presentation in Iceland as well. I was very impressed about, you know, all the things that she's been doing in the space. I just wanted to hopefully get some of those cold nuggets on the show today in terms of, you know, what is happening in terms of Rufus, in terms of AI, in terms of how to do a branding and really make your pros stand out and hopefully you can implement some of those things so and your brand to have a successful 2025. So Daniela, welcome to the show. How are you doing? All right. Speaker 1: Thanks for having me. Speaker 2: Glad to be here. Pleasure. Thank you so much for being here. You had fun in Iceland? Speaker 1: I had a blast in Iceland. If anybody is considering going to one of Kevin King's billion dollar seller summits, I highly recommend going. I was actually quite shocked at how well it was planned and how curated the group was. And I'm sure you feel the same way. It was like such a really well thought out, small, event that was just top notch from start to finish. Every single detail. Speaker 2: Very impressed. I agree. I think the opportunity to Also midway or like-minded people that fly from all different countries that come into the same roof and share ideas and have fun together. I think it's definitely something that is unique and you know every single year Kevin knocks the ball out of the park you know so it's like always. Speaker 1: I know and like fun fact If Kevin's listening, I may have heard, he might have like sneak in a little detail that in the future he hopes to do one in Italy in a castle someday. And if that's happening in 2026, sign me up. I want to be first on the list. Speaker 2: I know, me too, me too. Anyway, it's a pleasure to have you and as I was saying, you know, I have seen all the great things you represent in terms of I had to really make a product standard from the competition. I think that's really the direction most brands have to go, especially in 2025, you know, with the competition increasing now, having this situation with the tariff, where pricing is going to be a huge indicator towards if your product is going to succeed or not. The only thing that you have in your hands now is to have good content and branding. When it comes to that, we have seen a lot of things happening. Rufus has been, you know, one of the major topics in that department. People trying to figure out how to optimize when it comes to Rufus, if Rufus is here to stay. And, you know, the conversations are very interesting. And I think I would like to start with that when it comes to today's episode, because I'm sure with your clients and everything that you're doing, The first question most people are asking you is like, how do I make sure all my content is Rufus proof, right? So I guess let's start with that. What are you seeing when it comes to that? Speaker 1: This is such a great question. My favorite hot take for this, by the way, is there's an e-commerce leader that I was chatting with and he has a huge catalog and they've obviously been doing different types of tests with Rufus and whatever. And he said, you know what, if you were doing listing optimization right from the beginning, there's nothing new that Rufus is telling you. Like 95% of it should have been done in the first place if you were doing it right. But there is still little gems that you can still pick up from Rufus. So some of the best practices that you can do right now that are very simple is going to your competitors' pages and seeing what shows up in Rufus as the top searched items. Ask Rufus what the top searched items are, top questions people have about this specific brand and this specific product. Take screenshots of all those questions. Do that on your own listing as well. Even from the main search results page, conduct some queries just like that, asking it what are the best types of products for this specific, you know, what are the best brands for this specific type of product? Why are people searching or why are those the top brands and kind of get those queries going and try to understand what Rufus is paying attention to. Then in order to bring that stuff into your listing, you just want to make sure that the creatives and the copy that you have on your listing is actually answering those queries, right? Like it's really simple stuff. Like for instance, we're working with a brand right now and a few of the top questions were, does this have dairy? Is it gluten free? And is this vegan? Well, If you're doing, if you already had that in your listing copy, Rufus is going to be able to scan that and say, and give the proper response. But the other thing you want to do is if you know that's a top question, a lot of times customers are not going to read the text anymore. I mean, they were never were reading the text in your listing anyways, if we're being honest. So if we know this is a top question, then we should definitely be answering it visually as one of the first, second and third things. And I'm here to talk to you about how to make sure that your product image is visible in your product images and in your A+. You never want to assume that people are only looking at your product images and you never want to assume that they're only looking at your A+. So you should definitely put it in both places and make sure that it's visually able to see it. And then Rufus will be able to basically show that image to them when they ask that question in the search results, which is great. Speaker 2: Yeah, I think I totally agree. One of the reasons why Rufus has been taking so many people by surprise is because in the past, a lot of