Learnings From Starting & Scaling Reverb To A $275M Acquisition With David Kalt | EM Group Chat #191

Welcome back to Everything Marketplaces, where we talk with the founders and leaders from some of today's top marketplaces. So this is episode 191, which is a really great group chat that we just had with David Kalt, who's the CEO of Pango Books. Pango Books is a leading marketplace for buying and selling used books. Prior to stepping in the role of CEO at Pango Books, David was the founder and CEO of Reverb. Reverb scaled to become the largest marketplace for buying and selling music instruments and gear. It was notably acquired by Etsy in 2019 for $275 million. So this is a great chat with David. We're going back to how he started Reverb, some of the early challenges. He walked us through the growth and eventual acquisition by Etsy in 2019, shared some of his key learnings, what led to him joining Pango Books as the CEO. We did a deep dive into his leading marketplace for buying and selling used books today, got to learn about the growth. David shared how they're leveraging AI, how he thinks AI is going to impact marketplaces going forward. Lots of tips for early-stage founders, and we also had a great group Q&A. I really enjoyed this conversation. You're going to find it great. Watch to the end. So David, you have some really incredible marketplace experience. And as I mentioned to you right before we started recording, you know, have led the way for many of us building vertical marketplaces with Reverb. So I'm really excited to have you join us here today. I want to start off by saying huge thanks for taking the time to do so. We have a lot that we're going to dive into with all of your experience previously starting Reverb and now with Pango Books. But I thought it'd be great if you can start off by briefly sharing your background for those here that might not know you.

Awesome. Thanks, Mike. I'm really excited to be here. And also, like, I'm kind of new to the Everything Marketplace community and very excited to see such a robust community of people that are sharing ideas and leaning on each other. So I'm glad to be part of it as well. My background, you know, been doing this for a while building software and tech. So I got a recording engineer background, wanted to, you know, mix records like Jimmy Iovine for Tom Petty. That was the That was a big goal. After a year of doing that, realized I wasn't going to really, I didn't have the skill or the resilience for that. And I started writing software and started coding in the '90s, early '90s, started my first software company in '93, '94 in the travel space. It was a B2B CRM before CRMs existed for tour companies and travel agencies. In '96, as Marc Andreessen launched Nescape, it was pretty clear that my customers were not going to be gatekeepers for travel bookings. So I sold that business in '98 and sort of pivoted completely to options trading. Ironically, I had the money I made from my first business, a few million bucks. I lost most of it trading options at Charles Schwab. And I said, this is a real crime. Retail investors are being taken advantage of. And I started an online brokerage firm in 2000 with a few partners called OptionsXpress. We really were addressing a product market fit for lots of other retail options traders who had similar experience and was able to grow that very quickly with very limited capital. $2 million we raised and took it public in '05, $2 billion value from a $2 million capital raise. So I come from this like extremely efficient use of capital background. And then ultimately we sold it to Charles Schwab in 2010. Then I pivoted and said, I want to get back into music somehow. So I bought a guitar store in Chicago called the Chicago Music Exchange. That store led me down the path of buying and selling gear, music gear and musical instruments led me to the unfortunate world of eBay and the challenges of buying and selling on eBay as a retailer as well as a musician. And I started to hear frustrations and community and that really, at the same time, after 2 years of owning Chicago Music Exchange and growing that business, I was very much inspired by Etsy. Etsy was the first in— this is 2011, 2012, the first player that said eBay is not Amazon. Like they are not gonna win this for every category. They clearly, and it was, we used to call this in 2012, the dismantling of eBay. Category by category because it wasn't obvious. In hindsight, it's so obvious, but then it wasn't obvious that eBay wasn't going to dominate in all categories. At that moment when I saw that Etsy built community around crafters and people doing it, it was so obvious to me that musicians needed that same community. I dropped everything. It was middle of 2011 and I started prototyping Reverb at that moment. I had bought the domain almost 3 years earlier, wanting, knowing that it was just a good music-oriented domain. So we really launched it in 2012, um, with this idea that musicians needed a better place to buy and sell gear, more trust. The other thing I realized early— oh, I'm sorry, I'm going too far in the intro.

No, no, this is actually great. So we'll dive into a few of the topics in more detail here. But yeah, it would be great to, you know, uh, go back to some of the early stages for Reverb, if you can kind of walk us through it and some of the early challenges.

