We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Blaine Bolus
00:01:43 - 00:01:56
What's up, DTC pod? Today we've got Sam Radner, who's the co founder and CEO of Showroom. So Sam, I'm going to let you kick us off. Why don't you tell us a little bit about your background and what you guys are building.
Sam Rattner
00:01:56 - 00:02:22
Yeah, sure. Prior to showroom, I was in a much different business. I was in the online gambling business. Last company was called Victory. I had dropped out of college to build that business. Essentially, we had sports betting products, we had table game products, blackjack, bachelor, stuff like that. We were an authorized gaming operator of the NBA and Major League Baseball. We had what's called market access in a handful of states the licenses to operate an online sports book.
Sam Rattner
00:02:22 - 00:02:56
We had also gone through traditional compliance and at the time, every legal state. We didn't yet have market access in every single one of them, but we were through compliance. This was kind of at the peak of the market, if you would. There were also the peak of the streaming and gambling wars. You had everyone, like DraftKings and Fanduel and all the streaming companies like Hulu and ESPN. Plus Hulu had been in the market with what it felt like every athlete with Hulu sports doing all those commercials. And so at the. It made a lot of sense to try to align ourselves with a partner.
Sam Rattner
00:02:56 - 00:03:20
We had gone down the path of talking about a joint venture with some larger companies in sports and media. Long story short, made a lot more sense to sell the company. So we sold the company of Fubo TV. I stayed for about nine months to help transition that business. I was the chief operating officer of the gaming division. We opened up our headquarters here in Chicago and then kind of moved on to start showroom. But that's a little bit on the background.
Blaine Bolus
00:03:20 - 00:03:31
And what's your personal background? How did you even get started with the business in the first place? What inspired you to build it? Yeah, just take us back a little bit further to how you guys actually started the business.
Ramon Berrios
00:03:31 - 00:03:34
Well, he was 23 when he sold it, so that was very early.
Sam Rattner
00:03:34 - 00:04:05
Exactly, yeah. I had always played a lot of online poker, so I knew the ecosystem a little bit. I knew the community. The community was interesting at the time. Most of the sites were offshore sites, and so you were essentially in group chats or on the side panel of a poker lobby with people from all around the world. None of it was English. So I plugged in some Google translator plugin thing to figure out what people were talking about. And everyone was talking about how the US was about to legalize online gambling.
Sam Rattner
00:04:05 - 00:04:28
They were saying next year. It was 2017 at the time. This happened in May or June of 18, if I recall. And so I was like, that's interesting. I feel like the market will evolve similarly to what it did overseas, where you bet in Tottenham game stadiums and stuff like that, versus here, where it's, like, frowned upon. I think I said, on my first million, I think I said something. Know, if your grandma buys a lottery ticket, you don't think anything of it. If she's taking the bears, you're like, grandma, what are you.
Sam Rattner
00:04:29 - 00:04:54
There's. There's still a stigma here in the US a little bit, although it's, you know, I think a lot of the major leagues now? I'm not sure. I think Major League Baseball now allows some of these sports folks to know brick and mortar operations at the ballparks. And I know the NBA and NFL are working on it. They haven't done that already. So that was kind of the thought process is I loosely knew the online gambling ecosystem from being in it myself, and then kind of saw a market opportunity.
Ramon Berrios
00:04:55 - 00:05:19
So the dynamics are so different than fashion shopping, which is fully evolved if you don't look at it through the lens of AI, and then you don't have these sort of loopholes and regulations to build modes and get exclusive licenses and access. So tell us, what is showroom and what was the opportunity you saw here, given how different it is?
Sam Rattner
00:05:19 - 00:06:08
Yeah. So there's always a should you get into a business that's in a similar industry that you're in, or should you get into something, whether you want something new or you're bored of it, or you just think there's no more market opportunity? While always more daunting, I think that's where a lot of value is, because you have zero inherent bias to how a market works. All you care about is what does the consumer want? And sometimes, if I rewind, had I known everything I know now about outline gambling, I would have talked myself out of ever building that business that would have been impossible to build. And candidly, most early stage startups didn't make it out with an exit for market purposes. It's capital intensive, cacs through the roof. So regulatory barrier to entry. So I think it's inherent. A lot of value is actually sometimes in the.
