Ralph Shackhardt, William Blair, Internet Analyst here. Thanks again for attending our annual Growth Stock Conference. Before I forget and do my value add, please check our website for all disclosures. So, get that done. Today we're really excited to have CEO and founder David Wright and CFO Jason Beasley here from Pattern here to present. This is their inaugural William Blair Conference. They went public in September of last year. Stock is up about 40% since the IPO. They've had two exceptionally strong quarters being public. you always leave a little bit of dry powder, but they did a really excellent job the last two quarters. It's a tech company that positions itself between brand partners and over 70 online marketplaces, primarily Amazon, but they also have Walmart, Target, and a bunch of also international marketplaces as well. The business model is pretty straightforward. They take inventory from the brands. They have many patents and very proprietary technology to optimize listings. And they also manage the logistics to deliver the products to the consumers through third-party platforms. And then also, David will get into this, but it's a really interesting company. They really sit at the forefront of what's going on with Agentec AI and commerce. So I think that's obviously fairly topical at this conference as well. So with that, I'm going to turn it over to founder David Wright. We're going to do a presentation. If there's a few minutes left, we'll have Q&A, and then the breakout room afterwards. I think it's Jenny B following this as well. David, go ahead and turn it over to you.
Okay, thanks, Ralph. Well, thanks for having me. This is my first William Blair conference. It's a pretty phenomenal execution by the team. I think they do a really good job here. Maybe quickly, by way of background, my background is all on the tech data science side. I never really thought of starting a company, you know, until my cousin, she started up this little girl's headband brand, called Emma Jane, and I was talking to her husband one night, we're going through some spreadsheets, of course, and he's like, you know, the data is rather available, and it's quite formulaic, he says something along these lines. And then, you know, he started helping out on a couple little things, you know, trying to figure out, okay, what is your true return on ad spend? and how might you increase conversion, what everybody else is doing, and you don't have to have, what I liked is I didn't have to have very many original ideas, I could sort of see what was going on and just reverse engineer the data. And then we started helping a few other people, now we're $3 billion in, and we have 22 offices around the world, and amazing team, and we just continue to keep executing. So we'll walk through it a little bit here. So, what we do is we essentially help brands like the headband brand with profitable growth across global e-commerce marketplaces. This is the footprint now. These are our quarterly numbers there on the left, as you can see, and I think I can walk around a little bit, right? I'm miked up here anyway. I'm not very good on a podium anyway. I'll just bounce over here so we can walk through a couple things. things. So one of the numbers that I think is the most key is, you know, that net revenue retention. In my opinion, that is how you know whether the machine works. If a brand has been with you for years and you can continue to grow them, and if you think of total digital growth across the world, it runs, depending on the source for the data, somewhere in seven to 10% range. So we're actually growing our existing brand's 127% and that includes any attrition we might have as well. As long as we keep doing this, when I tell people you don't even really need a sales team, that will start speaking for itself. And so 43% year-on-year growth, great numbers overall. This is what we've spent years and years doing. We're collecting now, you know, upwards of, you know, near a trillion data points, you know, on a monthly basis. Actually, I guess weekly basis. 41 patents either issued or pending. These are our numbers. And since I, we took a little money out of the IPO because we had to, it was a big board conversation and I ended up on the short straw of that. And, you know, so then you're like, what do I do with this money? Like, you guys are doing it. I'm like, okay. So I started looking at stocks. And Jason Beasley, our CFO, is like, you know, he says, this is one of the most beautiful revenue charts I've ever seen. And now I'm starting to look at a lot of companies. I just told them this morning. I'm like, you know, you're right on this. I can't find them. But we've had great success over the years, very consistent growth. But the thing is, it's somewhat natural. It's a massive space. And as long as you do a very good job representing brands, you just continue to grow. So you can see we continue to get more and more profitable. Our business is a bit of a game of scale as well. The bigger we get, the more our costs go down, which allows us to sign more brands. The price point is just more attractive. Then you can see free cash flow generation. So here is the thing