So yeah, like the, the sort of like aha moment for me and why I was able to, after making a pretty significant investment in Chicago Music Exchange, the retail store that I bought, clearly I realized my background is in tech and I love building software, building a new, you know, building an e-commerce platform for a retail store that had inventory. It's not, there's really no drive in that. I wanted to go build something, but I needed, I needed a product that I could get passionate about and that there was a community that needed it. So like that Etsy moment was a real aha moment that like, Yes, build it and they will come. And so we started off with this, you know, basic strategy and Dan Orkin, who I know has presented here before, is on the call and he was an early Reverber in the beginning. He drove content strategy. But the whole idea was building a community for musicians and building a whole lot of trust and a whole lot of data around what was a very opaque industry in terms of, let's call it, used musical instruments. And we knew that eBay was doing around $1 billion in GMS at the time, GMV. And we were, we were going right for that. And we also knew that it was pretty much a dealer-driven market. And our vision was to create more of a peer-to-peer alongside dealer-driven market. So peer-to-peer and B2C coexisting, which is a challenge for any marketplace, right? Because you do have these distinct products, someone making a living doing it, someone looking to generate extra cash. But we thought we could marry those 2 by being— by creating great content, fair pricing, and building trust. And just an example of some of the lack of trust in our industry: people, sellers— no, no, buyers would buy an instrument, post-sale, and these are $1,000, $2,000 musical instruments, post-sale, call up buyer, call up seller who was maybe a musician who didn't— wasn't a professional seller, find a nick on a guitar or an instrument and extort money from the seller after a transaction. This happened over and over again on eBay, making it so you wouldn't want to sell on eBay if you were like, wow, that guitar was perfect. Like, there was nothing wrong with it. I took every photo possible and they're extorting them. And they— and what do they do? Or we'll send it back. And no one wants to take it back then. That was the first thing we did is we said, if we can't remove that sort of that extortion that exists on eBay, we aren't going to have And we started kicking people off and we started really spending time removing bad messages, removing profanity, removing the sort of ugliness that existed on eBay. And that might sound like obvious to people now, but it's, it wasn't, it wasn't so obvious that you actually had to enforce and police and create a community just like you would any sort of community. And that was, that was sort of big learnings in the beginning. The other advantage I had clearly after OptionsXpress, I had not only capital but access to capital. I'm not gonna be naive about that. I ultimately raised $50 million at Reverb. We sold it for $300 million. We used $50 million basically to drive Google PLAs to dominate and get a 30 ROAS when our competitors like Guitar Center and Sweetwater and other music retailers that weren't marketplaces we're probably getting a 15 to 20 ROAS. So that sort of ROAS arbitrage was really, really important in our ability to go from, let's say, an idea in 2012, 2013 to by 2021, they're doing almost $900 million in GMS after Etsy bought them. When we sold to Etsy, we were doing around $650 million in GMS and around $60 million in revenue. It's doubled since Etsy bought it in 2019. Etsy 2 weeks ago sold the business to Fender for roughly between $80 and $100 million. My sources have told me, and Reverb's doing over $100 million. And just, this is the bad part of the story for people in the marketplace world. Unfortunately, we sold it at 5 times revenue. Etsy was trading at 8 times revenue in 2019. Etsy sold Reverb at less than 1 times revenue. at $110 million of revenue and around probably $8 to $10 million of EBITDA. And Etsy's trading at 2 times revenue. So timing is everything. Clearly markets compress when growth stops, right? Marketplaces were sexy in 2012 to 2018. I'll say goods marketplaces. Goods marketplaces are less sexy now from an investor point of view. Investors are chasing other things. Doesn't mean they're not good businesses. I got back into the marketplace business because I believe they are great businesses. If you can really build a moat around buyers and sellers and bringing them together in a unique format, that's really, really hard to recreate. It's not impossible, but most people don't have the resilience or the determination to stick with it. And that's what I've seen in marketplaces. I backed a bunch. I meet with a bunch of marketplaces all the time and most people just don't have the resilience. It's 2 to 3 years before you start getting any sort of positive signal to say that you would do that. And then the capital and the preservation of capital to reinvest in your marketing and growing that snowball, right? So I'll stop there, but I could go on for days, Mike. So I'll stop and pause and let you interject.

Yeah, no, this is great. I feel like this is like a masterclass on building marketplaces kind of compressed in. And you got lots of head nods when you mentioned the resiliency that's required as well. You know, so I, uh, kind of going back to Reverb though, so across some of the various, uh, you know, stages, you know, what were some of those, like, I guess, like, you know, kind of key points and, and stages of growth for the company and then maybe kind of challenges as well?

So the first, um, the first 2 years, I really wanted to prove it out. Having had a success, my investors at OptionsXpress had made a lot of money, and I was coming off of an investment where I'm a believer. I'm, I'm fine with failure, I am, but I'm not a believer in spending other people's money foolishly and I'm not huge on the VC community. Okay. So I'm not like— I don't believe that VCs and private equity add a ton of money. Now, there are some great ones out there that have operating experience that can be very beneficial. But in general, capital is capital in my view. This is once again, I'm very opinionated and these are my opinions. So I was not gonna raise money until I had real proof of concept because the people I put on my cap table are like friends and family and people that I've made money. I'd rather work for them than a VC. I'd rather work for people I care about, but I want high probability outcomes. I'm not looking for moonshots. I've never looked for moonshots. So for me, the first 2 years was literally proving that ROAS out, getting it for, to a high ROAS so that I could go spend money efficiently to get more customers and to proving out retention. The retention gets better over time, but early on we had really good retention signals. And the other thing we had was we had an amazing ROAS observation early on that what the customer clicked on in our Google PLAs, and I'm throwing out a lot of acronyms assuming that these are marketplace people and they know them and you can ask me questions later, but Google PLAs, Google ad product ads that What you clicked on wasn't what you bought. You clicked on a beautiful $10,000 guitar, you bought a pack of strings, or you clicked on a pack of strings, you bought a $3,000 guitar. We had the power because of the inventory, because we were used and because we were aggregating such unique, interesting inventory. We had the power of taking a click and turning it into something else. It made our algorithm on the PLA side harder. But it made our retention greater because there was so much discovery opportunity. Combine that with Dan's amazing insight in terms of content and engaging people once they came here and giving them reasons to come back and a sticky experience rather than the product being the only driving factor to sort of get recurring traffic.