Sam Rattner
00:06:08 - 00:06:38
I don't know what, I don't know because you're going to hit the roadblock. If you build the business, you're going to hit the roadblocks either way, whether you knew about them or not. And sometimes it's just better to not have. Sometimes it's better to just run into a glass wall and not have known, like, eventually we're going to have this problem, because then when you run into it, you're like, oh, I didn't know about, like, how do we solve this? Not like this daunting. I've been waiting for this thing to hit me like a train. And now here we. Showroom is. And I think, Blaine, you said it well before the podcast started.
Sam Rattner
00:06:38 - 00:07:20
Showroom is what everyone wishes something like Google shopping was. And how I look at online shopping, especially in our market, which is fashion, footwear, jewelry, accessories at the moment, at least, is we used to walk up and I'm from Chicago. We used to 2030 years ago, you would walk up and down Michigan Avenue and it would take an hour or two to shop. You'd have to go into store by store. You'd have to find what you want. Then Google came around with traditional search, and now you can find what you want in 2 minutes. Over the last 20 years, Google Search has become very big box brand. If I would, knowing very few people are going to outbid Calvin Klein or Tommy Hilfiger for the keywords underwear or denim jackets.
Sam Rattner
00:07:20 - 00:08:00
There's all this live shopping. Now you have TikTok shop, you have other things, and you start to realize people are scrolling and scrolling and scrolling, and all of a sudden it's been an hour, and you've been scrolling through products on your phone. You could have just walked up and down Michigan Avenue. And what it really is to me is it's actually more entertainment than it is transactional. Even though you might be scrolling for so long that you then stumble into buying something, you don't necessarily have a ton of brand equity with it. And a lot of those products are like kitchen gadget esque, if I would. And so they're less reliable distribution channels for the merchants themselves. They're much better for TikTok to take fees than they are reliable for the merchant.
Sam Rattner
00:08:00 - 00:08:30
And for a user, if they're looking for something, they find what they want, and then they want to see reviews of it. Well, then something like TikTok or YouTube is great. But when you're trying to discover what you want, I don't think that's the best way to discover. And in my opinion, there's times for entertainment and there's times for when you want to buy what you need. And I think most people want to get on, find exactly what they want, and get off. They don't want to sit and look for the dress. They're looking for the pair of shoes they're looking for for an hour and a half. If they find the exact thing they want to look for, then they'll read reviews and they'll watch videos.
Sam Rattner
00:08:30 - 00:08:37
But I think it's the product discovery aspect of shopping that I think is pretty broken with traditional search.
Ramon Berrios
00:08:37 - 00:09:13
I mean, yeah, it's the impulse versus intent purchase, and the conversion of the intent buyer is the best one there is. Even given how successful the GMV of TikTok is growing, the conversion rate compared to how many people are watching TikToks every day is not as high as an intent buyer. I know you have this story of this meeting you were in, where you basically saw the opportunity and realized what was going on with sort of verticalized channels, like pages specific. Can you tell that story?
Sam Rattner
00:09:13 - 00:10:02
Yeah, sure. And I always like to focus on the consumer because just because businesses and merchants might want it doesn't mean consumers want it, but it does give some validation that there is a propensity for the market to pay for it should you solve the problem for consumers. And I was in a meeting with an advertiser who ran a direct to consumer cowboy boot brand company, and she made this comment that stopped me in my tracks in a meeting. And she's like, yeah, we could do something like that, but we waste fifty cents of every dollar we spend on Google. And I was like, whoa, whoa, whoa, whoa, whoa. What do you raise my hand? I said, excuse me, hold on a second. How is that the case? And she's like, well, we pay almost $4 to own the word cowboy boots, to show up first on Google shopping. But half the time, someone looks up cowboy boots.
