Stop Spending Thousands on Product Photography (Use This Instead)

In the past to do a fashion photo shoot, you have the model, you have the makeup artist, you have the stylist, you have the production manager, you have the lighting guys. There's a lot of people and a lot of cost. And now, so basically what you're saying is that I either choose a model from your stock library of models or I can shoot my own model and then I just need to shoot the individual pieces of clothing or accessories and then I can interchange these at will. Is that kind of how it works? >> Exactly. We try to yeah simplify it as much as possible and the input image can be very simple. It can be taken by an iPhone, you know, like just like a very simple image. You can select the model from the library or describe what kind of a model that you would like to see as well. And like consistent poses, backgrounds, injecting like all the details. >> Who owns the copyrights? I've heard different people say different things. >> What's the key to getting the quality that you're getting? I mean, anybody can go to a lot of these tools and create images, but the quality, like looking on your website here, is is really good. You're watching Marketing Misfits with Norm Ferrar and Kevin K. >> What's up, Norm? My good old buddy. How you doing, man? >> Yeah, we haven't done this in a little while. >> Uh, well, we do it every Tuesday. What are you talking about? >> Well, we do. We do it every Tuesday. >> It's We actually put an episode out every Tuesday, but we Some people don't realize we record these in in batches. So sometimes it's not the Tuesday that it actually comes out that we actually record it. But uh that's that's all good. You you've been uh since since we last recorded, you've been uh doing a few things. I've been doing a few things. Uh I guess uh and I'm sure we smoked a few cigars since then. >> Oh yeah. Yeah. And the Big Smoke Miami, which you missed. Sorry about that. >> Yeah, but I was in London smoking a biki when you were at the Big Smoke in Miami. I was in London cuz I can't get that in the States. Bahikis, for those that don't listen, Bahikis are a Colombian cigar that's actually um well Nor. I'm sorry, not Col. Why did I say Colombian? Uh Cub, they're a Cuban cigar. I guess I live 10 years with a Colombian on my mind. They're Cuban cigars uh that uh are very expensive and very hard to get. And in the United States, you can't buy Cuban cigars. So, I was in London. So, I went to a little lounge and London happens to be one of the more expensive places in the world to actually buy cigars as I found out and Norm had warned me. And so, I spent a pretty penny uh for it's about 700 bucks, I think. Um 7800 bucks for a single cigar. And uh you know what? It was actually really good. It's not something I'm going to smoke all the time, but it's something that was really good. And you know what? Since you're such a good buddy, Norm, I actually brought you one back, too. >> Get out of here. >> I have two of them. They're the smaller ones. I smoke the bigger one. This the three. You get the $400 one. >> Oh, all right. This is the $400. >> Next time I see you, which I don't know, maybe in August for Market Masters or something like that. Uh I I got I got a treat waiting for you. >> I'm on a flight tonight. Just'll be there shortly. >> But you know, you know, have you seen though what's happening with images lately out there? Have you seen like the new Codex thing that uh OpenAI put out and what you can do with images? And you know when Nano Banana 2 Pro I think came out like 6 months ago, everybody was jumping up and down going this is the best thing since sliced bread. Now everybody's like, "Ah, Nano Banana Pro 2, that's child's play." Now look at this codeex thing from Open AI. It's just seems like images and and video in the AI world are just moving at a rapid rapid pace on what you can do. >> Yeah, absolutely. And you know, our guest, I was doing some research. You know, usually we come on and we just rapid fire questions, right? But I decided I'm going to do some real like research on the website. And it was phenomenal. Wait till you see the photography and the images that you can drum up with AI now from her site. So, we'll get right into that. Yeah, it's it's we're going to be talking about imagery today and images and how what you can do with AI uh to create uh your image stack. And um so we have a we have someone coming on that has been doing this for a while. She's Ivy League educated, helped to grow a couple other companies and now she's got a company called Savannah uh with one N. Savannah AI.AI and they're they're doing some images. So we might as well bring her on. Norm see if you can pronounce her name. That's That's what the That's to see. See, I'm going to give you that shot. >> Well, here's Zarah Soy. >> You did pretty well. I think it sounds good to me. Yes. >> But I got your first name, right? >> Yes. >> Okay. And how do I say your last name? >> Soy Salt. Yeah. >> All right. Perfect. I'll still mess it up. >> How you doing, Zar? Nice to have you. >> I'm good. Thank you so much. I'm very excited about this uh podcast. Thank you so much for inviting me. Oh, you're very welcome. >> What's going on in imagery out there right now? I mean, what's just just for those that are listening that, you know, it's always surprises me. There's about half the audience that's always like, "AI is the greatest thing since sliced bread." And the other people are like, "AI sucks. Don't use AI. It's really bad. It's horrible." >> It's a fad or a trend? >> Fad. It's a trend. It's taking away jobs. So, photographers and stylists are going to go out of business. So what what's happening though when and and what's your opinion of not not necessarily your tool and what you're doing but just overall of the whole AI >> uh environment right now. >> Yeah, I mean it's uh pretty like fast evolving like really fast changing uh environment right now. Uh like things are like growing getting like advanced exponentially. uh we've never seen this kind of a pace in any kind of technology before. So it's really really fascinating but yes it is also frustrating in a lot of the uh like kind of yeah uh areas that yeah we need to be careful about but uh yeah in just in general like things are going like really fast and in terms of image generation as well uh the quality has like surpassed like a really high um like level. So it's started to take place like in more real life like cases uh like industry examples uh like it started with more like kind of memes like more you know like fun images content but nowadays like um as you mentioned like a lot of the models are super advanced right now. So it is being used for um like real use cases by brands by a lot of companies, by advertising, by um a lot of visual production related uh projects. >> And when I went over to your website and I saw the models and then uh the different uh clothing that they were wearing, >> you can't tell the difference. It's it's 100% You cannot tell the difference. >> It's crazy where it's gone. Yeah, it's >> Did you see the one with six fingers, Norm? >> Oh, yeah. And Yeah. >> Remember those days a couple years ago? You you'd do something, it'd be like some deformity somewhere. >> Yeah, those days are gone, I think. >> Yeah, that's for sure. It's been like pretty hard to like distinguish it if it is real like actual image, actual human model or if it's uh generated by AI. Um and then yeah, we've been working with uh like fashion companies, apparel, clothing brands to basically help them um to do like their photo shoot processes much more kind of