people had the belief that optimizing a listing was pretty much keyword stuffing and just put as many keywords as possible, just put some random images and that's it and the intention and the message beyond. Behind this, you know, copy of the content wasn't really resembling the actual semantic meaning that you need to, you know, actually connect with the right kind of avatar that's looking to buy your product. And I agree. Now it's all about making sure your images are on point. You have the right kind of people on the images that reflect the avatar, the lifestyle, which is something I get a lot of questions and I want to go a little bit deeper into that. Like we're seeing that Rufus now has not the The potential, of course, to read your images and depending on the lifestyle and who you have in your images, that's effectively who is going to potentially recommend the product to. So how important you're seeing now, more than ever, the elements that are taking effect on images. Do you see that being a huge factor when it comes to affecting the results when it comes to refills as well now? Speaker 1: For us, I think the type of creatives that we've done have always leaned in that direction anyways. They've always been very lifestyle-driven, very model-driven, very direct-to-consumer vibe because We know that that's the thing for us that separates our clients from the pack. That being said, most of our clients are direct-to-consumer brands, right? So they're ecommerce-heavy, less of like the private label type brand. So when they do use this type of creative, it does help them with their sales. We had an interesting case study that we released yesterday on our site with a brand called Jiyu and it's a Korean skincare line. And the interesting thing about that is They were initially running ads related to the actual product and the search terms most related to what the product is. But when we think about shifting towards like what our consumer is actually searching, typing in the search box, we need to be thinking through like a problem solution lens. So like for me, if I was searching for a product like Jaiyu, which is toner pads, right? It's basically toner pads for your skin that you wipe on your face. I most likely wouldn't be searching for toner pads. I would be searching for whatever I heard that this product does as a solution. Right. So like I would I would probably be searching for And we're going to be talking about pads for anti-aging, pads for anti-wrinkles, pads for less wrinkles. Right. And so when you think about that, we want to make sure that we are using that sort of language in our creatives as well so that we're showing like these are the this is a problem that people are facing. This is the solution that we're giving. So, tying those benefits into your story, like weaving them throughout your creative. That's one thing that we noticed when we pivoted to that type of language in the creatives. It started serving up in Rufus, right? Whereas like if you're just putting toner pads, that's really only going to help you show up for questions related to toner pads or searches related to toner pads. But there's so many other ways that people search related to the problem that they're trying to solve. So we need to be thinking through all of those types of problem solution languages that people are using and that's how Rufus is going to be serving that up. Speaker 2: Awesome. And I guess once all this optimization is done and all this is implemented, how do you make sure, I guess, from an algorithm point of view, that's where Rufus is actually reading? Because I know some people now bring into the conversation things such as Amazon Comprehend, Amazon Recognition. Is that something that you also recommend? Like once you do all this homework and do all the images, as you mentioned, run into a system like that? Speaker 1: Personally, I don't think you need to. So I was really excited by recognition and we were doing early tests on it. But at the end of the day, when we think recognition was first using it, it wasn't super accurate. Now it is. And you can generally assume that it's reading your image accurately now, right? Especially since a lot of people are going to be getting into making their images with AI and you're instructing the AI on exactly what you want the image to be. So assuming Rufus would be able to read exactly what the image is supposed to be, right? So it will be translating almost equally in the future. So for us, we haven't been really focused on using recognition, but it is really easy to see if Rufus is pulling up is understanding your images, right? Because it's going to serve up the image as a response to that query, right? So if I went into the Jiayu listing and I type in, does this product help me with my wrinkles or with anti-aging? It better show me some sort of solution in the form of an image, right? And it does. And so if you go and ask that and Rufus is giving you a proper response that is correct and is showing you the right image, then you nailed it, right? Speaker 2: Exactly. Speaker 1: Ideally, every question that you prompt into Rufus should bring up one of your images. Speaker 2: Yes, or the copy at least, you know, be able to transmit that to your messaging, which I guess, The question that people is asking themselves is, of course, everything makes sense. And they start to understand, you know, the methodology behind how to optimize for real foods. But the reality is a lot of people as well, they really struggle to understand what is the avatar they're actually targeting. And that's something, you