that I thought we would be a nice solution for some smaller brands. And I remember quite early, we had a meeting that I thought was somewhat miraculous with Panasonic early days. And we met with one of their divisions, he's president of the division, gave him, you know, talked to him and I'm like, hey, how big is your e-comm team? He says three. I about fell off my chair because I was like, wait, Panasonic, and he's like, well, we have 15 divisions, they all run separate. it, and I've learned since then all of these brands and companies generally do this. And 70% of brands have teams of 10 or less. And so what they need to do is optimize our whole tech stack is based on that formula right there. So revenue for a brand is traffic. How many eyeballs can I get to see the product? When they are there, do they convert? At what price? And is the product available in region? and within how many hours, like if it's available same day, it's better than next day, all of those types of things. You can start playing this formula out and a brand has to do, you have seven people that have to do that across the world or even multiple marketplaces in region and as soon as you breaking this formula down, which it's quite fun to do. You start war rooming this a bit and you say, okay, what are all the levers of traffic? You see the top left one there, for those of you who can see it, probably know one, but it's influencers. You could literally have the entire team of seven at a brand just focus on that. But if they didn't have some pretty good technology, they would probably underperform. They might underperform anyway. And then you start realizing, okay, what are all the levers for conversion? What are you doing versus, say, your competitors are doing? What type of content archetypes are you using? Is the stats and studies type of image convert better than a close-up image? You can basically digest the world of data and start serving up to a brand, and brands have tremendous success, but there's no shortcut here, really. You have to go through this process, and it's where Pattern has become a pretty phenomenal solution for the brands. So I'm going to walk through a little bit, and if we were to dive into a tech demo, which I would feel most natural on. We can walk through some of the tech here, but this is where our patents are either issued or pending. One of our first ones, you can see Destiny there. So Destiny, as an example, we make about 15 to 18 million bid changes at a keyword keyword phrase level across all of these marketplaces a day. So the idea behind Destiny early days is we said, hey, if you have a good product at a good price and you're competing against other products that naturally a consumer we believe would pick this product over that product, why do they not win? Could we identify across a myriad of 30,000, 40,000 keywords or keyword phrases, which now in the AI world you convert to semantic intent, we would say, where would this product have mathematical propensity to win over another one? And then we would chase it with ad spend to see if we get traction. If we don't, we abandon ship, go on to the next one. But you guys are probably familiar with machine learning. The output of machine learning is called a label. So our label would be winnability. By the way, if you have a machine learning model and it chases organic winning across across a marketplace. We own the patent on that for every marketplace in the US. People don't quite appreciate that until they really get deep in the space. But that's destiny. That runs 15 million bid changes a day. We finally got this patent, finally through the whole process, government-wise, called True ROAS, which has been a burr in my saddle forever. However, the idea here was, our early headband brand idea, my cousin was like, hey, supposedly my return on my ad spend is 5.6, right? And I was like, he just put in a million dollars, because your 60% margin, you should just put in a million dollars. And she's like, it doesn't work. I'm not getting the ROAS that they say I'm getting. And I'm like, I at first thought, you know, there's no way that the marketplaces could get away with misleading people, you know, on some of those numbers. But it's very hard for a brand to figure out, if I put a dollar into the machine, how much do I get out? And I think the problem is, is almost every technology out there takes some percent of ad spend, or an agency takes a percent of ad spend. So from the beginning, we've said we will never make money on ads. because I just think you won't ever chase the real answer. So he mentioned our model. The way we monetize is we actually buy the inventory. So all ad spend, they pay for, but it's complete pass-through. So then I can chase true ROAS. We have quite a bit of technology behind that. But I think we might be the closest in the world I know of to the right answer. It's not perfect, but I don't know of anyone closer, and we could walk through the reasons why there. And then on the conversion side, there's quite a bit of a very interesting tech here. You know, one is a content, we're essentially taking all the world of content, taking all of their conversion rates, comparing