{ "version": "1.0.0", "segments": [ { "speaker": "Speaker 1", "startTime": "0.0", "endTime": "19.904999", "body": "And then on the e commerce side, we're actually... We've measured a 15% higher conversion on e commerce PDP imagery compared to traditional photo shoots. And they're 80% to 90% cheaper. I think to your point earlier, that wasn't achievable two years ago. Like, there's no way." }, { "speaker": "Speaker 1", "startTime": "20.945", "endTime": "46.905", "body": "It's like, you don't have to have seven rounds of samples. You can just do all the six rounds digitally, and then you just get one final sample. You could imagine in another year that could be 25% higher conversion, because you're able to show videos, like very high quality videos, and like personalized content as well, which as we know, like, still is a completely greenfield in the space." }, { "speaker": "Speaker 2", "startTime": "48.344997", "endTime": "82.015", "body": "Welcome to the Retailgentic podcast where we explore the intersection of retail, ecommerce, and AI agents, which we call Retailgentic. Our mission is simple. We are shining a bright light on the new agentic AI shopping megatrend which will disrupt the trillion dollar digital retail market. The stakes are high. Google, OpenAI, Microsoft, Perplexity, Amazon, Apple, Meta, and Anthropic are all fighting to be the Consumer AI Shopping agent winner." }, { "speaker": "Speaker 2", "startTime": "82.095", "endTime": "98.99", "body": "There are gonna be some tough decisions for online brands and retailers to make. What happens to checkout? How should payments work? Could you continue to invest in retail media networks? What does the world look like when over half of your traffic is not from humans but from AI agents?" }, { "speaker": "Speaker 2", "startTime": "99.63", "endTime": "107.775", "body": "These existential questions are the tip of the iceberg. On this podcast, we're talking to industry experts, analysts, and influencers to get answers to" }, { "speaker": "Speaker 3", "startTime": "107.775", "endTime": "107.935005", "body": "these" }, { "speaker": "Speaker 2", "startTime": "107.935005", "endTime": "123.15", "body": "questions, identify trends, and best practices. This is Retailgentic, your podcast partner in navigating our AI shopping agent future. And now here's your host, Wingo." }, { "speaker": "Speaker 3", "startTime": "124.11", "endTime": "160.88501", "body": "Welcome to this episode of Retail Gentic. My guest today is Cheryl Liu, Founder and CEO of Raspberry dot ai. About every six months on the show, I like to mix it up from our normal front end demand gen, discovery, agentic commerce world, and look at how AI is being used in unique parts of the stack. We've looked at AI is being used in store. And on this episode, we have Raspberry AI, which uses AI in the product creation part of the journey, largely in the fashion category, but some others as well." }, { "speaker": "Speaker 3", "startTime": "161.205", "endTime": "198.82999", "body": "What they do is they help the designer visualize the product, show it to potential users in a very realistic kind of a way, and basically get product market fit very quickly, get feedback very quickly on design ideas. It's a really fascinating discussion how Cheryl came up with this idea, how they've implemented it, and some of the ramifications it has for product creation and also all the way into discovery. So there's an unintended consequence here that if you can make products very quickly, the discovery can feed into that part of the mechanism. We discussed that. Hope you enjoy this." }, { "speaker": "Speaker 3", "startTime": "198.82999", "endTime": "221.595", "body": "I learned a ton about this part of the way products get designed. Cheryl is a fascinating entrepreneur and I think you'll enjoy this one. Welcome to this episode of Retailgentic. I'm your host, Scot Wingo. Today, I'm excited because we're going to take a little bit of a diversion from our normal talking about what's going on in the world of the front end of a Gentic commerce." }, { "speaker": "Speaker 3", "startTime": "221.595", "endTime": "228.81", "body": "To help us on that journey, we have Cheryl Liu. She is the Founder and CEO of Raspberry AI. Cheryl, welcome to the show." }, { "speaker": "Speaker 1", "startTime": "229.37", "endTime": "231.85", "body": "Thanks for having me, Scot. Great to be here." }, { "speaker": "Speaker 3", "startTime": "232.73", "endTime": "244.815", "body": "Yeah. Before we jump into what the heck Raspberry AI does, let's talk a little bit about your background. How did you decide to do this crazy thing of founding a company?" }, { "speaker": "Speaker 1", "startTime": "245.295", "endTime": "272.04498", "body": "My background kind of straddles both retail and technology. So I started my career actually very focused on retail. I was at a private equity firm called KKR, where I was investing in consumer retail companies. Think about all the big mall based chains. And at that time, we were helping all these brands go online, go omnichannel, because that was the big revolution at the time." }, { "speaker": "Speaker 1", "startTime": "273.72498", "endTime": "298.605", "body": "And that's really where I learned of how the sausage or the products were made. And that's what eventually inspired me to start Raspberry, is to really change and modernize the product creation process. But before I started Raspberry, after investing in these retail companies, I worked in the tech industry. So I was at Amazon. I was at DoorDash." }, { "speaker": "Speaker 1", "startTime": "298.605", "endTime": "317.14", "body": "I was at a Series D e commerce startup, and mostly focused on bringing machine learning into those companies to really transform decision making and operating at scale, as well as bringing retail categories onto those platforms." }, { "speaker": "Speaker 3", "startTime": "317.22", "endTime": "326.82", "body": "At Amazon, were you more on the AWS side, the... What they call the store side now? Delivery? Which part of the giant empire of Amazon did you touch?" }, { "speaker": "Speaker 1", "startTime": "327.085", "endTime": "355.71002", "body": "I was in the AWS division, and specifically, was working on machine learning instances. So this is way before now where it's like the demand for these instances is way too high, so high. But, no. At the time, like, we were helping a lot of, like, self driving car companies and, like, big tech companies, get resources to to train their models better." }, { "speaker": "Speaker 3", "startTime": "356.54498", "endTime": "358.625", "body": "Yep. They still needed GPUs. We didn't realize that." }, { "speaker": "Speaker 1", "startTime": "358.625", "endTime": "359.905", "body": "They still needed GPUs. Such" }, { "speaker": "Speaker 3", "startTime": "359.905", "endTime": "362.465", "body": "a