efficient, you know, like cost effective also. Um and then helping them like to to like streamline that process to make their lives easier, >> you know, right now. >> Yeah, go ahead. I was going to say right now the prompting, you know, sometimes it's still this long, right? >> Mhm. >> With your system, does it chop the prompting down in half or is it uh still you got to write a book? >> No. Yeah. So, uh the default is actually there's no prompting at all. Uh it's we basically so when we started this platform uh we worked with a couple of design partners like big brands like pretty closely and then understand their pain points their like main workflows. So we basically like uh created more automated solutions in the platform uh particularly designed for their you know like daily um like needs. So that's why they actually don't need to write any prompt like they usually just upload a clothing image let's say it's all template based and then they select the model background um and then they can inject like uh integrate their own branding like visual DNA as well like you know background color lighting like these kind of details and then it generates the results like ready to be used uh for their let's say online shopping like uh online visuals for their website landing page social media. >> Kevin, did you hear I I was outside having a cigar yesterday and I was watching uh some news cast and um Anthropic is being valuated at a trillion1 trillion dollar. >> Yeah. Yeah. And yeah, that Yeah. A trillion dollars and then Elon Musk is about to be the first trillionaire once SpaceX goes uh um does their IPO. But yeah, so >> yeah. Yeah, it's it's crazy what the valuations are happening on here. So on Savannah though, you sp mostly feature it's mostly focused on apparel and clothing. Is that correct? Or does it do other things? But I'm looking at your website. >> It's primarily like models and clothing. Is that what your your niche is and your focus is? >> Yeah, that's like the main focus. I would say that's where we started uh applying the solutions but then we've been also expanding our focus onto like makeup, beauty, uh a lot of accessories, jewelry companies, uh and also home goods too including furnitureures, rocks, carpets or so. >> So do you have a background in fashion or anything or or where where did you where does this come from? not directly but uh yeah I used to work at consulting before uh and then during that time I um had the chance to work a couple of like fashion companies retailers uh globally like back in Turkey in in Europe in Milan in Singapore. So um I got to observe uh like and see like their processes and kind of like pain points like uh across their operational processes. That's where it came from to me. Yeah. So, let's walk it through a photo shoot. So, in a in the past to do a fashion photo shoot, you have the model, you have the makeup artist, you have the stylist, you have the production manager, you have the lighting guys, you have the set, you have uh someone's got to do all the wardrobe management, the steaming of it, everything. Uh, and then you have the studio or location. And so, there's a lot of people and a lot of cost involved in doing that. I Norm and I both have been involved in that that kind of stuff in the past. And now, so basically what you're saying is that I either choose a model from your stock library of models or I can shoot my own model on a it looks like a lot of these on your website are on a white psych background or something. Uh I can shoot the model and then I just need to shoot the individual pieces of clothing or accessories >> uh uh on a flat background or have a tool do that or do it a little light studio or something and then I can interchange these at will basically in different poses, different backgrounds, different situations. uh for the seasons or for uh the the type of customer that I'm targeting or or whatever. Is is that kind of how it works? So, you upload >> a bunch of pictures of shirts or bracelets or shoes or whatever it is and then you upload you say this is the model and then it just puts it all together. >> Exactly. Yeah. It's uh we try to yeah simplify it as much as possible and the input image can be very simple. It can be taken by iPhone, you know, like a phone um just like a very simple image I would say or some of the brands that we see, they use the images from their factory uh from the sample uh products for example. Uh it doesn't have to be like only you know like flat or like on a clean background. It can be put on a mannequin. It can be put on a table or so. Um and then you just like take a picture of it, upload it and then uh yeah you can select the model from the library or describe what kind of a model that you would like to see as well. And the focus is yeah mainly creating like the PDP product detail page visuals um and like consistent poses, backgrounds, uh like injecting like all the details together and not only like for like one product like one SKU but the system is designed in a way to um to basically able to produce all these images for like tens like hundreds of SKUs at once. Hey Norm, I've got a quick question for you. I'm trying to manage all my affiliate and creator programs from Amazon, from Shopify, from Walmart, but it's just a freaking mess. I mean, I've got reporting coming from here and there, all these different Slack messages. Do you know if there's like a unified dashboard where I can do this all in one place? >> Yeah, absolutely. And you're right, it is a mess. A lot of brands are complaining about that. But there is a place that has a solution. is called Lavanta and they let brands recruit partners, track performance, manage payouts, send product samples, and even run creator programs across every major marketplace all in one place. And guess what? Brands can spend less time on tools and more time making profit. >> Is that the one that you sent me a link for like a 10% off coupon or enterprise plan a few few days ago? >> You got it. Oh, cool, man. I think I've got that link here. Was it lavanta.ioisfits? >> Yep, you got it. >> le van na.iomisfits. >> Awesome. I'm going to go uh go hit them up right now and get that 10% off. >> Perfect. Me, too. >> You know what this reminds me of? Back in 1995, just prior to that, when you had to go and install a computer program, you had the batch file and the config file and if you installed it improperly, it would slow down your computer. It was a nightmare. >> And then 1995 came around and it was plugandplay and it was just so simple. And I think that's where it started. Like for back in 2020, 2023, you still had to create these big config files, prompt files. And now what you're doing, it sounds like it's more of a just a plugandplay. >> Yeah, it's true. Uh the goal is like to basically provide that like batch generation um as easy as possible uh to the brands because I mean a lot of the current platforms they do like single image generation you know like their point solutions and you have to like write a lot of prompt things like that but it's not uh very easy to create you know like hundreds of product images like same high quality consistent results. So um yeah that's where our platform comes in to help with that basically uh once you configure like set up the requirements. Um you get all the results in a batch. Yeah. Very fairly quickly easily. >> How sorry Kevin I was just going to do a followup. Uh how is your technology different than all of these other AI platforms out there right now that are producing photos? >> Yeah. So we have our own customtrained model particularly for clothing actually uh we have our