Those are great to hear about, and I'm sure it's gonna lead to some questions here. And when we get to the group Q&A, you know, kind of looking back though now, you know, what were some of your, I would say, like key learnings from Reverb that you've taken with you to Pango Books as well?

I mean, right, my key learnings from my last 2 big exits and tech companies is that as sort of a coder, product-driven marketer. That's who I consider myself. I got frustrated with the correlation between more people and lack of product innovation. And VCs will tell you just the opposite, higher, higher, higher. Well, now in the world of AI, maybe not, but for the last 20 years, it's been take as much capital you can and throw people at it. And either it's my personality, my leadership, how I work with people, I'm very, very into the details, very, very grounded in obsessing about customer experience and product experience. Therefore, I found as I added more people, the correlation between enhanced product experience, enhanced customer experience, and people didn't correlate at all. In both of my companies, as I got to 250, 300 people, product suffered and my frustration level suffered. And that's literally why I called Josh Silverman at Etsy and I said, you wanna buy Reverb? I'm like, I've done what I can. 300 people, $800 million, just $600 million in GMAS, $50 million revenue. I think you guys could be a better shepherd. And that was, that was that conversation. So Why I'm so excited about Pango is what you can do with Claude 4 and Cursor in coding. And as a CEO who is a coder who loves building product, I am just blown away by what I can do in 2023, '24, '25 relative to the past. So I think it's like the ability to build scalable marketplaces by, by really understanding how to build. Is we're in a whole new, whole new league now. It's also going to create tons of disruption for marketplaces that aren't very innovative and don't know how to adjust their engineering and their product teams to today's environment. But that is my, my biggest takeaway is, uh, and I'm actually might get back involved in Reverb. Reverb sold Defender. I'm friendly with Defender people and I've had conversations with them. So there might be an opportunity and it would, that would be the one thing I would do is How to reduce that engineering team to the most critical engineering team that can build that are 100xers, not 10xers. Like you need 100xers now. And that's what I'm building out. We do 20% of the order volume that Reverb does, but we do like 1/30th of the revenue because our books are like $12 versus $300. But we're handling that with 2 engineers, not not 80 engineers. And I really believe that's the winning model of figuring out your efficiency so that you can scale your business at, at, at levels that are, have not been seen in, in history.

Yeah.

No, those are definitely some great learnings and I'm, I'm glad you brought up the topic of AI. So that's actually one I wanted to, you know, spend some time on. So it's probably a good transition, you know, into discussing, you know, how you're leveraging AI at Pango and maybe how you think AI will impact marketplaces as well.

So there's a sort of obviously leveraging AI on the build side, right? Either working with your engineering team, your engineering co-founder, whatever it is, making sure they are using the best building tools to either refactor code or build new code. The other is automating your soft things. Like, I think, I'm not sure if Grant's done one of these yet, but Grant, who's the founder of Pango Books, And he's been my mentor. He actually built the product and taught— got me back into coding, ironically. And I'm mentoring him on marketplaces. But he went and built this tool called Filter, which really helped us do some of the things like content moderation, product moderation, automation, finding holes where customers weren't getting emails or customers were being bad messaging between customers, like those kind of things. These are things where we put a lot of resources at Reverb and Pango in trying to make a good experience, not just letting the marketplace happen, right? Trying to control the bad stuff and eliminate them. And AI is amazing for that, right? Not that it replaces the humans, but it enhances the humans' ability to process more, find more, and act on more things. So I'm a huge believer in the tools there. And the other thing I think in the AI world is One of the things working against us as small marketplaces are subscription SaaS tools that we need, whether from Google Cloud to Algolio. I could go on and on to Segment, Twilio, blah blah blah blah blah, Klaviyo. These tools get more and more expensive as your business grows. It's it's literally it's extortion in terms of how these enterprise contracts are negotiated to the point where your infrastructure and your serverless technology becomes. Really a huge expense, almost up there with your marketing expense in terms of how you're gonna scale your business when you really start to get 10,000, 20,000, 100,000 monthly active users. AI is gonna help you build those things yourself. It's gonna help you replace extortion from SaaS companies that are not going to be able to keep up with the pricing versus what you're gonna be able to build on your own. So those are like, 3 of the things that I'm really focused on in terms of AI. Yeah.