big thing now. If only you had just taken all your money and put it" }, { "speaker": "Speaker 1", "startTime": "362.465", "endTime": "364.465", "body": "in know. Only yeah, exactly." }, { "speaker": "Speaker 3", "startTime": "364.465", "endTime": "380.72998", "body": "Hindsight, is twenty twenty. It's sad. Has it been weird being a machine learning practitioner to see this happen? It's gonna be kinda surreal, like, the the the curve that that... It was, like, so slow going, and then it just went, you know, exponential." }, { "speaker": "Speaker 3", "startTime": "380.72998", "endTime": "381.53", "body": "It's crazy." }, { "speaker": "Speaker 1", "startTime": "382.49", "endTime": "407.19", "body": "Yeah. I mean, I think it's it's quite insane. I think especially on the consumer end of things because, you know, back then, we were, you know, working with, like, really, really large multibillion dollar tech companies. And now, like, people are running those things on their their Mac mini, you know, or, like, there's new technologies that allow you to run GPUs on, your phone. So it's it's insane." }, { "speaker": "Speaker 3", "startTime": "407.19", "endTime": "416.015", "body": "Yeah. Who who would have ever guessed that? Yeah. Tell us about the the the the little origin story of Raspberry. So you were sitting there." }, { "speaker": "Speaker 3", "startTime": "416.015", "endTime": "421.855", "body": "Was this after you went to that later stage company and decided to do your own thing?" }, { "speaker": "Speaker 1", "startTime": "423.135", "endTime": "428.99002", "body": "Yeah. Yeah. So I was actually at DoorDash at the time." }, { "speaker": "Speaker 3", "startTime": "429.07", "endTime": "430.11002", "body": "Oh, DoorDash. Yeah." }, { "speaker": "Speaker 1", "startTime": "430.11002", "endTime": "438.51", "body": "So we were our mandate there was to deliver any kind of product within thirty minutes." }, { "speaker": "Speaker 3", "startTime": "441.155", "endTime": "442.915", "body": "Hot and fresh as if it was Hot" }, { "speaker": "Speaker 1", "startTime": "442.915", "endTime": "443.55502", "body": "and fresh." }, { "speaker": "Speaker 3", "startTime": "443.95502", "endTime": "445.635", "body": "Than you can avail it yourself. Yes. I've heard" }, { "speaker": "Speaker 1", "startTime": "445.795", "endTime": "446.035", "body": "Yeah." }, { "speaker": "Speaker 3", "startTime": "446.035", "endTime": "450.355", "body": "Tony I've heard the CEO does a very good he's very passionate about that." }, { "speaker": "Speaker 1", "startTime": "451.28", "endTime": "467.86502", "body": "He is. Yeah. And we did spend a lot of time, on the operational piece. I remember at the time, like, people were talking about, like, oh, do you really need your Levi's jeans? Or, like, do you really need, like, I don't know, like, something from Sephora in, like, thirty minutes." }, { "speaker": "Speaker 1", "startTime": "468.025", "endTime": "477.86502", "body": "And, I don't know. I think, like, at this... Like, we just... We're so lucky today as consumers. Like, everything's moving so quickly." }, { "speaker": "Speaker 1", "startTime": "477.86502", "endTime": "501.76498", "body": "Right? So... We're so used to getting like an Uber in two minutes, and like our hot food in thirty minutes, and like Amazon same day packages. So anyways, I was I was there, and I was exposed to some of the first image generation models. And at the time, obviously, they were really, really poor quality." }, { "speaker": "Speaker 1", "startTime": "501.76498", "endTime": "541.70996", "body": "They were generating seven fingers. And they were the faces were just really, really distorted. It didn't listen to you most of the time, whatever you prompted. And construction was really bad. But I just felt like you kind of could see the future if this technology would become really, really good, that a lot of things that are tied to the physical world in the product creation process and in the e commerce process and in the marketing process, like, everything you have to wait for the physical sample to arrive." }, { "speaker": "Speaker 1", "startTime": "541.70996", "endTime": "571.0", "body": "And if that doesn't arrive, your photo shoot's delayed. And if the sample is slightly different than the latest version, that means you're shooting the wrong thing for your e commerce site. Or sometimes you just don't get the shots you want, and then you have to do a whole reshoot. And so, yeah, I just could imagine a world where none you didn't have to do any of that. And you wouldn't have to deal with all of the constraints of being in a physical world." }, { "speaker": "Speaker 1", "startTime": "571.0", "endTime": "604.565", "body": "And that you could really kind of create physical products similar to how we create software products. And yeah, that was really what inspired me to take the leap. I'd been thinking about this problem of how physical products were created for so long, being having invested in those companies. It usually takes like twelve to eighteen months for these fashion companies from the designer's idea to when we as consumers can buy that product. And that's just real..." }, { "speaker": "Speaker 1", "startTime": "604.565", "endTime": "651.93", "body": "Way too long in this modern day and age. And these brands are really feeling it, both from a sustainability perspective, but also from a operating margin perspective. The fastest brands in the industry, these huge conglomerates that have been built up over the last decade, they're able to take that twelve to eighteen month calendar and condense that down to three to fourteen days. And my PE days, we saw what happens when you can't compete effectively in the global market. And so that was really the inspiration is, okay, let's use this generative AI imagery technology to bring a lot of this physical process to the digital world." }, { "speaker": "Speaker 1", "startTime": "653.97504", "endTime": "667.09503", "body": "And, you know, really, of going through seven rounds of physical samples, you can experiment with every single type of prompt material color you want in a matter of minutes, and you don't have to wait for the sample to come back from overseas." }, { "speaker": "Speaker 3", "startTime": "667.67", "endTime": "687.08496", "body": "I'm a big Amazon nerd. I started a company that was in their marketplace, and I always thought it was really interesting how they operate as a company. One of their management principles is kind of the start. You know, when before they develop a product, they started the press release. So it's kind of like you you you get alignment on the finished product, then you're build it, which is, you know, we don't do that in the physical world." }, { "speaker": "Speaker 3", "startTime": "687.08496", "endTime": "700.23", "body": "So