own data set and then we've been working on it uh for a while now. Uh but also at the same time we build uh certain pipelines for particular solutions like particular functionalities. Um so we are also using using lots of different models too. But the key is like the basically the the source like the key secret is basically to be able to create that workflow and that gives the best results for different use cases and for different types of products. Like for example, creating a flatlay ghost mannequin is a different um challenge versus creating like on model images is different. And for different product categories, they have like lots of different details as well. Uh we do a lot of design related workflows too. So it's all about like finding the best solution and then piecing them together and then creating like a robust custom workflows for for the user. And then you need to also know you need to have like the domain kind of knowledge expertise like to understand what they actually need. Uh so it can become like repeatable and and quite convenient to use. >> So like I could go do this with Chad GBT right now. I could upload a picture of a shirt and a picture of a model and say put this dress or this shirt or blouse or whatever on this model and put her on a beach. >> But that's one off. but to do that at scale and that's where your agentic stuff comes in and across a whole platform across a whole bunch of different media that's another animal. So that's what you come in and automate. And then it sounds like >> you're kind of like the po.com or the Higsfield po.com for text where they >> you use PO and it goes and uses different models behind the scene depending on the need. Uh and then or Higsfield that uses different video models depending on what you're trying to achieve. So you your agentic system kind of figures out this is the goal. we need to use this model for this, this one for this, and it puts it all together and creates the type of imagery that we see on your website. That's pretty good. Is that basically the general gist? >> Exactly. Yeah. Exactly. You summarize it really well. Yes. Mhm. >> And so so that's that's what makes because I've always been I've always said that there's a lot of people out there that are in the video or image tool space and every time a new LLM, you know, Cloud 4.8 comes out and it's all of a sudden going to have video in it. I mean, no, sorry. Image creation, which it doesn't have right now. If it does, I'm just saying if it does, that puts a lot of like a thousand companies out of business. >> But that's not not what you're doing. What you're doing is the more of the let's do this at scale and and and make this a production machine type of thing, right? >> Exactly. Like think about like there are brands that we work with. I mean like there's so many of them like uh all around the world like they you know like fast fashion retailers or like you know uh they produce like you know hundreds of products like they launch like one of the companies like that we work with they launch 500 products on a weekly basis including you know like women's men's kids homegoods categories and it's not possible to you know like to sit down and then like upload every image of the product being launched like and then you know like add prompts like one by So it's it's becomes like a nightmare and it's actually at that point it's kind of easier to do just a normal photo shoot from scratch uh the physical process basically uh which is yeah which is like a big operational hassle for them already and um we basically automate that in the in the platform as you described and uh this is also just like the starting point like for our uh like company and so we start working with the companies like the brands within that which is creating a lot of costs saving to them like up to like 70 80% and also like speeding up time to market too like instead of like spending weeks like they can launch their products like literally in a couple of minutes to hours. Um and then the next step is also deploying these visuals um via back end to their websites. again like hundreds of products. And then we also run uh sales related like performance analytics on top of these two which is like the ultimate goal is like to create like a more um like a dynamic storefront for you know like either small mediumsiz brands like marketplaces because they actually have a lot of information data you know and we can help them to to leverage to utilize that um in order to create more like personalized experiences on the shopping websites via you know like email marketing things like that that's where we are heading towards. So you could do stuff where the like I know Sheen I think they do some of this. >> They don't even have the product. They just would create an image. If it's a some sort of swimsuit top, you know, and they want to do 50 different colors. Instead of having to manufacture 50 different colors, >> they'll manufacture 50 different designs and colors. Put them all up and see what see what sells. And then they'll one of their factories in China will manufacture that. As soon as the one with pink and blue stripes sells, they make it. Uh versus having to do it in advance. Is is that's your tool be used for that kind of stuff as well or is it used for that kind of stuff? Exactly like in terms of colors, in terms of patterns, in terms of you know like how they present the visuals as well like in terms of models, poses, uh different types of like backgrounds and settings like we can analyze like all these uh and then like basically provide them the best uh like suggestions like how you know like to to maximize like their um like sales performance. That's like yeah what we are building and then testing and they can do a lot of AB testing uh within this kind of a uh platform as well. Um yeah >> do you find it's not all about the clothes necessarily and the accessories that actually sells it but it's also a lot of the in the AB testing it's the model's expression or if her hands on her hip or at her side or if she's facing left or facing right or slightly. Do you find that those little idiosyncrasies can make as big a difference as the color or the look of the outfit? >> Yes, exactly. It's a it's like a it's a complex uh problem and there's like so much details in it. Uh and then that's yeah that's what we are basically working on right now. That's like the more difficult problem but we need to start working with the customers uh to help them automate the existing um process and then on top of that providing them more um refined and more like onbrand um specific solutions like uh improvements in terms of the visuals. But I guess this >> this could probably be on the customer's part as well like just split testing. >> Yeah, for sure. Yeah, I mean it's a self-serve platform like they can generate literally anything they want based on their also you know like um like in size like I mean they know their customers the best. So like they can definitely like create different types of models etc like types of the background settings and then they can definitely test it. And what we are doing is like to to achieve this in a more um automated like more technical way instead of like them testing things out like one by one. uh we will be able to provide a system to them that runs it for like again like hundreds of products uh for them and then it will also become like for example some brands also they come to us they're like oh our customer uh profile is like we we basically want to expand our customer profile into like more gen Z younger generations like how can we attract them uh so and