know, I find all the time when I have calls with people, like you ask them, what is the actual persona you're trying to target? And sometimes they don't know. They don't know if they're targeting specific gender and specific age, location, profession, lifestyle, All of that. And sometimes it's a question that can really make the difference in terms of how you do your branding in the first place. So I guess from the experience that you have with working with all these brands, like how you sometimes answer that question if you don't know who is actually your avatar, you know? Speaker 1: Oh, that's such a good question. And we do get this a lot. So there's a couple of tools out there right now that can help you understand this. One is if you're already selling on Amazon, you can go into your brand analytics and pull the demographics. It'll show you if most of your buyers are male or female, what age group they are, all of that. So that's amazing. If you're not already selling on Amazon, then you don't have this data at your disposal. What you can do is use third party tools. One of them is GigaBrain, which searches Reddit Reddit threads. The other one is called voc.ai, which is searching Amazon reviews. And what you can do is basically pull the reviews or any sentiment related to your category or your product or hopefully your brand name if people know about your brand name, if there's enough awareness about your brand name. If not, there's not going to be enough. But if there's generally enough conversation happening around your type of product or your category in which you sell, You can basically ask in GigaBrain AI, very similar, you can have a chat very similar to how you have a chat with ChatGPT. You would say pull up all the conversations around Korean skincare. Great, so it pulls up all those conversations, right? Now it says, now you can tell it, pull me all the conversations Pull together based on all of these conversations build me a psychographic analysis of the people who are buying Korean skincare products and it'll pull all that data and say most likely based on these discussions. This is this is the person. This is the age group. This is this is the types of things that they're into in their lifestyle. These are the places that they go. These are the types of things that they care about. These are the things that motivate them to make purchases, right? And using that, you can start to understand who the buyer is, right? If we know that this person is most likely into other wellness activities like yoga and healthy green juices, well, great. We're going to show a healthy gym girl We're going to show more of those yoga settings or that type of lifestyle yoga vibe in that imagery. It just helps you paint the picture of the type of person and the types of things they like to do so that you can build those into your creatives in a really smart way. I love GigaBrain for that. And there's other tools that are out there, right? It's just pulling as much of that data that's out there into one place so you can start to understand it and analyze it and build that profile. Speaker 2: Now, when it comes to Rufus, of course, a lot of the conversation has also been shifting towards how we use AI to make the content around our brand, right? In terms of our main images or A plus content and all of that. And there's a lot of mixed feelings when it comes to that because I strongly think that AI is still not in a position that could fully make that for you. I still think that you need humans to really do the final touches and the brand guidelines and really give the strategic approach. But AI can do a lot of the heavy lifting sometimes. So I guess what I want to What I want to hear from you is, given that you, of course, are doing this on a daily basis, I'm curious to know how AI has been changing your workflow. Are you using AI into creating some of the stock images that you use for your designs? Just give us some examples because I'm sure a lot of people are listening to this. Sometimes they're not aware of what they can use AI for. Maybe this can open the Pandora box for them. Speaker 1: Yeah Really great question and high-level thoughts on on AI in general is I think everyone right now thinks that or the the messaging right now is that all of us creative agencies should be scared. We're gonna lose our jobs. I can tell you no one on my team is scared at all. We've been we've been using AI in our imagery with our client projects for Probably almost a year now. And we use it in selective cases and with selective categories. It's not a great fit for every product. The technology is great, but it's not 100% where you can absolutely not tell the difference anymore. You know, it's like the difference between when people first started using renders on Amazon and they started using really bad renders. And of course, Amazon's like, don't do that. We don't want low quality renders on Amazon. And so they told sellers they couldn't use renders. Well, we know most sellers, especially in the pet category, use renders. It's not that people aren't using renders. And it's not that Amazon doesn't want renders at all. It's that they don't want low quality renders. And that's what's happening with AI is that everyone has access to these tools now. And if they start putting out bad quality renders or AI imagery, that can actually hurt your listing. So that's one of the things that we need to consider is that what we're putting out, is it quality, right? And so that's really one of the lenses. We had a client that came to us, I would say a couple years ago now, and