And Grant actually shared a breakdown post with us as far as like 10 examples. So it was, you know, cool to hear some more from you as well. So, and then, you know, kind of related topic to that is, you know, at Pango, I can imagine that it's probably like a lower as far as like the AOV kind of compared to like Reverb, for instance. So, you know, how does that kind of, you know, impact the, as far as the economics of the marketplace and even maybe even like how you approach marketing and growth?

Good question. So yeah, so Reverb had around a $300 AOV with around us about maybe a 7% take rate. Bingo has a $12, or actually we're up to a $15 AOV with a 20% take rate. So higher take rate, smaller transactions, much smaller transactions. The biggest point of friction in the Pango world is it's a purely peer-to-peer marketplace. I don't have professional sellers. So you're literally buying from individuals. This is both the strength and a weakness. You're buying from individuals and you're buying from their inventory. Unfortunately, if you're looking for 3 books, the likelihood that they have those 3 books is pretty slim. So therefore the shipping cost using USPS is a friction point. The $4 or $5 for the first 2 pounds is a friction point getting people to buy multiple books across vendors, right? So solving for that is probably our biggest challenge and our biggest opportunity of getting more sellers, greater inventory. So you're literally able to buy more from, from a seller and figuring that out from a marketing point of view. We do spend, I'm using a similar strategy with Google AdWords. Now I'm actually deeper into the, the Instagram product marketing using our merchant feed and Instagram, getting really good conversions on that as well as on AdWords. Words, we are an app-first model. Reverb was more of a web-first, bigger AOV. People still want to buy on a computer, smaller AOV, they're comfortable buying more on an app. Repengo is almost 70% app usage, 30% web and web mobile usage. Hence, the goal is to use POAs and Instagram catalog marketing to get your first transaction to get you to be an app user, to be a seller and a buyer. The other thing that we learned at Reverb is buyers, buyers that become sellers are way more valuable than buyers alone, right? Buyers become sellers, become committed to the platform in ways that are hard, that are hard to repeat, right? They feel connected as a seller and therefore they're a buyer. So building that sort of two-sided market is your core users and those core users stick with you And they're way more loyal and they have greater retention. The other aspect of that marketing of that thing is that when you create a community like we didn't like make this happen at Pango. And I'm not sure how Etsy made it happen with the notes and the cute wrapping and all that, but it's happening at Pango and it's a community thing, and people are actually willing to pay a little bit more for the shipping, knowing that they're buying from an individual rather than some large conglomerate and they feel this connection of human inner, human commerce. And Etsy tries to sell that as well. This idea that you're literally buying from another human being or you're sharing a book, you're sharing your literary desires with someone else who has similar things, right? Those are the kind of things in an AI world, those human connections, that sort of like community building, those are the things that I believe are gonna win in the AI world, right? Leverage AI to like build and scale your business and, and, and do it profitably. But like figure out those touchpoints that like will scale well way beyond something that some AI model could go build a book marketplace that's very lacking in sort of human personality and, and elements like that.

No, certainly you mentioned lots of topics that we're gonna, I could easily kind of, you know, go down the rabbit hole and get super deep into. You know, one though that I did kind of want to, you know, zoom out and you mentioned as far as, you know, the individual kind of sellers more on Pango versus, you know, more of the kind of resellers and even businesses on Reverb, you know, so as far as like when it comes to, you know, resale marketplaces and advising earlier stage founders, you know, how do you advise them think about those like initial kind of supply strategies and like who to focus on? Because I feel like that, you know, has a significant impact on the marketplace as well.

Yeah, I mean, there's not a right or wrong In that, in this case, it really depends on sort of where your passions lie. I do believe that, like, I personally, OptionsXpress, retail investors. I tried doing institutional business and never loved it. Reverb, I owned a guitar store. Like you would think, actually, let me go back a step. My first software business, my family owned a travel agency. I learned that business at the age of 8. I started building software for travel agencies, I realized my family's business was going to be out of business in the next 3 to 5 years. It didn't stop me from building software. When I started Options Express, like it was online brokerages were going to replace brokers. So like this whole idea of, of thinking about, you know, where your business is going And these trends are really, really important. I lost my thought there for a sec, Mike. It was—

I think what I was trying to ask about is, you know, it sounds like the kind of supply strategy for Pango is more like individual sellers, whereas Reverb, you know, is more kind of like, you know, resellers and then, you know, starting from owning a guitar shop yourself, like the businesses, right? Brick and mortar.