then but they want to keep their current audience too so what we can do is like based on every uh online visitor profile whomever like visits browse the items we can uh identify some of the characteristics and then we can personalize the models for example based on the visitor based on their uh information or so. Uh so yeah that's where we are kind of like going towards in terms of personalization localization as well like a lot of the uh like fast fashion retailers they like do one photo shoot or two and then they show like the same models same images like across all the countries that they sell uh like 40 to 50 countries or so like when we check them but because they don't have time or budget like to bas to to localize those visuals for different uh big markets like let's say for Middle East, for Europe, for Asia like even like sometimes I speak to some um some like teams in Europe they're like oh we need to our focus is in Germany but we need to show different models in Spain like even like in like certain uh areas like it becomes like more differentiated. So yeah, >> I know what you're thinking. I know you were grinning looking at me talking about that thinking of you have a model like me, right? My my generation the fossil the fossil model with a beard. You can say it. >> I mean yeah you can anyone can see any type of model if >> that's that's what a tool that the power of AI that I think a lot of people underestimate is what you just said in in marketing. I think it's one of the most powerful tools. Norman and I have a company called Dragonfish where we're doing a bunch of image and text based stuff and the amount of content that we can spit out at scale and very fast is just ridiculous. And what you said >> with going across, you know, 50 different marketplaces, the models need to look different. Or even if within the same model marketplace, if I'm selling a t-shirt, the black shirt that Norm is wearing, >> I need it on an old guy like him. I need it on some grunge looking hippie. I need it on uh you know a 17-year-old uh runaway looking person. >> Uh I I I need it on whatever. And you can just do that. And so that way all the targeting on the Facebook ads or all the targeting on the website >> is aimed at. So when someone comes there like that's me. They see themselves. They don't see like that's some pretty stick figure model or whatever that they can project themselves into that person that they're seeing that that's someone like me. And you can do that at scale that you could not do in a photo in a in a in a studio. Exactly. Exactly. It's because it's not scalable. But right now there's a solution to be able to do that like to basically create more culturally resonant like assets and then so that it would become more um yeah relevant to the to the audiences and yeah even like in terms of like styling for example for different types of like you know like as you um like do more purchasing like more you know like shopping from a particular brand like they know your taste and so so they can actually like come up with more um like styles com combinations of their like collections like based on what you actually like. So instead of like you try to like piece things together. So it will get there basically. >> Well, if I was a fashion retailer on my personal website, if I was selling uh I would actually once I've identified who you are and I know that avatar, the next time you log into my website, you're not going to see the standard photos. You're going to see I've got 10 different photos of uh the black t-shirt that Norm's wearing. And if I know that you're a a guy like Norm, you're going to see guys like Norm all throughout the website. Um if if you're a young >> That'd be quite the website, Kev. >> That would be uh or I know that Norm loves cigars. You might see a guy No, in a lifestyle picture, smoke a cigar, you know, >> fishing or whatever it may be. Um, so that that's the power of what this what this is. But what do you say to the people though that are like, >> "Well, that's AI. You're tricking us." >> Mhm. I mean, yeah. Like um brands, the companies, they shouldn't trick their customers. Uh like, you know, they need to obviously like disclose it if they use AI. Uh but a lot of the brands, they use it. uh we see it from the biggest like brands like the retailers when they use it for their PDP or so or for campaigns uh like more editorial visuals. Uh but I think yeah the laws regulations should definitely like keep up with the with the advancements happening. So uh there wouldn't be like much of the you know like gray area. So in terms of like you know like everyone would know what they need to comply. Um but yeah so that's like I think like that's happening uh hands in hands you know with these new with the AI generated images a lot of people are trying to take off the watermark or they're just not disclosing and that's first of all Google that's against their terms of service so you got to keep that on there. Secondly, I believe it's the law that you have to tell that it uh somewhere on your website say that the images are AI generated. >> I don't think it's the law in the US. It might be in Europe. But what Google has just done, they just announced uh at their most recent conference is they're actually water they're actually uh watermarking, not watermarking um like >> encoding >> encoding that this is made with AI. So it's a new standard that they're trying to get everybody to adopt that actually you would embed in the images. So it's like it's embedded. You can't take it out. You can't go in Photoshop and like right now on nano on nano banana on Gemini on nano banana you can take out the little diamond that's like the little Walmart. you can uh watermark. You can actually take that out uh in Photoshop or something, but they're doing something where you can't. And so you're going to be able just mouse over something and like, yep, says created with AI and you'll know instantly. That's one of the things that's that's coming out. But I've seen studies where if you say there was a study I had in my newsletter that uh a scientific study where they said if you say this product was designed with AI, it actually hurts the the conversion versus uh Even though that a lot of stuff is designed with AF, if you say, you know, designed with AI, it actually can hurt conversion. So, it's going to be interesting to see where this goes, but I I think it it's widely accepted and you just have a small minority who are a little bit a little bit against it. Um, it's more the you know, some of the creative types like the photographers and models may say this is not this is putting us out of business. Um, but at the same time, you still need the base models. So, you still need, you know, you need the Turkish girl, you need the Spanish girl, you need the German girl, you need some base stuff. Um, and I guess you, the AI could actually create these base images now, too, without you actually having to pose. But, so it's it's an it's an interesting time, uh, where we're going. >> Who owns the copyrights? >> Um, well, whoever generates the image, they own it. They own the copyright of that. But >> so if I'm if I'm generating something on Gemini or Nana Banana >> or whatever platform >> because I' I've heard it I've heard different people say different things that the platform the LLM owns it or no you can just go and publish it and you own the copyright. Well, like I'm not an expert on this, but as far as I know, like um whoever generates the the output uh they own it like uh but I mean of course like as long as it's not someone else's like digital, you know, like identity