it to what you're doing, and every nuance. We're like, hey, there's three of your competitors that include the word organic, and we also use eye-tracking software where we'll say, where is it placed? What's the likelihood somebody saw it? We serve all of that up. We just run it through a model, and we say, hey, we believe that the word organic would help you move conversion by 20 bips, whatever, right? And you have some slight improvement there, and for a brand, it's really quite material. So, this is the moat, you know, in an AI world is that, you know, 13 years of data collection. Okay, so this is the business model. Until AI really came around, I spent like a full decade getting beat up over this model. Everyone was like, why don't you just sell the technology? SaaS is the best way to go. And suddenly, everyone changed their mind on that, which I quite like. But so what we did, you know, early days, I'll go back to that Panasonic meeting, early days we basically said, hey, is there a subscription model or we'll buy the inventory. Served it up to Panasonic, the guy said, let me take it to the board, he went to the board and he came back and his quote was from someone on the board, he said, take the check, smile and wave. So they essentially were like, hey, let them write us a check for the inventory. And if you think about it, we make more total contribution margin this way. It's super sticky. And we can get scale on the logistics side. And if you look at that 127% NRR, we don't have to go and renegotiate every time we build. We just built the portal, which is AI-generated product photography, there's a dome and whatnot. and we don't have to go back to every brand and say hey to use the portal it's another you know five hundred thousand a year it just as long as it increases sales then you know we get that same margin and the brands are happy because they get a bigger paycheck so net net this has turned out to be a pretty phenomenal model I wish I could say I thought of you know it was this strategic idea in the beginning, but it was two things in the beginning. One is I was sitting with a few brands, and I was like, no, this is going to work. I guarantee this is going to work. At one point in a meeting, I was like, I'll buy the inventory. If it doesn't work, it's on me, and the guy's like, okay, and then I realized I made more money that way anyway, so we kept doing it, but it turned out to be a great model. and now we handle export import and just this logistics piece I'll just walk through one piece of scale here but if if you just take marketplaces you have to take a good and you have to inbound it generally to a marketplace to their fulfillment right that process is quite complex and it used to be you need to take a product from a brand and you inbound it to one location say for Amazon, and they spread it around across the network. Then they came up with the idea, hey, we need to have, we need to get closer to the consumer, so you need to send it to three locations. So if you have 100 products, Amazon will tell you, send 74 over here to New Jersey, 20 to California. So you have to break down all the packaging. Think if you're a brand how complex this gets, and the pricing to inbound all of these goods. So for pattern now, and now it's 14 to 18 locations now that they have regional nodes and so forth, my total inbound cost is 11 cents for every item across the board to inbound. I don't think anyone can touch it if I can do full truckloads and we're 95% full truckloads because of scale. So you just start realizing, okay, if you're a brand, how often can you do a full truckload? And you have to order on a weekly cadence, or you won't get the same same-day metrics. So this has become a pretty phenomenal moat. Okay, now, what I talked about in the beginning, this NRR, this number. Of course, we sign new brands, but I'm just going to focus over here on the NRR number. It was up from 115 the prior year. 75% of that growth is technology related. So some lever in traffic conversion, some improvement. The rest is we're tripling international and then non-Amazon growth, that'd be TikTok shops and Amazon coupon is doing amazing as well. And then the last bit of that would be a brand saying, hey, I gave you this product line, can I add these three more? That would also be NRR. So a lot of brands will say, hey, why don't we start with, you know, it's quite easy if you get in with a brand to have a conversation like, oh, well, I don't want to take a big risk. Maybe, you know, but I don't have a team in Japan or South Korea. Let's start there. So you start there and you expand. Okay. So this This is, the scale has become a phenomenal driver for us. We track what we call a cost to serve. So for every dollar that runs through our machine, we say, how much does it cost us to operate that dollar? Now a combination of technology, AI, logistics, just general overhead that we don't need as we scale, we're about 30 to 40% down on that number over the last few years. So this becomes, I'm starting to become convinced that a brand cannot run at the same price point as us with the scale. You know I use the one logistics example but they also don't have