it's kinda like you've you've applied that Amazon principle to this. It's kinda like, let's at least get a picture of what we wanna build and, like, suss all that out. You can't touch it and feel it, but maybe you can kind of get close. You can, like, get a mood board and" }, { "speaker": "Speaker 1", "startTime": "700.39", "endTime": "700.71", "body": "Yeah." }, { "speaker": "Speaker 3", "startTime": "700.79004", "endTime": "713.835", "body": "You know, little pieces of fabric from the thing you're gonna do and get, like, a lot closer, before... Deep deep, you know, pre answer a lot of the questions you wanna answer before you go build it. Yeah. I I... That's very cool." }, { "speaker": "Speaker 3", "startTime": "713.835", "endTime": "716.955", "body": "I'm a software guy, so everything has to have a software analogy." }, { "speaker": "Speaker 1", "startTime": "717.995", "endTime": "731.64996", "body": "I wasn't even thinking about that press release at Amazon, but that makes so much sense. The... Actually, the the the inspiration here was... Is that some of these, like, really fast companies I mentioned, that that is their business model. Right?" }, { "speaker": "Speaker 1", "startTime": "731.64996", "endTime": "733.08997", "body": "That Yeah. We're able to have" }, { "speaker": "Speaker 3", "startTime": "733.17", "endTime": "746.08496", "body": "You're specifically talking about she... H and m kinda started fast fashion, but I think they were, like, six to eight months. And then Sheehan's, like, hold my beer, now they're just like, boom. They see a trending TikTok thing, and they can have something on the site. Yeah." }, { "speaker": "Speaker 3", "startTime": "746.08496", "endTime": "753.08", "body": "I almost think they just put it on the site and then make it base... It's almost like a Kickstarter at that point. I don't Oh, yeah. Exact sequence of things, how things" }, { "speaker": "Speaker 1", "startTime": "753.08", "endTime": "754.36005", "body": "happen. But It 100%" }, { "speaker": "Speaker 3", "startTime": "754.52", "endTime": "756.52", "body": "like it's almost totally inverted. Yeah." }, { "speaker": "Speaker 1", "startTime": "756.52", "endTime": "762.92004", "body": "It's a 100% is. It's exactly what you said with... About the Amazon press release. It's... They're putting the product online." }, { "speaker": "Speaker 1", "startTime": "762.92004", "endTime": "790.96", "body": "They're testing it, seeing what kind of demand there is. They have a lot of machine learning going into like the, how people are clicking, how people are viewing, how people are buying. And then they make the buys that way, versus the traditional way of like, you just guess what people wanna buy eighteen months from now. And now that's like basically impossible because of the way social media is transformed, of like how trends... Trends like, you know, some of them are still coming from runway, but a lot of them..." }, { "speaker": "Speaker 1", "startTime": "790.96", "endTime": "797.755", "body": "A lot of the big trends are just coming from, like, social media and, like, grassroots. So..." }, { "speaker": "Speaker 3", "startTime": "798.555", "endTime": "800.395", "body": "Yeah. Cheryl and Hailey Bieber, according to" }, { "speaker": "Speaker 1", "startTime": "800.395", "endTime": "813.05005", "body": "my daughter. Exactly. Exactly. So it is really important now to to be able to test and and align on the final before before all the efforts and cost of creating it." }, { "speaker": "Speaker 3", "startTime": "813.37", "endTime": "821.29004", "body": "Did did you always wanna be a founder, or you just kind of couldn't get the idea out of your head and said, darn it. I'm gonna have to go build this thing." }, { "speaker": "Speaker 1", "startTime": "821.945", "endTime": "831.945", "body": "It's a really good question. I actually... I I wanted to start, like, a Shopify two point o. I was always, like, very fascinated. That's why I love your podcast." }, { "speaker": "Speaker 1", "startTime": "831.945", "endTime": "849.74506", "body": "I was always very fascinated by, how people shop and how that... That's changing and all the different modalities. But really, I just felt... You know, I did a lot of exploration. I talked to a lot of Shopify sellers and a lot of brands." }, { "speaker": "Speaker 1", "startTime": "849.825", "endTime": "869.64996", "body": "And I asked them all, what would it take to three x your business? Like, how do you think you could three x your business? And it was never around the shopping element, because Shopify solved a lot of those problems. It was always around product creation. It was like, I can't create product fast enough." }, { "speaker": "Speaker 1", "startTime": "869.64996", "endTime": "889.385", "body": "I have no means of testing my product. I have no idea what people are gonna wanna buy twelve months from now. No data source can like tell me with any sort of accuracy what people will wanna buy. So it was really, it's like photo shoots are so expensive. Like marketing campaigns are so expensive." }, { "speaker": "Speaker 1", "startTime": "889.385", "endTime": "908.035", "body": "Like it was really around, like, getting these products created and to market, which seemed like the big problem. So that's why I felt compelled to solve it. And I think we're still very early days. Like, we have been selling the product for two year... Two and a half years." }, { "speaker": "Speaker 1", "startTime": "908.515", "endTime": "913.715", "body": "We're working now with four sixty brands. But, yeah, there's a lot more to go." }, { "speaker": "Speaker 3", "startTime": "914.035", "endTime": "921.32", "body": "That's amazing. That's quite a two and a half years. So that means you started in early 'twenty four, right?" }, { "speaker": "Speaker 1", "startTime": "922.04004", "endTime": "922.84", "body": "Started we" }, { "speaker": "Speaker 3", "startTime": "923.08", "endTime": "925.72003", "body": "launched The Image Generation was not good in February of" }, { "speaker": "Speaker 1", "startTime": "925.72003", "endTime": "926.60004", "body": "'twenty four. I" }, { "speaker": "Speaker 3", "startTime": "927.0", "endTime": "932.365", "body": "had Mid Journey, a Dolly. It was it was a stable diffusion. Right? Is that" }, { "speaker": "Speaker 1", "startTime": "932.52496", "endTime": "934.845", "body": "Yeah. That was the big one at the time. Yeah." }, { "speaker": "Speaker 3", "startTime": "935.005", "endTime": "946.6", "body": "But it was still magic. You were like, you know, I just you know, I bought I can do stick figures. But I've always had this vision of better images than than I was still human. And then they translate it. And then we're like going circles a little bit." }, { "speaker": "Speaker 3", "startTime": "947.48", "endTime": "961.565", "body": "And