Sam Rattner
00:10:02 - 00:10:37
It's a third grader looking up cowboy boots. For their third grade presentation. They needed a photo to print out from Google Images. They had no intent of buying $1,000 pair of cowboy boots. Previously, she had been a senior marketing executive@cars.com and she said, and this was the first time I had heard this term candidly, which was, that's why vertical search is so great. And I said, what do you mean by that? And she said, when she was@cars.com people Google Hyundai for all kinds of reasons. They might be looking for a video. They might be looking for the photos of the new interior.
Sam Rattner
00:10:37 - 00:10:53
They might be looking at their stock price. If you look up Hyundai on cars.com, you're a buyer or you're a seller. It's the only reason you're on there. You're not scrolling through for entertainment like on TikTok. You're not looking at the stock price. Maybe you're looking to buy or sell a car. And so the intent is through the roof. Everyone on it is buying.
Sam Rattner
00:10:53 - 00:11:56
And so the brands love it because the car brands and manufacturers or sellers, I don't know exactly how the business work, but also sellers, they know that the only people on it are looking to buy or sell, which is good. I think this idea of vertical search, I had never, ever in my life thought about it this way. But all these vertical search platforms, or vertical discovery platforms are actually Google's competitors. Up until maybe this Chat GPT integration, I used to think Bing was Google's competitor. Google's competitor was anybody that stole search traffic that would cut them off, that they didn't go to Google because they just went straight to Expedia or Kayak or travelocity, or they went straight to Pinterest. Now, there's some secondaries and there's some primaries I would call Expedia and kayak and travelocity secondary search engines. Some people might start their search there because they're a big rewards user, but a lot of people start their search on Google flights, and then they have you stumble into Expedia because they paid to be there. Most people do not stumble onto Pinterest from Google images.
Sam Rattner
00:11:56 - 00:12:43
Pinterest is a primary search engine destination as to where people chose to go for photos and visual inspiration. Most people, we will be more like Pinterest. Most people will not be stumbling onto showroom from Google shopping, showroom will be a primary destination as to where millions of people choose to start their product discovery to find what they want to buy in our vertical. But I thought it was very interesting that I had never thought to myself that just because someone calls them a search engine is actually not who the search engine's competitors are. The search engine's competitors are who would, if this other company did not exist, had typed in search that on Google. That's Google's competitors. And so they still get a ton of the market share from all the secondary ones like Expedia, because they're paying to be there. And then it's kind of like an arbitrage game between them on just keywords.
Sam Rattner
00:12:43 - 00:13:08
But I would call the Pinterests of the world stockx and footwear. You could say like, oh, eBay sold shoes for years before Stockx. Well, the reality is stockx didn't just paint a new logo and say, we're in this business now. They had domain products and services that were better than know. They had sneaker verification. They'll warehouse it for you. They'll handle transactions. They'll do all these things that eBay wouldn't do.
Sam Rattner
00:13:08 - 00:13:27
So it's inherent in this vertical search. So everyone on Stockx has intent, and then the products and services are better than what the alternative was. You have to have both. And so that's kind of how I think about vertical search businesses. And out of all the verticals that I think have terrible experiences, that consumers are clearly looking for alternatives. I think it's online shopping.
Blaine Bolus
00:13:27 - 00:13:30
So why don't you tell us a little bit about online shopping?
Sam Rattner
00:13:30 - 00:13:30
Right.
Blaine Bolus
00:13:31 - 00:14:08
When you say online shopping, obviously great opportunity for vertical search. But on the contrary, there's a ton of different marketplaces that already exist where you can shop in a more traditional sense. Think more of your big retailers. They kind of become aggregators. You've got blogs, you've got curation platforms, you've got Amazon. So when you're launching this platform, where do you start? What brands do you work with? How do you index inventory? How do you think about creating your store? And how do you make your store and your platform really stand out from all the other places that someone can go if they want a more vertical shopping experience than a traditional Google search?
Sam Rattner
00:14:08 - 00:14:36
Okay, so first I'll talk about the markets we don't compete in. We don't compete in fast fashion, cheap fashion. Anyone who's trying that, you will probably lose that game. To me, that is a, assuming the executive teams are smart and rational people is a capital game. Timu will win that game. Sheen will win that game. The capitalized players with great talent will win that game. So that's one end of the spectrum.