all the data. The more the more brands that we sign, the more data we get on the entire ecosystem and it just starts feeding this fly well where price comes down, our opportunity keeps going up, and the space is quite large. So just back to our favorite chart here. So this is a little bit of history. We've acquired quite a few companies over the years. We acquired this company called Current Tech. They were a platform for social commerce. And then we acquired this company here that was called Practicology early days because they did international e-commerce consulting and they had quite a presence internationally. You can see the name in the beginning. I remember drawing that logo for my partner on the first our very first lunch and I was like hey what do you think of this name and she was like I think it looks pretty good and so we were really proud of ourselves and then we hired a legit CMO CFO, or a CRO, and some meeting where they're like, you know that ISERV name, it's bad. So then we landed on Pattern here in 2022. But we've always been, I mean, we had patents around natural language processing machine learning before anyone really cared to even talk about AI. So we sort of had a natural transition into the transformer models, which gave us an enormous us. That was almost just pure luck, I guess. Okay, so a couple things. We just launched what we call pattern intelligence, which takes all of this and rolls it into an interface. I am super proud of this. If anyone's really interested in going deeper, I'd love to spend some time with you on this. But if you think a little bit about what is your favorite enterprise If you think about it, you're like, maybe some of the best ones I guess would be maybe a Salesforce, Workday, and cloud native. But most of us are like, those are bad. Not a bashful. It's just a very hard problem to solve, because you build an interface the way you think it might work best. And then the LLMs came around and we're like, okay, I enjoy this experience of chatting, but now I'm missing maybe core metrics. So we built a push-pull model that will essentially say, okay, we believe here's the key things that you should care about with regards to traffic conversion. Some of these other things, we serve them up in an interface where you can just converse around the data. and then it meets you where you work. We're fully integrated with ChatGPT. So you can just sit in ChatGPT and use Pi natively there if you want. You can use it via Slack. You can use it wherever you work. Your approvals will drop into your email. It will drop into your Slack. If there's a set of content, we're like, hey, here's a new set of content. We'd like the brand's approval. Drop it in over there, and then it will pop up and say, okay, here's the core messaging pillars. here's all the data as to why. If they would like to ask questions about it, they can interact with it. I think that this is where, once you look at this interface, I think you'll realize, okay, you will see over the next three years, five years, most companies, I believe, switch to an interface that will look somewhat like this, and they will be able to roll most of their technology under that umbrella where they can have an interface that people enjoy. So happy to talk more about that. I might just highlight the portal here which is at the top. I don't think I'm overstating this to say from AI generated product photography like a Gemini or Nano banana does not cut it. It does a phenomenal job with the basics so like you like hey I want to show my product in a you know with a jogger but as soon as you say hey I want to show them holding the product the machine starts to fall apart and we've demoed this a million times and I have a hundred and fifty thousand SKUs that I have to run globally so I need this to work so we actually built some hardware started you know just you know underneath the desk with a blanket and we were essentially taking you know some of the backdrop that we were copying Hollywood a bit. And so now this is a dome and it's about 16 feet and we've actually got lighting. We have a SpaceX engineer that we've engineered the entire thing and has a robotic arm and we can now automate the process of shooting photography. If I get about 60 to 80 images out of the portal. With that data, it's reference data enough I can train a LoRa model, so a low-rank adaptation model. And from then on, I never need to shoot that again. So I can use that and create product photography that is very high quality in any language, anywhere around the world. And you can't shortcut this. You have to shoot it somewhere in order to get that type of quality. So that's a pretty interesting and fun advancement that we have. Okay, I guess we have, maybe I finished a little bit early, but no one will complain about that. So I'd be happy to answer any questions either now or later, however you want to do that, Ralph. Well, you know, it doesn't make any preference that way. It's just where the consumers are. As a matter of fact, our non-Amazon numbers are growing faster than Amazon numbers, but internally we don't really have... Investors always love that when we show diversification there, but