yeah, so you took a very substantial bet that that would get better. I can definitely say that. And it did. The good news is it probably got better faster than you even would have ever guessed, I would imagine." }, { "speaker": "Speaker 1", "startTime": "961.565", "endTime": "966.68506", "body": "And that's I was just going to say it got better faster than I expected." }, { "speaker": "Speaker 3", "startTime": "967.16504", "endTime": "986.565", "body": "You seem to be a very humble person, but I'll brag on you. The I was looking I always go and look at fundraising history because that's a good indication of one of the things to look at how well you've done. And your seed round, you had kind of a who's who of Silicon Valley in there. But let's start there. How did that come together?" }, { "speaker": "Speaker 3", "startTime": "986.565", "endTime": "1002.94", "body": "Did you just roll right in and pitch and they all said, sign me up? Because I imagine VCs, they totally get the commerce side of things. But I bet on the other side, they probably haven't thought much about the product creation side. Did you have to kinda educate them on that?" }, { "speaker": "Speaker 1", "startTime": "1002.94", "endTime": "1007.74", "body": "Right on. Wow. Yeah. I wish I'd talked to you before fundraising myself." }, { "speaker": "Speaker 3", "startTime": "1007.89996", "endTime": "1014.91504", "body": "Well, I don't know how to solve it. You clearly have, Sue. I... I'm just... I can guess you had to kinda bring them along and help them understand that piece." }, { "speaker": "Speaker 1", "startTime": "1015.235", "endTime": "1042.7449", "body": "I think you're very spot on. Like, when... You know, initially, there's a lot of education in the process. And I think that's always the case if you're working on something truly transformational, because it's just a very different way of thinking of doing. And it's like, what we're trying to do at Raspberry is is kind of insane." }, { "speaker": "Speaker 1", "startTime": "1042.905", "endTime": "1065.2", "body": "Yeah. To get to the final PDP imagery, to get to the final marketing campaigns, to have designers who had that initial idea actually create billboards. At some of our clients, they do that. Like, two years ago, like, the marketing department would be like, that's crazy. Why would I have a designer create a billboard?" }, { "speaker": "Speaker 1", "startTime": "1065.2", "endTime": "1109.465", "body": "Like, that's blasphemy. And so, yeah, I think it was similar with VCs. I think we are very lucky in that through kind of like the education system and and, you know, being at Stanford, we had a lot of connectivity to venture capitalists and, you know, also having been part of really great technology companies like DoorDash and Amazon, I think we had just a good sense of what it takes to to build an amazing technology company. And then I always think there has to be some connectivity. Like, at the end of the day, I think a venture..." }, { "speaker": "Speaker 1", "startTime": "1109.865", "endTime": "1132.925", "body": "Getting investment from venture funds is education, but it's also like they have some predetermined predetermined biases as well. Like, some of our seed investors, like, they... One of their daughters was a fashion designer, and then another one You wanna jackpot on that one? Yeah. Yeah." }, { "speaker": "Speaker 1", "startTime": "1133.245", "endTime": "1153.5599", "body": "Exactly. And then the other one, she had... She also had a major consumer fund. So she was investing in a lot of brands. So she she could see firsthand from the brands that she invested in, like, how much money they're spending on sampling and how much time and how that's always delayed and how much time how much time and money they're spending on photoshoots." }, { "speaker": "Speaker 1", "startTime": "1153.88", "endTime": "1180.7399", "body": "And then for the series a, which was a $23,000,000 round led by Andreessen Horowitz, again, we one of the deal team members, his whole family was in fashion. And we demoed to, I think, one of his sisters, and she was like blown away. She was like, I've never seen anything be able to do this. And this would have like... She had at that time like exited the industry." }, { "speaker": "Speaker 1", "startTime": "1180.98", "endTime": "1185.3799", "body": "She was so fed up with it. But she was like, yeah, if I had this, like, I wouldn't have left." }, { "speaker": "Speaker 3", "startTime": "1185.62", "endTime": "1203.0", "body": "Doesn't get better than that. We know Alex Rampell really well. I've seen Brian Kim do a lot of he's on the consumer side, right? I consume a lot of their content. They do that monthly or quarterly consumer AI kind of chat." }, { "speaker": "Speaker 3", "startTime": "1203.16", "endTime": "1204.76", "body": "Yeah, it's a great team over there." }, { "speaker": "Speaker 1", "startTime": "1205.16", "endTime": "1207.3201", "body": "Oh, yeah. Alex is great. We love him." }, { "speaker": "Speaker 3", "startTime": "1207.3201", "endTime": "1221.6149", "body": "Yeah. And Brian. An e commerce guru. He can go he talks I have to always have a venti coffee before I talk to him because he talks insanely fast. His brain operates even 10 times faster than his voice can carry it." }, { "speaker": "Speaker 1", "startTime": "1221.615", "endTime": "1226.095", "body": "Yes. The team there is very intelligent. Yes." }, { "speaker": "Speaker 3", "startTime": "1226.4149", "endTime": "1236.75", "body": "Okay. So you get this fundraising. And then how hard was it to kind of walk through the journey of landing your first customers? And then we'll jump into some case studies of the cool things going on now." }, { "speaker": "Speaker 1", "startTime": "1236.91", "endTime": "1265.4", "body": "We got our first customers through a combination of investor intros and LinkedIn outreach and going to conferences. I would say some of the bigger ones, we probably got through conferences, just being face to face. But, yeah, I mean, we definitely... I think it's... Even in the early days, it's, like, always been quite a competitive space." }, { "speaker": "Speaker 1", "startTime": "1265.4", "endTime": "1280.2849", "body": "Right? Like, mentioned, like, Midjourney and Dolly and, Stable Diffusion. I think two of those companies no longer really play in this space. And then Midjourney, I think, is now like a biotech something. Mhmm." }, { "speaker": "Speaker 1", "startTime": "1280.525", "endTime": "1290.15", "body": "Yeah. But there's always... Yeah. Now that, like, my point is this landscape's always changing. There's tons of competitors." }, { "speaker": "Speaker 1", "startTime": "1290.15", "endTime": "1312.2599", "body": "And even in the very early days, we were always up against many different competitors and... Or in... Indirect