Sam Rattner
00:14:36 - 00:15:37
On the far other end of the spectrum is high end luxury. Big, big retailers, especially the online aggregators like italist that have all these european brands, and we know them pretty well, they'll always win the online aggregator retail game. And then if you have a brick and mortar footprint, Macy's, Sachs, Neiman Marcus, Burberry. So we sit from our inventory perspective, it sits in the middle of, we're not high end luxury where everything is $2,700 coats, but we do not compete in fast fashion. From the value prop perspective, the reality is a lot of these merchants, the retailers, when I say a lot of the retailers, they have more skus than some of the larger brands, but not by much. I think a lot of people think Macy's might have 10 million fashion skus. Not the case. It comes off that way because their properties on Michigan Ave.
Sam Rattner
00:15:37 - 00:16:09
Are massive. But that's actually not the case. They usually carry a couple hundred brands, and so it's like more limiting than people realize. And that's why they go from revolve to Nordstrom to Neiman Marcus, because one of these retailers, for whatever reason, didn't have what they were looking for. So how I approach the problem is every retailer has their own initiative. Some are in certain sectors, some are in know. They focus on men, maybe more than women, whatever it is. But if I'm a consumer, it's frustrating to have to go to merch, to brand by brand and retailer by retailer.
Sam Rattner
00:16:09 - 00:16:59
I wish I could just communicate with all the retailers I just want to look through all the inventory. That and traditional search is very rudimentary. It's non conversational. Right now, how search works up until these language models has been the single input results and the single output of results, those outputs are unrelated from one another. And these engines don't really learn more about me over time, and they don't know exactly what I'm looking for. Even if I have a couple of follow ups. That should give you a big clue of like, I'm looking for a dress and then I'm looking for a dress that might go well with cowboy boots. And then you keep going, well, if every single output was unrelated from one another, what do we know? Well, we know that they might be going to Nashville or they might be into that kind of fashion, but a traditional engine doesn't know that.
Sam Rattner
00:16:59 - 00:18:02
And so I think the value prop is in that you can have this ongoing conversation as opposed to this rudimentary search where you're going back to the search bar on Lululemon to redescribe what you're saying, and then it can actually rationalize what you want, provide options, and then you can have this ongoing conversation. And then as you progress and do that, it knows more about you over time. And so I think that's really the value prop, is I think people have brand fatigue. There's more brands from a fire hose than ever before. That said, consumers clearly are more willing than ever to buy a product from a brand they've never heard of after a five second TikTok. But everyone at the company's opinion is consumers are more interested in buying products from new brands, but they don't want to sit around and scroll for 2 hours and go through all these different sites, keep tabs open so they don't lose something, and then they accidentally closed it out. And so to have an infrastructure where you can kind of search throughout the entire Internet at once in a conversational way, and then it actually be transactional versus Pinterest. I think only 20 something percent of all the images on Pinterest are actually linked out.
Sam Rattner
00:18:02 - 00:18:26
They're usually linked out to some Instagram post because most of the content is from influencers, not from brands. So you'll find what you want and then you find yourself going back to the Google search bar to try to describe what's in this photo. And while you could use something like Google lens, candidly, I think most women between 25 and 45 don't use Google lens. And so they'd much rather a place where only can you find exactly what you're looking for, but then it's one click checkout right there. And so that's what I think the big value prop is for the company.
Ramon Berrios
00:18:26 - 00:18:27
Cool.
Blaine Bolus
00:18:27 - 00:18:53
And so for any marketplace or aggregator, right. You need to be able to serve both sides. You need to give the consumer what they want, but it's also got to work on the back end for the merchant. So how do you guys cater to the merchant? What's the merchant experience? Where does it stand now? I know you guys are kind of in beta and you're expanding and where do you see it going? And how do you benchmark yourself against something like a Google or like some of the other marketplace options that merchants can sell on?