I do believe you gotta follow like So back to the Option Express, institutional versus retail, Reverb, dealer versus individual. So my thoughts on that is I love a lot of littles. I love a lot of little transactions. I hate sales. If you like selling and enterprise and you want to go after like big customers that are going to bring tons of inventory to your platform, God bless you. I hate it. I'm not a salesperson. I'm not a SaaS enterprise salesperson. I want to build as many little transactions as possible. And if one customer leaves, I could care less. And if that customer is like picky and they need these 5 features or else they're gone, that isn't critical to how I think about business. So that works for my personality. And that's not right or wrong. That's just David's opinion. And David, knowing David well enough in my background to do that, I challenge you all to do the same thing. If you don't like lots of little customers and customer service that has to like deal with little $3 ACH transfers and people complaining about, you know, a little button that's, you know, 2 pixels too big, then you shouldn't get into a retail-oriented peer-to-peer type marketplace. Like, to me, that's like an obvious thing, but for most people it isn't. They just are trying to chase money. And to me, I've always focused on what I'm, what I'm going to— willing to stay up all night working on. And to me, I'm willing to stay up all working on— when I see 10,000 people on my site at one time, I'm like, that motivates me. If I had to go get one big customer to like pay, do the payroll, that would make me vomit. Like, I couldn't deal with that.

No, I really like how you're emphasizing, you know, being authentic to yourself as a founder, right? So I think especially with marketplaces, they can get very complex. And a lot of times we try to overcomplicate them and look at others for examples and, you know, kind of lose sight of that sometimes. So, well, cool. So I know we could, we could probably go on with some of these topics and I promised founders we'd save time for the group Q&A. But, you know, right before we do, do you have a few tips maybe for our early stage founders? I know you shared, you know, a few, quite a few along the way, but maybe that you'd like to share before we jump into the group Q&A.

Yeah, I, and I do talk to a lot of marketplace founders. I had a great conversation with a woman in Toronto last week and she had built a really cool marketplace and I was excited for her and I was like, I'm like, you need to get into the US market. Like, Canada is just not big enough, like for what you're doing. And it was hairstylists and it was really cool and it was a peer-to-peer thing and she had really momentum. I'm like, get into the US. So the advice can vary across marketplaces. The one thing that I stressed at the beginning is the resilience. I really do believe that these businesses take time. And when I say resilience, if there's not good signal, don't keep throwing money at it, like move on. But if you have good signal, but it's not great signal, signal, I'll tell you one other story. When I, Reverb 2013, one year into it, I had a call with Bessemer, big VC firm, and they were really excited because at one year we were doing $2 million in GMS with very little marketing. Year 2, we did $35 million in GMS. Year 3, we did $120 million. They kept like tripling. But after that first year I shared, they, they were very interested in making an investment. So I spent the time with them and I shared my cohort data with them after one year and they had the audacity to come back to me and say, your cohorts don't look that great after one year. It was like a 20% retention, which I thought, we thought was actually pretty good, but we didn't know, we didn't know enough about marketplaces and marketplaces were still very, let's say goods marketplaces relative to eBay were pretty naive there. They passed on the retention and our retention ended up being some of the best retention in goods marketplace history. And I remind that to people that you're dealing with these, you know, institutional investors that will literally tell you what, like they think your data is telling them at that early signal. And it was so off. So to, that's back to the advice is resilience, resilience, resilience. If you have signal, if you are— customers are coming back, then you gotta figure out how to beg, borrow, and steal to keep getting to the point where it makes sense to raise money. It rarely makes sense to raise money in those first 2 years. You need to beg, borrow, and steal to get to the point where it makes sense to raise money.

That's gonna be a real, really helpful tip. So I guess, you know, to expand more on the, on the note of resilience, do you maybe have like a story, you know, from that, from the early stages or the early days at Reverb? of resilience and kind of what it took to give us a kind of insight into that?

Well, one, that Bessemer story is a great story because that's just motivation. Like, great smart investors who, you know, really does like, okay, I'll show them. Like, those are the kind of things I, even at OptionsXpress, I mean, we only raised $2 million and most of the people that turned us down were like supposedly really smart investors. And we were like, okay, we were Barron's Best Online Broker 5 years in a row, beating out Schwab, E-Trade, and Ameritrade, the 3 of the biggest brokerage firms, Right? So the resilience is not focusing on the money raised or the, or appeasing investors. The resilience is focused on wowing customers and creating raving fans. I read a really inspiring post on LinkedIn yesterday just talking about the AI and how building marketplaces or building product is going to become very, very easy in an AI-driven world. But building community and building customer connections. So like for me, the resilience also is looking at the testimonials, going into Intercom or Zendesk and looking at the customer experiences and saying, wow, you actually care. You have a customer-driven thing. The other thing I haven't mentioned is that all of these organizations, OptionsXpress, Reverb, and Pango, I invest heavily in customer experience and I go in and respond to customers and we all do. It's part of the culture. Customers want to know that someone cares and that someone is there. And they can tell that by the customer service. They can tell that by the product design. They can tell that by the marketing initiatives and that they're human marketing initiatives, the emails, the SMS strategies, whatever it is. So to me, the resilience is spending the time and having and enjoying spending the time enhancing the customer experience. And it's not a hire. It's the CEO founder, period. If you don't enjoy doing that, you shouldn't be doing marketplace work because it is, the devils are in the details. The great marketplaces, you know, Brian Chesky, love that guy. And Chesky is like, you know, to me, like there's still proof that people, you know, he's a descendant of the Steve Jobs generation just with sort of a different style about him. But like, that's the, the model that I grew up with in software and tech, the people that actually cared about product and they're obsessed about it to the point where customers notice. And that's really, really the goal.