we're not really trying to. It's just TikTok Shops is becoming quite an interesting platform. We continue to do international expansion, Kupong is doing amazing. You know, but where the consumer is, so I'll give you Mexico as a good example. It's about 50-50 Amazon MercadoLibre right now. I don't know who's going to win, and we represent, say, a Pandora jewelry there, and so we're just brand focused. We're trying to help Pandora win, and wherever the consumer ends up choosing, we don't care. So we don't have to do the consumer chase. All of those, you know, we're this middle layer. We're not spending enormous amounts of money chasing the consumer, all the customer acquisition costs. We're just layering on to wherever they're winning. Well it's a very, very deep partnership, so we don't do anything. First, we require exclusivity to the channel or else we won't engage. So the contract that they will sign with us is, hey, Pattern gets Coupang South Korea. So marketplace, geography, those two things we require exclusivity on with the brand. They'll sign that, and we don't sell it anywhere other than that spot. They can watch us very closely if they want to. We're not a distributor in any way that way. We're just trying to help them win on that one channel. It's not really a problem that we can solve, particularly for a brand. We help them. Because of antitrust reasons, one seller of a good can't have any enforcement mechanism them that would be considered horizontal on, say, another seller of the good. It has to be vertical. So the brand, if the brand doesn't like the pricing that exists there, so we do connect them with a law firm we think is amazing, and we do pay that bill. So we have, over the years, we've said, okay, this is a big enough of a problem, and brands are not investing enough money to solve it, and if they don't solve it, we won't move any goods. So the tech won't even turn on. So if we're not the seller of record, for one, I don't get all the data, because the seller of record gets all of the data advantages, and then our bidding engine, they don't work if we're not the seller of record. So we have to solve that problem. It's a great question you call out. So we just connect them with this law firm, and then they work together, and we pay the We're not allowed to know what they're doing or how they're talking about it. We just know that the law firm has generally been able to solve these problems for other brands. Well, the thinking is, I don't know if it's right, but the thinking is if digital is growing 7 to 10 in totality, to hit 115, you have to nearly double that, especially if you take attrition out of our numbers. So I think it wouldn't be fair to overstate that to an investor to say, hey, we expect it to be 127 when that is sort of a crazy good number. Now, as we develop technology, we start saying, okay, yeah, this is going to be a great quarter because we're going to move conversion by 30 basis points across the board. We've moved conversion since I think our first earnings call from about 15% Now it's about 18% across the entire portfolio. So that will have an impact. Right now, we just keep winning. But I think we're happy with a 115 number because it's still doubling what's happening in digital out there. So we think in the long run, that's a good target. But we keep beating it.
One last question.
And the question is, is it mostly, did you say opening new marketplaces? Oh, yeah. The fee for service versus inventory model? Yeah. So there's two areas. There's one line item in our revenue numbers that we call SaaS services or SaaS logistics and other. that line item grew 173%. And we've actually started selling that logistics middle mile piece I talked about. We actually are moving more volume now from non-pattern owned goods through that middle mile offering than pattern owned. And we charge a margin for that, but it gives us better pricing power across that whole channel. So that's in that number. And then we only take on large brands on the inventory model, because say a brand is doing a million dollars. And you've seen our EBITDA last quarter, I think it was 6% million dollars. You're not making enough money on an inventory model to make that worth your time. And so in the past, we've always said no. So anybody who's small, we've just said no. And we've had enough requests for just the technology solution. And I was just at a conference recently and I was like, well, I guess maybe I'd do it if you guys paid me a shit ton of money. And the guy in the audience, he's like, he goes like this, he holds out his wallet, literally like that. So I was like, well, talk to that guy on the sales team. And so there is just a need for the smaller brands, pure tech solution. So, we have done that, and we have a lot of brands, but that's been more recent, and that you'll find in just that SaaS line item. But we do require the sales team to go through a few more hurdles. It's not the model we want to be for the larger brands, but it's a great solution for some of the smaller ones.
Unfortunately, David, we're out of time. Thanks for the great presentation. Thanks for the questions. The breakout session is that Jenny be upstairs.