their entire listing was made with 3D renders. And they weren't bad renders, actually, they were pretty good. But something about the listing was losing people, like it was just not trustworthy because you're looking at an entire listing of basically 3D generated images that are not real. It's like, have you ever seen those like renders of apartment buildings before the apartment building has a photo? Like you know it's not a real photo, it's a render. It looks fake. Yeah, it looks fake and you're kind of like, oh, I kind of get what the apartment's gonna look like, but you wouldn't go and spend like a half million dollars on an apartment without really knowing. Right, you're like, you don't fully it's like, it's like slightly off, you know, and it can be beautiful, but slightly off. And so that's what happened with this listing in particular, it was all renders. And so what we did at the time, because they didn't have a ton of photography, We took some of the photography that they did have. We took some stock imagery and we took some of the renders that they did have. And we kind of filled in the gaps to create something that felt more authentic and more natural. We brought in some of their UGC images. We brought in photos of the founder. We brought in shots of the product, like real shots of the product and it mixed it with the renders that were good. So it didn't feel like everything was fake. It gave you like a mix of everything. And that's what I think is gonna be a good Good principle to follow with AI as we're getting into this like, A, it has to be really top notch, really good quality AI. Two, it shouldn't be everything on your listing is AI unless you're in a certain category where it's really hard to shoot the product. It's really hard to, it's really expensive to shoot the product. It takes too much time to ship and shoot the product. Like there's things that are like, obviously, you know, we worked with a category of, it was like holiday ornaments. Well, When your back's against the wall and you need to shoot, you know, not just even a thousand ornaments, but let's just say 5,000 ornaments, you don't have time to be sitting there doing a photo shoot. You have to use AI, right? So there's certain catalogs, certain categories, certain things that we need to look at and say, when is the best opportunity to use AI to your advantage? And when should you not? And even then you can still incorporate more stock elements to make it feel more natural and not fully AI generated. I think that's going to be a really, really big, really big focus for a lot of people is like kind of making sure you're following those principles. But beyond that, I think I shared this in my in my presentation in Iceland. The thing that I'm even more excited about, I'm less excited about the one-off images for Amazon listings. That's fine. It's going to get better. It's cool. What I'm most excited about is from an agency perspective, we had clients that would come to us and they worked with us and they were like, we want to keep working with you, but we have We have over 5,000 SKUs, 10,000 SKUs and we need to get all of this overhauled and we need to do it for a budget. We need to get it all done and there's a time constraint. How do we get this all done? And our team was manually and still manually makes all the master templates for every client. So then to scale it out, we would have had to hire a team of junior designers that would have just been dedicated to that project working month after month after month. And I'm here to talk to you about how to get all those designs done, right? And then you have to quality check all of that and there's gonna be human error involved. And so that is super time intensive. Well, now with AI, we're basically able to train the AI on our master templates that we designed specifically for that client. And we're able to scale it out. I mean, we're talking like thousands of SKUs we can do in hours or minutes, Depending on what the limitations are with our design, we can get it down to hours or minutes. It's crazy what is available to us now and we used to have to turn those types of big mega brands away and now we can work with them. That to me is where it gets really exciting because actually if you think about it, that leaves our creative directors to do their best work. And we're at the front end, you know, and those junior designers, nobody wants to be sitting in a closet, like painting the same thing all day long every day. Like we're, they're artists at the end of the day, they want to be creative too. So like, I don't think in that case, AI is really taking anyone's job away. We weren't getting that job anyways. Now we're able to get that job and we're able to help those brands scale and do it in the best way possible because we're still creating this like very custom tailored thing for them, you know. Speaker 2: Yeah, yeah, I agree. I think the benefit of AI now is that a lot of the heavy lifting that you sometimes will have to sacrifice your time into like, you know, doing these images now you can do with AI, you can now repurpose that time to do more the strategic and the actual concept, which is where really the magic happens, you know. And I 100% agree. It's also the scale at which you can work. Like before, if you wanted to do 100 listings, it would take you hundreds of hours. Now you can reduce that significantly and make a better job because you're able to effectively, you know, deploy AI to do the heavy lifting and you just focus on the areas that will add those elements and make the difference from a branding