competitors or direct competitors, and we still are today. And I think what we've taken a lot of pride in over time is just really being a 100 customer obsessed. This is the Amazonian principle number one." }, { "speaker": "Speaker 3", "startTime": "1312.58", "endTime": "1313.22", "body": "Yeah." }, { "speaker": "Speaker 1", "startTime": "1313.38", "endTime": "1342.44", "body": "And, yeah, just everything we do, we we do for our customers. Like, most of our roadmap is driven by our customers. And, obviously, we're infusing our own futuristic state and where we think the future is going, particularly in regards to agents. But it's very much driven by them. And I think one of the things that we pride ourselves on is if a customer makes a great request, like..." }, { "speaker": "Speaker 1", "startTime": "1342.44", "endTime": "1356.475", "body": "And we think it'll help the rest of our customer base, we will turn that request around within a few days. And I think that speed of innovation really compounds and has made us the product we are today." }, { "speaker": "Speaker 3", "startTime": "1357.355", "endTime": "1380.28", "body": "Yeah. One more foundry question. Founders struggle for years finding what we call product market fit, the elusive PMF, the zero to one, as they call it. And here you are two and onetwo years later, you have four sixty clients. How did you so it seems like you skipped right through that whole drama, and you went right to product market fit." }, { "speaker": "Speaker 3", "startTime": "1380.28", "endTime": "1405.14", "body": "Any advice for future founders out there of how you was it because you just kind of tapped onto a pain that was so acute? There's this kind of part of product market fit is if you can find a pull, right? So if you've a product and they yank it out of your hand, they need it so much because you've solved their problem so, you know, so urgently. It seems like you just kind of nailed that on your first time out. Or did you have to do like a little bit of..." }, { "speaker": "Speaker 3", "startTime": "1405.14", "endTime": "1408.1", "body": "I don't like calling it pivoting or slight course corrections in there." }, { "speaker": "Speaker 1", "startTime": "1408.1", "endTime": "1424.26", "body": "Yeah. Yeah. I definitely think that the vision has very much stayed the same throughout, which is our our mission has always been to help brands create product that people love. And, I think that's just a really hard problem. Right?" }, { "speaker": "Speaker 1", "startTime": "1424.26", "endTime": "1454.6299", "body": "Knowing what people love, that in itself is a very difficult problem, especially if you're trying to figure out what that is in eighteen months. It's nearly impossible. And then creating the product, it's so expensive. And you're spending 90% of your time and efforts on products that will never actually reach the consumer. So we as consumers, when we shop online, we don't see like 95% of all the other products that they were experimenting with, iterating with." }, { "speaker": "Speaker 1", "startTime": "1454.6299", "endTime": "1489.73", "body": "They got physical samples for. They pitch to the retailers. I mean, we work with a lot of brands who have big wholesale businesses. And so a massive pain point for them is they walk into Nordstrom or Bloomingdale's buy meeting with 1,400 products, and then 45 are adopted. 14 So 100 products did not need to be created at all, which brings us to that Amazon, you start at the beginning or like She Sorry, you start at the end." }, { "speaker": "Speaker 1", "startTime": "1489.73", "endTime": "1511.1", "body": "She and, you show the end before doing all the work. So that's really what we're transforming. And I think the results have... Honestly, there is always, like, a luck factor in starting businesses. And, like, when we started the company, we didn't expect some of the case studies we got so quickly." }, { "speaker": "Speaker 1", "startTime": "1511.1", "endTime": "1567.63", "body": "So one example in wholesale is one of the largest sportswear companies in the world. I mean, sportswear is a very competitive space, they actually were able to double, like, 2.3 x their Dick's Sporting Goods orders because they were actually prompting live with the CEO of Dick's, and they were changing the product and, like, changing the styling of the product and, you know, putting different products together that really inspired, you know, the retailer to place a much bigger bet on their product. And that wouldn't have been able to happen without the fast nature of of Raspberry AI. Right? Because if you wanted to change the styling or you wanted to change, you know, some of the designs, like, you'd have to go back to the drawing board, go back a couple months." }, { "speaker": "Speaker 1", "startTime": "1567.63", "endTime": "1584.855", "body": "You'd, like, set up another pitch meeting with them. And then at that time, in that, you know, couple months that have passed, your competitors might have sold the golf line or whatever basketball line in that time frame. So" }, { "speaker": "Speaker 3", "startTime": "1585.255", "endTime": "1621.4601", "body": "Yeah. Yeah. It's almost in certain categories so I know electronics, mattresses, and then some of the home improvement, like Shop Vacs are famous for this. The the the manufacturers create actual retail specific versions of products and and they get relatively custom. That seems like an unlock you could offer because, to your point with the DICK'S Sporting Goods example, you can go into that buying team and kind of like create a custom product in a way that is way more efficient and doable than than you ever could before?" }, { "speaker": "Speaker 3", "startTime": "1621.4601", "endTime": "1625.2999", "body": "Has that become a new unlock you didn't really think about when you started this?" }, { "speaker": "Speaker 1", "startTime": "1625.78", "endTime": "1650.7", "body": "Yeah. A 100%. Like, exclusive lines, collaborations have been huge for us Because, yeah, it's so much effort, so much designer effort, merchandising, ecommerce, photographers, models, like so much goes into it. That you could just could never, you can... You'd never had the budget to experiment with all of that before." }, { "speaker": "Speaker 1", "startTime": "1650.7", "endTime": "1660.5549", "body": "So I think there's a huge positive, optimistic growth that AI is driving." }, { "speaker": "Speaker 3", "startTime": "1661.835", "endTime": "1666.6349", "body": "Yeah. What other case studies can you share with us about how folks are using the system?" }, { "speaker": "Speaker 1", "startTime": "1666.795", "endTime": "1702.59", "body": "I would say we've evolved over time. So when we started the company, we were laser focused on the design department. And as we started working with these