or like likeness or so uh which is like obviously not uh like yeah uh ethical use but I think like yeah if you just like create you know like an on model image and the model is like fully AI generated uh and then you own the product as well uh as a brand and then that picture belongs to you >> but I mean >> okay >> I think actually in the in the terms of service for like Google and some of the and open AI I know open AAI was this way a couple years ago they technically own the rights to it and they're assigning you a sub if you look read the fine print they're assign assigning you the rights to use it but ultimately they own it if they created it Um, but obviously, you know, likeness, people own their name, image, and likeness. They're NIL, so that there's that's different. You can't make a picture that looks like Brad Pitt and then pass that off as as Brad Pit. But, um, the I think in your case, since your tool is actually a it's it's actually a collage or a a combined creative of multiple tools. It sounds like >> that you're actually creating something original. I'm not like you said, like you said, you're not a legal expert. Neither am I. But I I would I would think that since your tool is creating something that's using different tools to create different aspects and different parts for different reasons. Um it's probably more likely that the legal rights probably reside with the creator uh versus the uh the the tool that's underneath it. But I I don't know that for sure, but that would be my my hunch. But that that's that's always an interesting and thing because you have you know as you probably know if you get create good images you get they get knocked off. I mean norm experienced this in the Amazon world where we we create in the past we would spend5 $10,000 on a nice photo shoot >> and then the next thing we know are images of our model holding the exercise devices all over China uh and all over websites in China and maybe swapped out with a Chinese face >> or something. And so that that's always uh that's always an issue when it comes comes to images and like and it's easy to copy and >> I mean even with Google with their new way with they're going to embed stuff all you got to do to get around that is take a screenshot take a screenshot it creates a new image and so it doesn't carry over. Uh so there's there's always workarounds but what do what so what do you where do you feel this is this image stuff is going? Do you think this is going to get bigger uh and more brands are going to start using this at at scale to just knock out massive stuff or what are you seeing behind the scenes at at at Savannah? Yeah, I mean to be honest like in the next couple of years um or so with this kind of like exponential like growth uh like the the advancements happening um I think this is just going to become like default solution uh like norm and and yeah there won't be like I think like many photo shoots happening or like photoshop or like you know like more manual editing tools or so out there like everything will be like quite you know like a lot of those like fragmented tools uh will become much more like streamlined in one place because it's it's a big like process right now like you know I mean you mentioned that like takes weeks to you know like organize everything and then get the results like do post-processing retouching like a lot of like you know manual steps involved And then but what's happening right now is like all these like steps are being combined into one platform. Um I mean that's what we are doing but it is going to become more common and like kind of the the yeah the just main solution that's how I see uh because yeah I mean it's it's just like it just makes a lot of sense to to use this solution like yeah. Hey, Norm, do you know any sellers out there that are just burned out during this uh ecom game? >> You know, I I know a lot of people that have talked to us, you know, when we go to events, and it's not only that, they don't know where to start. >> And who would you recommend they talk to? >> The first one that comes to mind is is Quiet Light Brokerage. And here's why. They're going to build you up. They're going to understand your company. And at the end of the day, you're going to know how to maximize your valuation. So, the very first thing you need to do is go and get your free confidential evaluation at quietite.com. They're going to ask a couple questions. Uh you're going to meet up. It's one- on-one with uh somebody over there and then, you know, let the games begin. >> Awesome. What was that website again? >> It's quiet.com. >> Awesome. I'm gonna head over there. your platform. First of all, it's affordable. So, you don't have to be a huge brand to be on your platform. And secondly, I don't think we've talked about video, but you also have the ability to produce video. >> Yeah, correct. Yeah, we provide videos too. So, any type of like visuals uh like static images can be transformed into a video. Um, and then we provide templates for that too. And then mostly like in fashion it's been mostly like the model models like turn around themselves or like you know walking towards camera like kind of like standard movements like 5second 10 second or uh it's been very popular for just like uh invisible mannequin of the product and then you know like it does like 360 like kind of turn around it. Um, so I mean even like just you know like creating like flatly or like this ghost mannequin visuals like brands like pay a lot of like money like uh to just like you know get those visuals but right now like in the platform they can do it like a dollar or so and then you know like per per product and then they can get their entire collection um like for for like a very yeah small budgets and it's always like at their convenience like whenever they want to use it, you know, like with just like one click literally they can get it done. Um yeah, so it's also pretty accessible and affordable. U yeah per image is like it's yeah pretty very low pricing. A buck. A buck. And like I'm an old guy, uh, like a fossil. And I remember Kevin, when we had photo shoots, like when when we used to do photo shoots, it could cost thousands of dollars. You know, you'd have to rent half studio for a day or half a day. Um, it was nuts. And now a buck. >> Yeah. Yeah. It's crazy. It's crazy. What? So did you did you raise money for Savannah or is you are you selfunded or did you >> Yeah, we raised we raised some money like preede kind of like Yeah. angel investors are involved as well. Yeah. >> So what's it you I mean you're a woman in a world that's dominated by dudes uh and in the AI world, in the fundraising world. So, what what it's it's refreshing to see a woman that's actually leading a company and leading in this because it's there's not a lot of them. What do you find that that's a difficult or uphill battle sometimes or they're not they don't take you quite as seriously um or that you don't know what you're doing or what or what what's the challenges that you've had to experience as as a woman in a in a heavily maledominated uh realm? Yeah, I mean like in any industry or in any of the jobs that I I worked um yeah women are like yeah it's mostly like male-dominated uh it was the same case for consulting it was the same case uh like for the for the previous startup that I I used to work as well so um I mean like yeah I'm definitely not the stereotype uh like kind of yeah founder in AI but I do have the very relevant experience like coming from the consulting like the kind of more fashion angle and then uh later I worked at this AI startup for about like four