No, that's great. I'm glad, glad you bring that up and help emphasize it for us here. So, well, cool. We're gonna go ahead and get into the group Q&A. Hey, Peter, do you want to come on? Saw you raise your hand.

Hi, I'm Peter. I'm build— I'm building Recast, which is— I came up with the name before I heard of Reverb. So, and we are a marketplace that specializes in, uh, in reselling fishing gear. Um, there's a very big enthusiastic, uh, fishing community out there. So I was wondering, how did you create your community? Is it— are you talking about the community inside the platform, or did you also create communities outside of your platform? And then the second step of that question is, what did you learn from building communities in Reverb that you're trying to replicate in Bingo Books?

Yeah, I think we tried to build community in Reverb. It's a really interesting question because the community is this idea that it was an industry, and I'm sure it's similar in fishing, where there's a lot of intimidating— let's call them power— like people, people that have real knowledge in the, in the music space, in the space. It's in the vintage space, and there's nomenclature just like there is in fly fishing and stuff that if you don't know it, you feel left out, right? And it was the kind of thing, if you went into even my guitar store in Chicago that the term they would use, oh, it's a burst, oh, does it have PAFs? Like, you know, within 30 seconds you were like, if you didn't know, you didn't know, and you were left out. Well, I didn't like that. One, after buying the guitar store, I didn't like that sort of snobbery that existed in terms of if you know, you know. So we took an approach that, like, that, um, let's explain this stuff. And Dan, once again pointing to Dan, like, it was part of our mission is to actually The folklore that existed in our industry was amazing and the history was great. Why not make it easy for more people to understand what is a PAF, a patent applied for pickup? What's so special about it? So we spent a lot of time really demystifying and we do that at Pango and we not as much at Pango because at Pango, the challenge is the literature world is so massive. And the interesting thing with Pango, and I didn't mention this, is our demographic is women in their 20s, romance and fantasy. It's it. And they want physical books. And BookTok was part of the inspiration that got people to do book reviews and do that. And they want physical books and they're promoting their books. And so we have this amazing community. I believe it's gonna go beyond that, cookbooks and, and, and nonfiction and lots of other things. But that is our core user. Reverb, we have these used and vintage people that are the core elements of why people buy and sell high-end gear. I wanted to grow that audience and sell guitars to people in Japan and anywhere in the world, and I can do that on Reverb. And we really, really emphasize that. So I don't know if I'm totally answering your question, but I'm giving you some— those are some of the elements of community. And the other element of community is customer service, is speak to them in their language Be friendly, remove intimidation, be transparent. Like, those are key community elements that, that resonate.

Awesome. That's a great question and be super helpful for all the consumer marketplaces that are always discussing community-driven marketplaces here. So cool. So I actually had another question that I wanted to ask for Brendan sent me in advance, and he was mentioning, you know, resale marketplaces can often struggle with unstructured inventory. And yet at kind of a Reverb and a Pango, you know, you've created more of a structured kind of SKU-based experience that'd be maybe more, you know, typical with like a shopping new experience, you know. So what was the playbook for doing that well? And then he also mentioned, how do you see AI shifting that going forward?

I mean, the advantage at Pango is we do have ISBNs, which is an industry that has some structured data. So clearly we can lean into that. ISBNs aren't the same as titles and there are lots of books that don't have ISBNs and figuring out how to discover and find your book still requires a lot of really good database and search technology. And, you know, understanding the distinction between a structured database and a good search to me is like the other, the other discovery that a lot of marketplaces either miss or don't spend the time on very early on at Reverb. And Reverb, on the other hand, And there's no— there's very unstructured taxonomy. And that was another one of Dan's drives is how to bring structured taxonomy to an unstructured world. There was no price guide, no active way to compare similar models that had different finishes. Every year the models would change. So we started to build that. It was a— it's a 10, 15, 20-year journey to build that thing, but you gotta start somewhere. And so building that sort of taxonomy early on But I do want to emphasize the search thing is important because I remember in 2013, our CTO or our chief product officer at the time, Dan, another Dan, is he, he said, no, we need to go to Elasticsearch. And this is way sooner than I thought. I mean, we were only doing $2 million in GMS. We maybe had, I don't know, a couple hundred thousand listings. And we started using Elasticsearch really early on, and that made a huge difference. And the same thing with Pango, like having a separate search platform and dedicating all this search intelligence that you can build to get people to find the product, especially early on, and scaling that is really, really important. And not just relying on a SQL database or MongoDB or any of that, like really investing the time in search and making sure that you have good search and that you have, you know, multidimensional search. Like you can search for authors, you can search for people that are selling it. In the marketplace world, there's this other dynamic. Are they buying the product Or are they buying from one of your sellers? Like we're finding at Pango that actually we are, people are following our sellers and they're leaning into, we have a huge part of our community that leans into who they wanna buy from before they're necessarily going and just doing a broad search. We didn't have as much of that at Reverb, but that's a good thing to have where you're this sort of mall as opposed to this sort of like warehouse, right?