Sam Rattner
00:18:53 - 00:19:45
Yeah, so we have revenue share deals, we have a couple thousand brands on showroom. We have revenue share deals with all of them. We have a fair amount of direct deals with the brands themselves, whether these are companies I had previous relationships with, or we've had a lot of companies reach out. We also have a lot of deals where they have pointed us to their aggregator partner companies like Rakuten Share, CJ, you know, in the same capacity of a company like Honey, who now obviously at their scale, being under PayPal and everything, they've got better integrations with merchants, but primarily started with, they went and got the rev share deals to put the parameters and the links from the aggregator networks and almost had no relationship with the merchants. So we have a little bit more of a handheld relationship at the moment that'll grow over time. Eventually, we'd love for showroom to be the merchant of record. I think that's going to be on the roadmap for 2025. There's some value in people still checking out at the merchant.
Sam Rattner
00:19:45 - 00:20:14
People get rewards points. They like doing those things. Not to say that you couldn't also do that going through showroom as the merchant of record, but that would allow us to operate more like Amazon than not, which is you're checking out on the platform, you're earning rewards with showroom, if you want to return things, you can return them through us. We don't want to touch inventory. I want to stay out of that game. I think that's why companies like rent the Runway have struggled, because they actually control the inventory. I think that's problematic. So we don't want to touch the inventory.
Sam Rattner
00:20:14 - 00:20:37
But if we have deeper technical integrations with merchants and we build out our back end council a little bit more, to have API integrations for logistics and returns and customer service, well, then we can be the middleman and facilitate everything. And candidly, you open yourself up to a much larger revenue opportunity by being the merchant of record. But until then, people don't complain, they don't have any issues, and they're okay with checking out at the merchant. Sometimes they like that better anyways.
Ramon Berrios
00:20:37 - 00:21:17
Yeah, it reminded me of honey now that you mentioned that. But I think what's interesting as well is brands have no insight into any data of their consumers, only through their DTC website. Like retailers are sharing no data now. There's like a few companies popping up, sort of starting to share data back. But I think in your case specifically, there's also a huge value proposition on here's what people searched to buy your product and here's how they're describing the experience they want to have and how that ties into your product. Like you said, bachelor in Nashville, whatever. That is really valuable data for these companies.
Sam Rattner
00:21:17 - 00:22:15
Yeah. If you are a brand and you have products at a retailer, you don't really know how the consumer got there and you don't really own that customer. With us, when you're on the back of the showroom, not only do you see the customer and you own the customer because as of right now, they're checking out with you. But then on the back end of once you integrate with us, you can see how people got to the product. And that informs a couple of things. The easiest example is if someone searched, like you just said, what dress should I wear? That goes well with cowboy boots to Nashville. Well, now they might have some insight into some other words they can buy on Google shopping that might lead people to a similar outcome. A bigger one, in my opinion, if you dig deeper, is if you realize 30% of all the people buying this product are actually coming from a domain you didn't think about, well, that could inform your creative team as to what content you should be putting in paid.
Sam Rattner
00:22:15 - 00:22:56
So knowing how they got there is really important. And you don't get that if you sell through a retailer as much. I don't want to say completely because I think some retailers share data, but there's also a lot of times where there'll be people who bought at convenience through a retailer, whether it's Amazon or Macy's. You bought something at the bottom that says other people have also bought this, so someone might buy it. You're going to realize the revenue because you got the sale, but that's a little bit different data than if they were primarily looking for the product. You should really want to know every which way someone bought your product who was looking for it in a primary fashion, and you don't always get that data.
Blaine Bolus
00:22:56 - 00:23:29
We are really excited to announce that DTcpod is officially part of the HubSpot podcast network. The HubSpot podcast network is the audio destination for business professionals, and we're really excited about being part of the network because we're going to be able to keep growing the show, bringing you guys amazing guests, and obviously helping you guys learn from the best founders, marketers and builders of the most successful consumer brands. So anyway, keep listening to DTC pod and more shows like us on the HubSpot podcast network@HubSpot.com. Slash podcastnetwork so how are you guys.
Ramon Berrios
00:23:29 - 00:23:45
Thinking of AI in know, you know, right now in the very early days, first implications we're seeing is like e comm, product photography, et cetera, but actually powering commerce. How are you guys implementing AI and how do you see it unfolding?