perspective. So it's a win-win on all fronts. I think the fear comes because You know, of course, there's always going to be people out there that the only added value was that they could only do the heavy lifting, but they couldn't do beyond that, like branding and concept and everything. And that's, of course, something that is becoming now more and more appreciated because of AI doing the heavy lifting. But yes, I think the examples that you gave is a clear indicator of how you can embrace it. Effectively, make your overall operation more efficient. Speaker 1: The other thing is if you look at split testing in general, that's the third area where we're using AI to our advantage, right? Think about the way that we design like A plus content right now. It's a really heavy lift, right? And it's built like a landing page. It's got carousels. It's really, really polished design. And same thing with our product images. We might create like a variant of one or two of the images, split test the stack to know the order, right? But beyond that, that's where we're kind of limited in our testing. Well, with AI, if we could feed the AI our master template, which we can do now, We can then tell the AI, I need, you know, 10 variants of this with this type of messaging and pointed at this avatar. I need these 10 variants of this one for these different avatars. And then you can literally just take those and split test them month after month on manager experiments. Or you could test them on your own on something like IntelliV or Product Canyon. Speaker 2: Yeah, I think that's a very actually good point like split testing is something that before it was so costly and sometimes consuming. And now with, you know, a couple of clicks, if you have the right tools, as you mentioned, that you cannot automate a lot of that, which that means effectively be able to, you know, improve your revenue profitably also faster, because now you're able to figure out your conversion issues on the spot, which before would take you longer. And time is money. Yeah. So very important. Cool. Now, I guess to start concluding, we talk about Rufus, we talk about AI and how that's changing the way we pretty much navigate the whole content concept. What are things you would like to briefly mention, I guess, that you think is something people should know when it comes to branding and content to be aware of for 2025? Yeah, I think one of the biggest things is just making sure that. Speaker 1: that you're not getting distracted by all the shiny objects, right? Like at the end of the day, we've been using AI for over a year. Has it changed our process and how we do things? No. Do we still stick to the foundations of what we do? Speaker 2: Yes. Speaker 1: Like it's still for us, there's three steps that we follow every single time. It's SEO and copy to help you get found. It's main images to help you get clicks into your listing and steal that click share away from your competitors. And then making sure that your product images, your A plus content and your storefront is built out to the best standards with the data driven information that you need for conversion. Like if you follow those three steps, You're golden, right? And then you should, I mean, the Rufus stuff is going to work for you. The AI imagery, you can supplement where you need it. Like those are the foundations that haven't gone away. Like that process has not changed how we build listings and we split test with our clients. We're still seeing over 80% of our clients that are increasing their sales anywhere from 20% to 2,500%. By following just that framework, just one, two, three. So if something's not working with your conversion, it's just getting back to those basics. It's not about thinking like, oh, I need to go do this whole Rufus special optimization. No, it's like, Just look at those three things. And Rufus is a part of that now. You know, like, it's not like you're having to change your whole way of doing things because of Rufus or your whole way of doing things because of AI images. No, it's like those those things just made it easier for you to do the main foundational work that you should have been doing in the first place. Speaker 2: Yeah, I agree. I agree. So I think, yeah, that's a perfect place to conclude today's episode, Daniela. I appreciate, you know, your time and your knowledge. And I'm sure, you know, some people might want to reach out to have a deeper conversation or figuring out how You can help with all of these things. So tell us how people can reach out and find you. Speaker 1: Yeah, you can find me at mindfulgoods.co. You can find me on LinkedIn. At the bottom of our website, we have a mini audit. So if you want myself or one of our creative directors to actually look at your listing and help you out with a bunch of suggestions of what we recommend for you, you can get that on our site at mindfulgoods.co. Speaker 2: Awesome, Daniela. It's a pleasure as usual and hopefully see you soon at some of the upcoming events. Thank you so much. Thanks so much. See you soon. Thank you. Bye-bye. Have a good one. Unknown Speaker: Thanks for listening to The Ecommerce Lab By Ecomcy. Be sure to subscribe so you don't miss an episode. While you are at it, we would appreciate it if you could leave an honest rating and review on Apple Podcasts, Spotify, or wherever you listen. That will make it easier for others to find out about the show and benefit from it. Want more? Visit our website at www.ecomc.com where you can get your first consultation for free. Or find us on Instagram, Facebook, and LinkedIn at ecomc.