departments, they brought us organically to more downstream departments, like ecommerce, like marketing, like visual merchandising. And again, like visual merchandising and store merchandising was not in our vision, initial vision at all, but they just started creating like the store concepts. Because every store has a different layout, and every store has different end customer." }, { "speaker": "Speaker 1", "startTime": "1702.59", "endTime": "1737.595", "body": "And so we have visual merchandisers at Raspberry now who say that they were spending nights putting on 50 different combinations just to find which one looked the best. But now you can do that all digitally. And then you can just spend your time putting on the one physical, final aspect. Right? So, yeah, I think in terms of case studies, it's really now about, bringing us more downstream into the other areas." }, { "speaker": "Speaker 1", "startTime": "1738.155", "endTime": "1738.635", "body": "And now" }, { "speaker": "Speaker 3", "startTime": "1738.795", "endTime": "1756.44", "body": "There's, like, different modules because it's a different customer. There's kinda like, here's your shelf design system from Raspberry. Here's your product design. Here's your digital marketing design. Or is it or is the designer wanting to pull in all those teams so they can kinda, like, envision the final product in the..." }, { "speaker": "Speaker 3", "startTime": "1756.52", "endTime": "1757.96", "body": "Yeah. On the shelf" }, { "speaker": "Speaker 1", "startTime": "1758.2", "endTime": "1758.6", "body": "ladder." }, { "speaker": "Speaker 3", "startTime": "1758.6799", "endTime": "1761.1599", "body": "On the mannequin on the billboard." }, { "speaker": "Speaker 1", "startTime": "1761.4", "endTime": "1793.805", "body": "Exactly. It's it's more the latter. I think that's Okay. One of the one of the amazing things about AI and the system in particular is, I think we're really breaking down silos and really increasing cross departmental collaboration. So one of the companies we're very inspired by early on was Figma and how they really was a truly collaborative multiplayer system for digital designers." }, { "speaker": "Speaker 1", "startTime": "1793.805", "endTime": "1831.425", "body": "We wanted to do the same thing for physical product designers. And I think it's gone beyond that because AI is so powerful, where designers can now also create billboards, and designers can also help with visual merchandising. Of course, there's still the expertise and specialties that all collaborate together. And that also helps with approval processes and just kind of the boring company corporate stuff. And, yeah, I think in terms of case studies, now 70% of all of our clients work with us in three or more departments." }, { "speaker": "Speaker 1", "startTime": "1832.4651", "endTime": "1862.5349", "body": "The most popular ones are design, ecommerce, and marketing. And then the ROI is, I think, number one is speed. Obviously, you're just able to do things way more quickly. People are condensing their design calendars, condensing their go to market calendars. One of the biggest retailers in the world, they said, in the twenty years that they've been doing private label, they've never been able to turn in design specs on time." }, { "speaker": "Speaker 1", "startTime": "1862.6951", "endTime": "1892.875", "body": "And after using Raspberry, they turn them in four weeks early. And they are also, because of that speed, you know, bringing on other collaborators. So even external collaborators, like the vendors and the manufacturers, are being brought into this collaborative experience. And I think the second area of ROI, which is really exciting to us, is revenue uplift. Right?" }, { "speaker": "Speaker 1", "startTime": "1892.875", "endTime": "1923.86", "body": "So DICK'S Sporting Goods example I talked about with wholesale, we see that with a lot of wholesale customers. Like, they're going into these buy meetings without 1,500 samples. Like, they're choosing, like, the best samples, but they can just show a lot of digital visualizations. And that's overall, you know, 1.5 to 2.5 x ing their wholesale orders. And then on the e commerce side, we're actually..." }, { "speaker": "Speaker 1", "startTime": "1923.86", "endTime": "1942.9149", "body": "We've measured a 15% higher conversion on e commerce PDP imagery compared to traditional photo shoots. And they're 80% to 90% cheaper and faster. So that's really great. I think to your point earlier, that wasn't achievable two years ago. Like, there's no way." }, { "speaker": "Speaker 1", "startTime": "1943.395", "endTime": "1971.5549", "body": "But now it's 15% better. And you could imagine in another year, that could be 25% higher conversion because you're able to show videos, like very high quality videos, and personalized content as well, which as we know, still is a completely greenfield in the space. And then, of course, like the lower cost. Right? It's like you don't have to have seven rounds of samples." }, { "speaker": "Speaker 1", "startTime": "1971.5549", "endTime": "1982.4199", "body": "You can just do all the six rounds digitally, and then you just get one final sample. And, yeah, all the different departments have that kind of cost factor." }, { "speaker": "Speaker 3", "startTime": "1983.46", "endTime": "2002.345", "body": "Yeah. For our audience, I spent a fair amount of time you have an excellent website, and it's I'll warn the audience, it's kind of addictive. So you have a whole you have an area where you can kind of see a sketch and then see the render of it. That's really cool. You can kind of drag a line and kind of watch it do it in real time." }, { "speaker": "Speaker 3", "startTime": "2002.345", "endTime": "2024.555", "body": "And then you have it looks like there is an area of your product where you kind of give it a prompt. When I was looking at that, was like, I can't show this to my daughter because she's going to say, I want to do that as a consumer. Do you see a world where the consumer can do that, where they can kind of prompt? I think someone there's a company is it Arcade? A company that's kind of tried it for jewelry." }, { "speaker": "Speaker 3", "startTime": "2024.555", "endTime": "2025.755", "body": "I don't know how well that's going for" }, { "speaker": "Speaker 1", "startTime": "2025.755", "endTime": "2026.3151", "body": "them. Really?" }, { "speaker": "Speaker 3", "startTime": "2026.3151", "endTime": "2031.3551", "body": "And there's been a couple of other folks that have kind of tried this, but you like, you know, kind of design your own product type of" }, { "speaker": "Speaker 1", "startTime": "2031.435", "endTime": "2031.995", "body": "Yeah." }, { "speaker": "Speaker 3", "startTime": "2032.155", "endTime": "2050.0999", "body": "Do you see a world where, I