Sam Rattner
00:23:45 - 00:24:32
So there's been some really good things I've seen on Twitter or X, I guess if I should. There was one, I'm spacing on the name of the company, but it was virtual tryons where you could take a product and put it on any size. So if you're a bigger guy or you're a bigger woman, or you're taller or you're shorter, you can put it on that. There was an associate at a venture fund, and the reason she really liked that is she said, I'm a black woman and I can't see what this would be like on me. But now I can change the skin tone of the model and now I can see it. So I think that's really great. Unfortunately, I think for some of those businesses, they're a little bit more of a feature than a company. Consumers aren't going to start paying for every vertical of the stack.
Sam Rattner
00:24:33 - 00:25:37
They want a place that has it all. So those are more things where we should integrate all that as a company like showroom than I think is as valuable in building it. What we think is valuable is in things that entirely change the trajectory of the journey and make it not a little bit better, not ten times better, make it a thousand times better. And so I'm very long the conversational component, I think GPT has made it easy that anyone with an Internet connection can now have a conversation with a chat agent that sounds organic, who knows a lot, vast knowledge. But most of the platforms today that are the broader models, like perplexity or others, they're not going to ever really dig in on the details of all these different verticals, they want to provide the infrastructure. And so I think we think of it as there are components that you can use AI that make the product marginally better. And if you incorporate 10, 15, 20 things like that virtual try on, then that's great. But them on their own doesn't completely change the trajectory of the journey.
Sam Rattner
00:25:37 - 00:25:44
So we start with the fundamentals of what changes the entire shopping experience and makes it 1000 times better. And we think that is conversational.
Ramon Berrios
00:25:44 - 00:26:02
Know that being said, I'm curious, did you ever consider not racing for the business, just given how cheap it is to run OpenAI and a bunch of other APIs? Did you assess that or was it always like, this is only going to work if we're just the top one?
Sam Rattner
00:26:02 - 00:26:05
When it comes to why I chose to raise capital, you're asking. Yes.
Ramon Berrios
00:26:05 - 00:26:06
For showroom.
Sam Rattner
00:26:06 - 00:27:19
Yeah. So I always think of what's the goal of the business. If the goal of the business is to build a really great 50 or 100 million dollar business over the course of five years, own most of it and sell it, well, then you should try, especially early, raise as little capital as you can, raise capital at convenience, get to revenues, potentially be raising debt. If your goal is to build, and not just because you want to, but because the business actually could, if executed, be a one 2510 billion dollars business, well, then the reality is there is 100 competitors in every space. And capital isn't always the core constraint, but it is a fundamental constraint, and you can only get so much done with little capital, candidly. And so how I think about it is, if I was trying to bootstrap a business, and I woke up one day and I saw that a competitor of ours raised a $25 million round and hired two senior executives, would I care or not? If I don't care, well, then you don't raise. But if you think, not only do you care, is it a big issue, and if it's a big issue, well, then you have to go raise capital and you have to really think to yourself before you start those businesses. Do I have the ability to raise the capital? Because if you don't, you're going to struggle.
Sam Rattner
00:27:19 - 00:28:06
Candidly, you're going to struggle. You can't bootstrap a business where all of your. I shouldn't say you can't. It's very difficult to bootstrap a business when all of your competitors are one in a market that's moving really quickly and are raising significant capital. There is a component to it where people over raise and then they don't know what they're doing and they, I guess whatever you say, you get lost in the sauce of having the cash and you're not focused on the customer. But not everyone who raises goes and raises $200 million and drinks mushroom coffee every morning. Even if they raise 510 $15 million, you can do a lot of damage with 510 $15 million against someone who's trying to bootstrap. So to me, it's, do you think not having the capital is worse for the business, could ruin the business because there's other competitors? And then you have to go raise.