years uh from the kind of beginning like from the founding team to the exit um the company got acquired like last year uh by one of the largest media companies so um yeah I mean definitely like I I think a lot of the investors they see you differently. A lot of the big kind of like very famous uh female founders they also explain like their experience kind of like oh like when are you going to get a child like you know like things like that. Are you married? you know like they they may they tend to ask like more personal questions like to see you know like to try to predict the future of the company if you're you know like as motivated uh as like the other you know like male uh peers I guess. Um yeah so it's definitely not easy I would say. Um but like as long as like you show the you know traction and then your relevant experience expertise in this space um and also the vision obviously um it's yeah it that's how you do it. Yeah. >> Are you out out of Silicon Valley? >> Uh we are based in New York in Manhattan. Yeah. And yeah I have a CTO here as well my co-founder. Um yeah, yeah. >> How I mean you you went to Harvard, right? You got your MBA from Harvard. Is that correct? >> Yeah, correct. >> There's some people that say you go to Harvard for the education, but that's a very small portion of people. Most people say you go to Harvard for the network. So, have you have you ex have you do you agree with that statement that the most valuable thing? Yes, Harvard's a good education. I'm not denying that. the most valuable thing is the network that that comes out of going to a school like Harvard or MIT or Stanford or something like that. It would you would you agree with that and how has that helped you in what you're what you're doing? >> Yeah, for sure. I mean it's definitely you know like you get to like connect with a lot of people um in that certain circle and um it's I mean the school itself is like you know it's all about like case study like um like case method and like it's mostly you know like everyone does their own prep and then comes to class and then you know like you discuss the case and then everyone has a different sort of background and experience. Uh so you actually get to learn from each other from the kind of like more academic perspective but also the whole experience is designed to to socialize like to meet as many people as possible and and yeah it's kind of like creating you know a sort of a club in a way that you know like it opens doors to you like you can um reach out to people you know you can there's it's a big community for sure like um Right after MBA, I moved to New York and then you get to still meet a lot of people and um and they are helping there, you know, like uh they want to help you. Uh it's it's definitely it's it's um yeah, it's a big community for sure. >> Where where do you see this going? Like two years ago, I never thought I'd see imagery like I'm seeing now. Where do you see this going? Um I mean yeah for me so basically how I came up with this was um was like at the previous startup we were building generative AI solutions mostly for media. We work with couple of big actors like Tom Hanks uh Kay Riley Robin Wright um the movie was called here and the director was Bob Zamez um the director of Forest Gum. So they wanted to experiment with things uh with AI and that's how like sort of like I got into this like more image video generation like world and I got to you know like experience the impact of it the power of it like instead of like spending years and a lot of manual work like really big budgets we were able to like deliver a lot of the really like high quality deaging stuff like for example Tom Hanks was acting himself as like 67 years old in studio and then on screen he would look like 20 years old for the flashbacks in the movie 30 years old like he was like you know driving all the mimics gestures like it was his performance but then he would just look you know 20 years old 30 years old or so it was real time technology um so that was a big project like I mean we did like half of the movie for that for four actors we made uh their younger versions and that was huge and then we've done like a lot of like similar projects like this too and then I was like this is happening this is a pretty big space uh like producing like hyper realistic um like content and and then after yeah we sold the company I started to look into you know like markets and obviously like ecom, fashion, retail, are huge uh like markets and kind of like known as slow to adapt like slow to kind of like apply new technologies, innovations. Um and then yeah, we've started working with couple of brands literally just like you know reaching out friends in the space. Um and then we've started building solutions for their particular workflows and it's been yeah it's been a really big momentum and every month or so like every couple of you know like like since last year in the last two years the the interest like the you know like the the a lot of the brands they are like really really interested in and they want to actually like integrate to their current um like daily workflows. Yeah. What's the key to getting the quality that you're getting? I mean, anybody can go to a lot of these tools and create images, but the quality like like Norm was saying, looking on your website here is is really good uh with the lighting, the shadows, the the the little details, the fine lines in the hair and where it's integrating into the sunglasses and all that kind of stuff. What is the key? Is it what's the I mean, what's the You don't have to give me your secret sauce, but what's the secret sauce? Is it in the prompting? Is it in the knowing all the technical side like you coming from this movie world where you had to create stuff that looked really good on a big screen or what what's the what's the key to actually making this making your stuff differentiated? I think like yeah it's creating like particular um like pipelines like solutions for each of the use case like as I mentioned like you know creating on model image is one challenge and then it's yeah it has a lot of like technical details in it like like you know the skin texture of it like how the you know like how you create an AI model how you render that item on the model, how you create the background, how you combine all these pieces together, how you also um kind of blend the companies, the like the brands uh like kind of the visual aesthetic into it. So, it's actually like it's it's not easy. It's um you need to like yeah work on it and then yeah, it's basically finding all these like challenges um difficulties like logo refinement is another challenge for example. uh we basically like find all these like issues working with our customers like very closely and then uh piece by piece like we basically it obviously like what what the platform today is very different than what it was last year. So we are like shipping features like literally on a daily basis. Um it's yeah it's all about being very uh on top of what's happening um in the industry and then also what's the actual need for the customer and then tying them together and it doesn't happen easily. Uh yeah, it's a lot of technical requirements like we have very strong um engineering team. Um yeah. >> So what happens with fonts? I know with uh some uh LLMs that I'm working with and other apps that sometimes font either on the shirt or on the logo will fall apart. >> Now how does that how does font affect uh images on your site? uh like the the font on the logo. >> Yeah, if I were to Well, it doesn't even have to be on a logo. It could just