guess, the visualization it's not easy, but it... It's one of the pieces of the puzzle. The manufacturing is harder, but it's getting digitalized not quite as fast as we are in our world because there's got a... You know? But there's three d printers of weave type products." }, { "speaker": "Speaker 3", "startTime": "2050.0999", "endTime": "2060.115", "body": "There's all kinds of things going on. There's obviously like the plasticky type of things. You wouldn't wear that. But, you know, do do you see a world where the consumer can kinda just design their own products?" }, { "speaker": "Speaker 1", "startTime": "2061.875", "endTime": "2081.905", "body": "Yeah. I mean, I I think the possibilities are endless. I think personalization has been like... We're so early in that. I don't see any brands really nailing what they could do with AI in personalization." }, { "speaker": "Speaker 3", "startTime": "2082.785", "endTime": "2085.665", "body": "M and M's is pretty good where you can put your name on an M and M." }, { "speaker": "Speaker 1", "startTime": "2085.665", "endTime": "2087.185", "body": "Oh, that's much more." }, { "speaker": "Speaker 3", "startTime": "2087.185", "endTime": "2101.8801", "body": "You can't change the iconic shape of an M and M, but Nike has done Nike has done a lot on this area. Their custom piece, I don't think they do it anymore. But for a while, they had a custom that you could go to the store and watch it be built. It was really cool." }, { "speaker": "Speaker 1", "startTime": "2102.2002", "endTime": "2106.76", "body": "Yeah. Yeah. I've I've been to that... Those kind of experiences at, like, Uniqlo. That..." }, { "speaker": "Speaker 1", "startTime": "2106.76", "endTime": "2108.12", "body": "That's cool. You can, like, make your" }, { "speaker": "Speaker 3", "startTime": "2108.12", "endTime": "2108.84", "body": "own t shirt." }, { "speaker": "Speaker 1", "startTime": "2109.945", "endTime": "2156.6702", "body": "But I'm even thinking just like, you like, you have a lot of data on on me as a consumer from my various social media, from my loyalty program, from my previous purchasing behavior. So you could serve personalized images to me that will inspire me to buy the product. But it's just the same old static images as we've had for the last twenty years. But, yeah, I mean, I definitely could see the brands, you know, doing more of custom custom consumer driven design. I think one of the lingerie brands, I think, was experimenting with that a couple years ago." }, { "speaker": "Speaker 1", "startTime": "2156.75", "endTime": "2164.1902", "body": "I don't think they have that system anymore last time I checked. But, yeah, a lot of possibility." }, { "speaker": "Speaker 3", "startTime": "2164.965", "endTime": "2186.66", "body": "The persistent problem in the fashion industry is fit because the human body doesn't want to be it's not the same for So eventually, we'll have a world where you can go get a body scan. And that's a whole controversial thing. Those are things I've never done well. But eventually, people will get comfortable with that. And then have the item made perfectly for you." }, { "speaker": "Speaker 3", "startTime": "2186.66", "endTime": "2192.5798", "body": "It's kind of like a Seville Road type of a situation, but more democratized." }, { "speaker": "Speaker 1", "startTime": "2192.74", "endTime": "2193.22", "body": "Right." }, { "speaker": "Speaker 3", "startTime": "2193.2998", "endTime": "2223.99", "body": "So it seems like you guys could play a pretty big role in that where the fit measurement part, again, digitizing, you could work with mid journey, it could be part of the body scan. Just kidding. The body scan part, it's relatively technologically easy to measure things. Then marry that together with your system, you could have the ultimate show the look type of a situation if you wanted to down the road too. Build a custom item for yourself." }, { "speaker": "Speaker 1", "startTime": "2224.895", "endTime": "2226.335", "body": "Yeah, exactly." }, { "speaker": "Speaker 3", "startTime": "2227.535", "endTime": "2273.91", "body": "The other thing I noticed while I was watching some of the demos and kind of the especially the prompt piece where you go kind of like sketch, add some different colors and textures, then render. Doesn't feel and feel free not to answer this specifically, but sometimes you can use a third party model or underlying technology. This feels like you had to go build a bunch yourself to understand, right, if I'm going to do certain types of leathers and fabrics and wool looks different than cotton looks different than satin, so on and so forth, the way it hangs, the way it glistens, then you can rotate them. All those problems are not easy. So I feel like you guys have had to go build a bunch of that yourselves." }, { "speaker": "Speaker 3", "startTime": "2273.91", "endTime": "2276.79", "body": "Is that a fair guess?" }, { "speaker": "Speaker 1", "startTime": "2276.79", "endTime": "2298.83", "body": "Yes. It is very fair. Yeah. We have an amazing team of computer vision and image generation specialists who've... We've trained a lot of our own models for exactly the use cases you mentioned, but also for construction and, like, physical feasibility." }, { "speaker": "Speaker 1", "startTime": "2299.55", "endTime": "2309.9502", "body": "Right? Like, you can create all these awesome artworks mid journey or whatever program. But at the end of the day, someone has to sew... Cut and sew that." }, { "speaker": "Speaker 3", "startTime": "2310.605", "endTime": "2316.205", "body": "Yeah. It it has turned turned back into a two dimensional thing that that can be actually made." }, { "speaker": "Speaker 1", "startTime": "2316.4448", "endTime": "2326.73", "body": "Yes. And we generate, tech packs as well, so, like, all the technical specs in addition to the images. So it really is grounded in the physical realities." }, { "speaker": "Speaker 3", "startTime": "2326.73", "endTime": "2352.25", "body": "So I can go from two d sketch, three d render, and then maybe build a prototype off that of some kind. And then you then go back into then I say, perfect. Then you generate... Then you to it back into my my... I would call it a bill of lading, but, like, all the subsequent pieces I would need to then go build that product, including the pattern, the the directions." }, { "speaker": "Speaker 3", "startTime": "2352.25", "endTime": "2357.13", "body": "I don't know what a tech package is, but I imagine it's something the sewing companies Yeah." }, { "speaker": "Speaker 1", "startTime": "2357.13", "endTime": "2383.0999", "body": "Yes. It's it's like the measurements and, like, the stitching should be this far apart and, like, this is a double stitch and that this is this is the ribbing on the hem. And yeah, it's very detailed