Yeah.

It's kind of a different approach.

That's really interesting to learn more about. So, hey, Paulo, Joe, want to come on?

Hey, Mike. Yeah, how's it going, David? Thank you so much for setting this up, Mike. Awesome. And great listening to you guys. I run a collectibles marketplace. It's focused on actually trading collectibles, so no buying and selling. It's literally just sending one item for another item on our marketplace. Pretty tough to kind of introduce liquidity in a market like this, of course, in a marketplace like this. One of the questions that I had was I really enjoyed what you were saying about investors and them giving us advice. And we did take money early on and we are getting a lot of advice. And some of the advice that we're getting is based off of authentication. They want us to middleman. So technically taking a product, putting it in our logistical warehouse, et cetera. And I'm kind of very against that because I believe that there could be authentication in different ways. You could build that authentication different ways. And I want to know, maybe did you have any problems with kind of setting up that trust and authentication in the early days? And how did you kind of go about building that via like maybe KYC or something like that?

That's great. That's a great question. First of all, I spent a lot of time with the early sneaker marketplaces in 2013, '14, GOAT and StockX. And I met with— I'm from Detroit originally, so I met with the StockX team and I was like, you mean you touch every sneaker and authenticate it? This is ridiculous. I'm selling $3,000 guitars. No one's touching them. Like, and I learned that really on my investors. My friends and family investors were like, well, how are you gonna authenticate this? Isn't there a lot of fraud? And yeah, there is plenty of fraud in the musical instruments category. I was gonna solve it by actually guaranteeing that we would give you your money back if you bought a fraudulent guitar. If I'm gonna take that risk, I better have really good listing standards on the front end. And this is before machine learning models. And Dan knows firsthand, like we were literally monitoring, reviewing, removing a bad image was a bad image. Like you could tell if people were stealing images from other stock photo. But what made our product unique was the image and the photography and the description. So over time, we started obviously using machine learning and today with all the AI tools, it would be like a walk in the park. But like being able to quasi-authenticate first and removing the crap. Like, that's like table stakes, right? Well, then the next level is, is can— where is the human intelligence and AI tools that you're using to like really further refine that? And we do this with Pango. I mean, we have people frauding us at Pango for like $12, literally creating fake books, selling it to themselves, doing one legitimate transaction. We're paying them out so that they can then go do fake transactions with fake shipping, USPS, things so they can get $12 from us maybe 10 times. And, and, and so yeah, the frauds exist everywhere. Um, I do think leaning into not authenticating is critical. I don't know exactly what kind of collectibles you're selling and when it does make sense for people to touch them, inspect them, and look at them. Clearly, I'd say the StockX and the GOAT model work Because eBay wasn't offering that, and it was a huge win. I think if you look at the TCG player model, I think you'll see a different scenario where that solved bundling. It didn't necessarily solve authentication. These are people that wanted to buy multiple cards at a time, and therefore having them in a fulfillment house solved more of a shipping friction than an authentication problem. Um, so I'm trying to answer your question lightly without like saying like there is one reason why you would authenticate or not. But if anyone looked at Reverb and saw $3,000, $10,000, $50,000 guitars, they would have assumed that we were somehow authenticating it, and we weren't. We were authenticating it with our knowledge, with our intelligence, and with our due diligence, but we didn't touch the product. We also were authenticating by kicking out bad sellers. And promoting good sellers, right? And like, that's it. If you think you're just gonna create a mall and let anyone show up and sell anything, well, then you're not really authenticating, right? If you're really gonna stand behind the people that are on your platform and in your mall, then you are authenticating. Do you need to touch it? I don't know. That's just another level of trust.

Yeah, I feel like on the topic of fraud, we could probably have an entire off-the-record conversation on So cool. Hey, Adam, do you want to come on? We'll try to get in one last question if we can.

Yeah, awesome.

Yeah, I'm Adam.

I run an outdoor gear rental marketplace. Great chat today, David. So I was just curious, you know, you spoke briefly about the low AOV value that you have on the marketplace for Pango, and I'm curious how that sort of affects or, you know, influences the strategy that you have when you're thinking about scaling up the supply side of your platform. And you've mentioned wanting to have a lot of small guys on the platform and that you'd prefer that. Well, if you're selling one book for $15, that's a lot smaller of an upside for a seller. And so I would imagine with a low AOV, you have to be rapidly onboarding these people at scale. And so I'm really curious how you think about that AOV perspective in terms of wanting to get tons of these small types of suppliers on the platform and what kind of strategies you guys are using and what you'd recommend?