be a saying. Um just just uh give you just an example. Uh coffee cup. There we go. Says Hawaii. Okay. And if I were to load uh upload this image, sometimes it would look like I don't know, it it just running into each other or it would break apart. Um, and that happens a lot still today if you put in a com like I I will deal with a lot of non alcoholic spirits and they're very very particular >> and they're very the font is very fine and nine out of 10 times I got to go back and redo it and redo it. >> Now do would I have to worry about that using your app? >> Um, we yeah we do have a particular solution for that. uh like most of the times as long as the input image uh reflects the font you know like uh in a good light or so uh we can yeah create that uh pretty accurately. Uh yeah it's it works pretty well. >> Yeah, you might have a new client. >> Yeah. Okay. We can do a live demo with your >> I know I told you you got to quit doing these pictures of you and Speedos. >> You got to quit that because that's just grossing the people out. I know you want to do one in a pink one and a blue one and a green one, but you you can't with with a little uh bracelet on your hand, but you you can't do that, man. You can't do that. >> It's better than you in the thong. >> And we'll stop it there. >> Yeah, you don't want to paint that picture. That's That's for sure. So, so I mean, so you said furniture and fashion are the two big things, but when we we were talking earlier, you said that you're actually reaching out to some like people in the Amazon e-commerce world and trying to get some of those guys that are that first. What So, is it just the fashion and furniture type of stuff there? Or you can pretty much do any kind of lifestyle pictures or models holding something or using something or a girl holding a power saw drilling into the wood or or what what what's your what's your targets or goals there? Yeah, actually like we um so we started like just working with fashion companies. Uh but then we were like checking like who's like signing up or so like who's like using the platform. We actually started to notice like a lot of different uh like product groups uh are coming into the platform like the companies. Um it's literally it is any anything any product category like red light therapy mats to like you know like the masks to um like yeah a lot of like homegoods stuff. Uh it is like yeah it can be applied to really anything. Um but we just like we provide very much more like automated solutions for particular groups but any uh product group can be also yeah generated in the system. Yeah. So, if I come to you and I I want to run Facebook ads in 40 different countries >> and I want I have my red light therapy mask and I want to run in 40 different companies countries, but I want the ads in Turkey to be look look a to look Turkish, like a Turkish woman. I want the ads in Lebanon to look like a a a Lebanese woman. I want the ads in Germany to look like a German woman. I want the ads in uh um Colombia to look like a Colombian. I can do I need to have 50 different models >> from each of those countries or you have all those models and those ethnicities I could just pick uh and then I can just say okay here's my mask here's a product shot of my mask here's the 50 countries or the 50 miles I go select them and then just hit a button and then boom it's all and I did set the dimensions and the size and the back whatever the other variables are >> and then a couple minutes later I h I have those entire things to download is that basically and that cost me 50 bucks You said a dollar an image or something or whatever it is. >> Yeah, pretty much. Yeah, exactly. So, we do provide like a diverse model library that you can select from like female, male models, kids models as well. Uh so, you can select any of the you know like size, age, ethnicity, any of the you know like characteristics that you would like to create. Uh just like yeah one by one you select them and you can select like different poses too uh from the templates. Uh and then yeah, you get the results like literally in in like maybe for the 50 of them it would take maybe like 20 minutes or so max. Yeah. >> Are these models AI generated models that were generated from scratch by AI or are they real people that you you paid and l they licensed the way they're imaged to be used for this? >> Yeah. So right now they're all fully AI generated. Uh but we are we are planning to move uh to that direction as well. Like for example working with model agencies like we can make agreements with them. Uh so if you know like any model would like to you know create their own digital like you know uh version uh like avatar or so uh we can you know like include them and then every time they are you know like like image is being used they basically make money out of it. Uh so we will uh we are like in the process of like making that type of partnerships too and also like for example if brands like you know or let's say like if you want to use a particular brand for your own uh visuals you can also actually upload that image obviously by making your own agreement with that model. Um and then you can technically create that particular model images too. But yeah, so far in the platform like all the models in our library are all AI generated. >> That could be a huge market with creators. If I'm a big creator and I got 10 million followers and I want to do I want to license my image out to all these brands, but I don't want to go stand and do photo shoots and all over the all over the world. but I can authorize them, you know, to actually use my name, image, and likeness um for a feed. That could be a huge business. >> And then other brands, if I'm if I'm a Nike and I and I want to work with uh this one creator, I don't have to go to the creator and say, "Hey, can I um I could actually go to a middle person like yourself. You're an agency and you have agreements and like, yep, for you can use Susie Smith for $10,000 for this ad campaign if you want and it's all right there." that could that could be a huge opportunity on both sides to actually leverage uh authority brand and everything all into one. Is that something that you guys are playing with? >> Yeah, definitely. We are um making some plans like to get there and we're already in some conversations with some of the agencies and then uh and there's actually a lot of demand on that side too from from them like from the models uh or so too because and some of them actually they want to create their younger versions uh >> of course >> yes >> that's your trick norm when you do your speedos just the younger the younger norm when you back when you were wrestling They're looking good. >> Yeah. So, there are some there are some like interesting use cases uh which is like which is happening. Yeah, for sure. >> And Zar, you're not supposed to agree with him. >> Oops. Why not? Yes. >> Definitely an interesting time to be living and to be in. And there's tons of opportunities out there when it comes to >> Yeah. AI and images and video and and you you name it. But uh and images are are one of the things I think are kind of leading the way um in a lot a lot of what's what's happening. >> But is there anything that you see that AI is still struggling with like on on the creative side? Mhm. Um I mean yeah like for sure um like capturing you know like 100% accuracy is like still a challenge uh when it comes to more like you know difficult items like when it comes to more like intricate details or so um so because you know it's all about um the system like predicting like a lot of the things and then when