Blaine Bolus
00:28:06 - 00:28:57
Another question I have for you, Sam, is kind of going back to what we were talking about, about vertical search and where people go to do separate things, right? One of the topics in AI that everyone asks about all sorts of AI products, whether it's shopping or whether it's consumer SaaS or whatever it is, is why can't Chat GPT just do this, right? I know Chatcheept and OpenAI and some of these models, they're growing up. They're adding in plugins so you can connect to Shopify and some other things. And you'll type in the search and it'll say looking up in Shopify or whatever plugin it's using and start to return you results. How do you differentiate yourselves from the default? Or do you see that as, hey, that's great chat. GPD is a good on ramp for people to start exploring conversational search, but when they really need to get what they actually want, we're going to have a better infrastructure and ecosystem built to tackle that problem.
Sam Rattner
00:28:57 - 00:29:57
Okay, so the big fundamental problem is if you can just plug in GPT to some data set and build a wrapper, well, then anybody could do it. The big problem in e commerce, specifically in fashion, footwear, jewelry, accessories, is that every merchant saves their data differently. Some merchants use Shopify as a back end, some use woocommerce, some have Magento. Neiman Marcus has some huge enterprise that they built ten years ago with some custom development company. How that data is saved, when you then crawl that data, it comes in differently. A language model, because there's some constraints to how many tokens you can have in system prompts, there's cost associated with doing different things, really needs to be trained on a single internal data set. So I think, for example, in hotel and travel, I think Expedia or kayak could crush this. They already have the data set.
Sam Rattner
00:29:57 - 00:30:36
They already have tens of millions of users. They know the data. They have it organized. It's one way you can train a language model to understand your data set. But what if you had 3000 different data sets? How do you train it? There needs to be a constant variable of the data. And so the example I gave recently on morning Brew was, if one merchant has a dress and that dress comes in green, they have a little green wheel next to the product. Another merchant also has a dress that only comes in green. But Marissa on the brand team decided that it looked silly to just have one green circle because without any others, it makes it look like we don't have options.
Sam Rattner
00:30:36 - 00:31:16
So she said, let's get rid of that and put green in the title of the product or into the product description. Well, now, when you crawl that data, I have one skew that looks like it's missing a color. And so you need to kind of standardize this data to be able to train a language model against. And so that's what's taken us a year to build, is we essentially, and there's dozens and dozens and dozens of things like color. There's sizing and there's pricing and there's shipping. And even within color, for example, one of the ways we solve for color with missing skus is we take out a pixel from the garment. We find a pixel that's within spitting distance of the RGB code, and we know what that color is. So then we can assume what color that one is.
Sam Rattner
00:31:16 - 00:31:51
Well, it gets really tricky when you have things like pattern, floral, metallic, polka dot. What pixel are you grabbing? And then even once you know that, if it's a top, you got to make sure you don't take the pixel from the jeans the model is wearing in the image URL. So color is one of 50 things that in and of itself are very complicated. The only way to do it is to have this standardized format. So think of us as like this big funnel, and all these brands save data differently. And you can't just try and change a language model on every single one of them. You need to get it all into a standardized format. So that's what we did.
Sam Rattner
00:31:51 - 00:32:20
And so, in my opinion, the real value in AI is either you have an existing data set that in and of itself is proprietary. Bloomberg would be a great example of this. They're starting Bloomberg GPT, or whatever they're calling it. They have all this proprietary data of the trading that's happened on Bloomberg terminals for the last, whatever, 20 something years. No one else has that data. And then they can build a GPT around that. So you can now ask it questions related to your financial information. That's the value.
Sam Rattner
00:32:20 - 00:33:13
Or is there something that's so broken in such a mess that even if they wanted to, the market participants would not solve because Macy's will only ever solve it for themselves. They might standardize their data, but why would they go solve it for Sachs? So you need like this independent company who has no bias and doesn't have any initiatives other than to solve it for the consumer and has relationships with all the brands. There's only a couple of companies that would have that. Google has that, but Google's keyword driven Shopify has that. But Shopify will only ever have what they have on the shop app, which is Shopify merchants, which they should. They're never going to push Nike. Nike doesn't sell on Shopify. And I feel like in absence of the big brands, it feels a little bit Alibaba esque because it's a lot of brands you've never heard of versus Google's all the brands you've heard of, but that's all they have.