Yeah, it kind of falls back into like what I was mentioning about how Reverb was enabling people to have their own guitar stores without having to go buy them. And it's the same thing with Tango is these are people that are like, oh, I want to sell some books for hobby. For fun. And that's why we have a higher take rate and we're able to have a higher take rate because they're less worried about the money. They're more— we're— they're more focused on, I sold a book. I bought a new book at Barnes& Noble for $25. I just got $12 for it. I now have $12 to go buy my next book. And these are people that love to read. The other thing that we added, and we had this at Reverb, Etsy took it away and we have it at Pango. And I'm a huge believer in a lot of marketplaces don't do this because it's hard to build and it's not hard to build. It's just more complicated. It's something that if you have buyers and sellers and you kind of do, you have renters, people that have product and people that use product is Bucks is keeping your money on the platform, allowing people to become both sides on the supply and the buy side. And we have that at Pango and Pango Bucks so that sellers are now generating money and they're using that money to buy books on Pango. So we're, what we're doing is we're kind of fulfilling this desire of generating some extra cash and then using that cash as my kitty. And people did this in the music, in the music musicians. What they sometimes had to do is they, they sometimes had to hide their musical instrument purchases from their spouse because it was a really, we called it GAS, gear acquisition syndrome. It was a real obsession and people had it. I, I'm experiencing the same thing with book people. These, these people just love books. They love physical books and they want more of them. So this idea of creating some sort of budget where I had to sell 3 to go get 1, I think is a really powerful paradigm. And so I'm seeing that work in Pango. That's how I work on the AOV. As far as like customer represent— yeah, like the onboarding is automatic. We onboard, you know, 800 new users a day, mostly on the app. And then the number of new users that sell a day, I'm not exactly certain on that, maybe 20% of that, you know. So a few hundred, a couple hundred people are brand new sellers And honestly, some of them go and list 20, 30 books and they don't sell a book for 2, 3 weeks. Are they upset? Maybe. Do, did their expectation change? But then some people list 20 books and they sell 2 within a day and they're like ecstatic, right? So I can't really control that. The other thing that's unique about the book thing is this, that probably around 50% of our inventory has zero bid. A John Grisham book from the '90s is worth nothing. Not a penny, not a dime, not a quarter. It's worth zero. And someone just listed it. Unfortunately, there's nothing I can do about that. Right? I, I want people to believe that they have something that's a physical book that has value, but it has no value. But then they could list a book that's worth $15, $20 because it just came out. There's limited inventory. There's limited used inventory, and we can dominate it. So I do believe it's a numbers game, and you focus so much on, ah, this is going to be a great lister, I'm going to get great inventory from them. You really have to build the systems and the process where people want to— they want to list on your platform. It's easy, it's fun, it's engaging. Back to that product experience, back to why people use Airbnb and they want to host. Like, Airbnb spends a lot of time making that hosting experience feel good. And I believe that that should be the motivation and that you're not going to please a lot of your inventory producers. You're not. You're going to have the outliers, but you shouldn't try to plan for that. You should almost try once again, like the way we did the tuners, like you should try to build a big haystack of inventory. Now, in your case, you want to make sure these are quality, the products are quality, that the renters are going to have a good experience, that the sellers are going to be there. And you then have to go Do the authentication work to do that, but you shouldn't try to limit it. The the size of the door—the door should be really. This is once again David's opinion: doors should be really wide. Authentication should be really smart. Yeah, totally.

Well, it's a great way to kind of wrap things up here. And last question. So you know, we're almost out of time. But you know, David, really appreciate you taking the time to join us here today for for the group chat. This is a long overdue and. And as I kind of mentioned, you know, towards the, you know, the start, it ended up being like a masterclass kind of packed in on, you know, building resale marketplaces. So, you know, we really appreciate taking the time to join us here today. And, you know, actually, I had one last question for you. And this is kind of my closing question that I like to ask. And it's, if you could go right back to before you, you know, started Reverb and entered the world of marketplaces, you know, what would you tell yourself about them?

Oh, that's a good one. Celebrate the wins and celebrate the sort of early adopters. I mean, the more that we learned about the early adopters, I thought it was going to be— I thought it was gonna be, yeah, but the next level are going to be, you know, really the holy grail. But it wasn't. It was trying to create more people that looked like those early adopters. When you have that signal of what your early adopters are, don't assume that you're gonna all of a sudden become like Amazon and you're gonna, like, solve it for everything. It's like, no, you have to, like, learn from those early adopters and try to try to figure out how to grow them into more of them and not necessarily assuming that those early adopters that they exist out on the street, they actually need to be curated. They need to be educated. They need to be like groomed on your platform to like what you want. And in my case, buyers and sellers, people that are buying and selling from each other peer to peer, like that's what I wanted. And it wasn't, it was the early adopters that gave me the hope. But it's also those early adopters that you want to learn from and you want to build on.

That's gonna be good parting advice for us all here. So time for a quick plug. Where can we keep up with you at?

LinkedIn. I'm pretty accessible on LinkedIn and Pango. Like, I'm in Pango all day.

So awesome. I'll include a link. So you'll definitely get lots of founders coming your way. So we once again really appreciate, you know, taking the time to join us here today. You, of course, have led the way for us all here and built building vertical marketplaces with Reverb. So we really appreciate it. And thanks everyone for the great questions.