it comes to more like complex patterns or you know like um textures or so like it's like still still like challenging like to to create the 100% accuracy on those. Um and then logo integration is like something that we do uh specifically like um on on on the you know like t-shirts or like dresses or things like that like there's a company logo and then it's been also like it's it's still a big challenge like to because you know like they have very particular requirements like in terms of the scale in terms of the the the details of it. So we are working on that too like we are providing a like a like an editing function to to apply to apply their particular logo. Um yeah and underwear is like it's there is a big demand coming from uh underwear companies for example and it's a challenge like to create the you know like the details around them as well. Um yeah but but it's it's covering yeah most of the main use cases right now like creating for example yeah diverse models like different sizes uh it's it's been very helpful for a lot of brands like to yeah to become like to be like more inclusive with their visuals for example. Yeah. >> Hey Kevin King and Norm Ferrar here. If you've been enjoying this episode of Marketing Misfits, thanks for listening this far. Continue listening. We got some more valuable stuff coming up. Be sure to hit that subscribe button if you're listening to this on your favorite podcast player. Or if you're watching this on YouTube or Spotify, make sure you subscribe to our channel because you don't want to miss a single episode of The Marketing Misfits. Have you subscribed yet, Norm? >> Well, this is an old guy alert. Should I subscribe to my own podcast? >> Yeah, but what if you forget to show up one time and it's just me on here? You're not going to know what I say. >> I'll I'll buy you a beard and you can sit in my chair, too. We'll just You can go back and forth with one another. Yikes. But that being said, don't forget to subscribe, share it. Oh, and if you really like this content, somewhere up there there's a banner. Click on it and you'll go to another episode of The Marketing Misfits. Make sure you don't miss a single episode because you don't want to be like Norm. >> Very good. Well, it's looking like it's getting to the top of the hour. Kev, do you have any uh other questions? >> No, just uh if people want to try the tool, they can try it for free, right? There's like you get a I saw there's some free credits if someone just wants to go to Savannah savv.ai AI and you can give it a a free whirl, right? >> Yeah, exactly. They can go to our uh website. Yes. Savana AI and then they can click on try for free and then they will be directed to the uh to the app like or either on desktop or mobile. Um they can create an account either with their Google or non- Googlele accounts and then we provide couple of free credits for them like to test it. Um and then if they like it they can yeah sign up they can get one of the plans to start with. It's also like just they can start monthly see how it works for them. Uh we also they can also reach out to us and then schedule a live demo. We usually help them like uh before and after as well like we do um on boarding. We help them like to we show basically how the the particular solutions work and yeah it takes like 10 minutes or so to get you know more familiar with the solutions and it's pretty fun after that actually like they can you know like we get this we get this feedback which is like which makes me very happy. They say that it's a very fun platform to use. It's pretty easy to use. So it's yeah that's that's the goal. Um a lot of people come to us, they're like, "Oh, it's like so exhausting like to write prompts and I don't even know where to start to, you know, like create something good." So that's what we provide to them like the starting point like a base. Uh in terms of design too, uh they can they've been a lot of the brands, the teams, they've been using it for design related solutions as well. Um so yeah, >> very good. Well, at the top of the hour, and it is exactly the top of the hour, uh we always ask our misfit if they know a misfit. >> Um like someone else like to >> to be put in put on the spot just like you. >> Okay. Um yeah, so actually a friend of mine uh she was on your podcast, Shan Shan, uh she I think like Yeah. Uh recommended me to you. Um I would say um a friend of mine Silvin Goa they have a very cool uh startup as well. Um they do more like UGC type of videos uh for any type of company and yeah their startup is super cool too. I can I can yeah uh recommend her. >> Very good. Oh we'll be reaching out to get that information. All right. Well thanks so much. This has been awesome. Learned a ton. Learned a ton about your company and it looks like it's just going to be a real success. >> Thank you so much. Yeah, it's been a pleasure uh speaking with you. Thanks so much for having me. >> Thank you for coming on. Appreciate it. >> Yeah. >> All right. Now, I'm going to remove you, I think, if I remember how to do this. There we go. All right, Kev, that was that was pretty good. Yeah, there's so much going on in the in the world of AI and imagery right now. It's hard to keep up with it all. It's like it seems like every day there's a new thing, but I think what she's doing is smart and future proofing where it's putting the agentic side in where you can do this stuff at scale. Uh and that that's, you know, something that you and I are working on. >> We just talked about it before we got on the podcast today. >> Exactly. And so, uh that's that's uh that's perfect. Uh, and I think I think that she's on to something there. And this is just a I think a good example of the proper use of AI. Um, and so so for those that are listening that are anti- AI, this is the right way to use AI right here. Uh, and to to scale. And it's it's the tools and the processes. It's not so much the slop and the crap that's out there. And that's that's what's uh sets everything apart. So if you haven't, go to savannah dot I'm saying Savannah. I think she's pronounced it Savannah. savana.ai and check check it out. And speaking of that, there's something else they should check out. What is that, Norm? They should check out something on uh the YouTubes or the Tik Toks or something. >> I think it's called that. Yeah. Or the Googles. Uh no, definitely check this out. Go to marketingmisfits.co, not.com. And you can always go to YouTube, check out our channel, and that is MarketingMisfits Podcast. Or if you just want to check out these three minute or less uh clips, go to marketingmisfits clips on YouTube. Or you can also read the the podcast. Uh we have a newsletter called Misfits Marketing Misfits. It's a misfits.news. So misfits with an s.news. And you can sign up every Wednesday. A brand new edition comes out. Features the highlights from uh this podcast as well as some of the past podcasts. And uh we're getting really good feedback on people that that like to read that and just skim over it and maybe you don't have an hour to listen to every episode, but you can actually check out uh the latest of what's going on in the marketing world and get some ideas there. So misfits.news >> and we're a little biased. >> Yeah, a little bit a little bit. >> We think it's good. >> We're getting we're getting some good feedback from people. Um, but yeah, so uh if you want more feed fe feedback or feedback from us or learn from us, we're here every Tuesday with a brand new episode. So I guess I'll see you again uh next Tuesday, Norm. >> Very good. All right, everybody. We'll see you later. >> Take care. >> Bye.