Sam Rattner
00:33:13 - 00:33:50
So I feel like we're not a platform for their backend. So we don't only push our own merchants, we're not Google shopping. So we're not only keywords, we don't care who the brand is, we want to show the consumer the best option and we're getting a revenue share. So all I care is I show them the best option. We're getting paid either way. And candidly, I think the conversion will be a lot higher because we're showing you exactly what you want, not what someone paid to be there. And so I think the value is either you already are existing Goliath company and you have existing data internally that's proprietary, that no one has, or you can go out and solve a big data problem to standardize it and then train a model against. That's where I think the value is.
Blaine Bolus
00:33:50 - 00:34:08
So is your North Star basically, like how do you think about it? Is it merchant quality or merchant quantity or. Right, like when you have to work with all of that stuff right now, are you focused on signing up as many merchants and working with as big of a catalog and standardizing as big of a catalog as possible or being very selective into what comes into your data set?
Sam Rattner
00:34:08 - 00:34:37
No, we're selective. Yeah, we do a lot of diligence. You have to have great product. You have to be building a brand. I think it's funny, we find that the brands that only sell women's products on average are significantly higher quality. And I think our conclusion to that, especially in direct to consumer view Shopify, is that there's a lot of guys building brands that are just like trying to arbitrage something from overseas. They're not really building a brand. So we find it a little bit more challenging with the men's brands.
Sam Rattner
00:34:38 - 00:35:00
Not to say there's not incredible brands, cuts, clothing, I mean, they're great, but we are very selective. If you just have the inventory of everything, the last thing we want to do is be sending a showroom user to check out on a product from a site that has no return policy and payment options are like through Western Union or want. There's some standardization. So we have a laundry list of things that we look for when we're onboarding a merchant.
Blaine Bolus
00:35:00 - 00:35:24
I love that. That's super cool. Really excited to see you guys continue to grow and launch and deploy the platform. Really excited to dig in, play around with it, maybe do some shopping ourselves, Ramon. Maybe we'll buy a thing or two and kind of, as we wrap up here, what are the next plans? I know you guys are getting ready to go into your big public launch, but what do you have to get right this year? What's on your roadmap for the rest of 2024?
Sam Rattner
00:35:24 - 00:36:08
Yeah, so we have a big waiting list. We launched beta the other day. So starting with a couple of hundred people a day, we'll start to add a couple thousand people a day, fix any critical feedback that we're getting on the platform, and then be in the market hopefully in February and March with we have hundreds of creators ready to make content for the company. We have a big short form social strategy across TikTok, YouTube and Instagram. So 2024 is growth mode for the company. We're capitalized. So we have solved the capital constraint issue of the business. Now it's can we execute? And if sometime in 2024 we hit certain metrics from a user and revenue perspective that it makes sense to raise a growth round, we will.
Sam Rattner
00:36:09 - 00:36:19
But we feel pretty good right now. But I would say late 22 and 23 was build the platform. 2024 and 25 is build the business.
Ramon Berrios
00:36:20 - 00:36:28
So Sam, how can people that are listening join the waitlist or merchants that might be a fit are interested to join? How can they find out more too?
Sam Rattner
00:36:28 - 00:36:45
Yeah, so the waiting list is still available at showroom store. We have a different domain for the beta. We will unlock you in the order that you're in line. You can also, if you're a merchant, reach out directly to me at my personal email. I'm being cognizant of saying that you.
Ramon Berrios
00:36:45 - 00:36:46
Can just say your socials.
Sam Rattner
00:36:46 - 00:37:00
Yeah, last time I did a podcast I woke up to 4000 emails. My Twitter is at Sam Ratner. Very simple. I give a lot of updates on the company there and I'm always in my dms so that's the best way to get a hold of me.
Ramon Berrios
00:37:00 - 00:37:02
Awesome. Thank you Sam. This was awesome.
Blaine Bolus
00:37:02 - 00:37:03
Thanks for coming on the show Sam.
Sam Rattner
00:37:03 - 00:37:05
No thank you guys. Thanks for having me.
What is Castmagic?
Castmagic is the best way to generate content from audio and video.
Full transcripts from your audio files. Theme & speaker analysis. AI-generated content ready to copy/paste. And more.