instruction manual for the manufacturing firms. And then, exactly, yeah, your point, exactly. There's also a parallel path, which is, Okay, we like this product." }, { "speaker": "Speaker 1", "startTime": "2383.0999", "endTime": "2415.95", "body": "We love how it looks. Let's create all the marketing campaigns. Let's create on body digital visualizations so that we can start to test that. So I think that's one of the really exciting things some of our more advanced customers are doing is they're using what's being created in Raspberry to test whether that's like their loyalty program, or some of them have apps where you can... You know, as you as a you as a consumer of that brand can can say, I like this." }, { "speaker": "Speaker 1", "startTime": "2415.95", "endTime": "2424.0298", "body": "I don't like this. I would like this if the color were different. I would like this if the trends were different, to really give them strong buying signal." }, { "speaker": "Speaker 3", "startTime": "2424.4448", "endTime": "2432.125", "body": "You mentioned earlier when you were pitching, you kinda had to help them understand the future state. What... Help us understand the future state. Where's where's this going?" }, { "speaker": "Speaker 1", "startTime": "2432.365", "endTime": "2439.66", "body": "So I think there's the where's Raspberry going in the next twelve months. But in AI Hard" }, { "speaker": "Speaker 3", "startTime": "2439.66", "endTime": "2440.7", "body": "to look out past that." }, { "speaker": "Speaker 1", "startTime": "2440.7", "endTime": "2442.54", "body": "Yeah. Things that I can" }, { "speaker": "Speaker 3", "startTime": "2442.54", "endTime": "2443.9", "body": "barely see six months on my world." }, { "speaker": "Speaker 1", "startTime": "2443.9", "endTime": "2444.54", "body": "I don't know I" }, { "speaker": "Speaker 3", "startTime": "2444.54", "endTime": "2445.82", "body": "don't know how you see 12." }, { "speaker": "Speaker 1", "startTime": "2446.6199", "endTime": "2461.645", "body": "Yeah. Maybe six months. Yeah. So I think in the next twelve months, we see Raspberry expanding in two key directions. One is expanding more into other retail categories." }, { "speaker": "Speaker 1", "startTime": "2461.645", "endTime": "2493.8652", "body": "So we have clients in the home, in furniture, in soft home, in luggage, in jewelry, in beauty. But there's so much more. Like, we have to build the reputation that we have in fashion in all those other retail categories as well, and really personalize the systems for those industries. Because none of those industries I just mentioned, none of them have a specific industry specific vertical solution. They're just using the general image models." }, { "speaker": "Speaker 1", "startTime": "2494.745", "endTime": "2521.665", "body": "But I know that the art of creating beauty and and furniture is much more complex than what a general image model can deliver. So that's one direction. And then the the other direction is, going deeper with the retail industry and really helping them measure everything that they're doing in Raspberry. Right? I think we we kind of think of it as ras..." }, { "speaker": "Speaker 1", "startTime": "2521.665", "endTime": "2540.765", "body": "We think Raspberry is more of a... You're not just ending with this beautiful image, right? It's really like a supply chain of the content. And so once this beautiful image is published online, you want to know, is it converting better? Like, which angles convert better for my brand?" }, { "speaker": "Speaker 1", "startTime": "2541.9648", "endTime": "2557.3", "body": "Which, like, kind of models convert better for my brand? What kind of lighting converts better for my brand? And that... The answer to those questions is actually we find is very different for for every single brand. Because..." }, { "speaker": "Speaker 1", "startTime": "2557.3", "endTime": "2589.72", "body": "And it's different for different groups. Like, so for for some men's companies, like, they have the best conversion when they're just doing flat lays or ghost mannequins. But for women's, like, they have better conversion when there's on body because they wanna see how that's draped and how it fits. So I think all of these questions have, since the Internet has existed, been very much like hearsay. Like, you hear from other brands or, like, you do a little bit of testing yourself, but it..." }, { "speaker": "Speaker 1", "startTime": "2589.72", "endTime": "2627.585", "body": "It's never had a robust way to optimize your images and to use your images as a, like, amazing asset to inspire people. So that's that's really, you know, where we see Raspberry is is kind of if you think about, like, SEO and GEO, there's been a lot of investment in those areas for very good reason. But I think the area that I see as the future for the next twelve months, for Raspberry at least, is really what I call IEO, which is image engine optimization." }, { "speaker": "Speaker 3", "startTime": "2627.585", "endTime": "2650.7", "body": "This comes up a lot in our world where we're trying to connect the product catalog out. The brand will look and say, where is that image coming from? And we're like, that's why we really need to get this dialed right. You spend all this time. And for whatever reason, the answer engine can't map what your product image is to the catalog in the way that is capturing your hero image." }, { "speaker": "Speaker 3", "startTime": "2650.7", "endTime": "2663.335", "body": "They're showing some kind of the classic is it's like the garment on a rug from an eBay listing. Kind what they found. And you're like, yeah, not really what you wanted there. Exactly." }, { "speaker": "Speaker 1", "startTime": "2663.335", "endTime": "2688.605", "body": "Yeah. And then I think in the beyond twelve months, there's just so much happening in the agentic space. And something that's really big on our minds is how are agents going to process images and process the listings and really understand those assets differently than humans. And how can we prepare brands for that reality?" }, { "speaker": "Speaker 3", "startTime": "2689.2449", "endTime": "2703.55", "body": "Yeah, absolutely. Awesome. Well, appreciate you taking us on a little bit of a founding story and helping us understand what it is Raspberry AI does. I love your future state and got to have you back on when you've built more of it." }, { "speaker": "Speaker 1", "startTime": "2704.43", "endTime": "2706.19", "body": "Thanks, Scot. This was fun." }, { "speaker": "Speaker 3", "startTime": "2706.5898", "endTime": "2707.63", "body": "Yeah. Thanks for coming." }, { "speaker": "Speaker 2", "startTime": "2707.6418", "endTime": "2716.442", "body": "Follow, rate, and share to stay up to date with Retailgentic, your guide to the future of AI shopping agents." } ] }