Lessons from two decades in two-sided platforms with Casey Winters (Eventbrite)

Please note: automated, so contains several mistakes

[00:00:00] Casey Winters: So what's unique about Eventbrite. Now, 15 years into the business is we actually are starting to drive a lot of acquisition to our supply. And that means we have the potential to change our messaging, our pricing, et cetera, over time to incorporate that acquisition value prop to the supply side and message that network effect.

We are in the process of building, you know, more than we could in the past.

Welcome to two-sided. The marketplace podcast brought to you by Sharetribe.

[00:00:38] Sjoerd Handgraaf: Hello, and welcome to the first episode of the second season of two sided. If you're new here. Welcome. And if you listened to us before, welcome back. And my sincere apologies for taking so long with season two. Let me just say that we have some different priorities here at Sharetribe, but I'm very happy to be back in a puck passing game.

Lots of people have reached out asking about the podcast and a new season. So during this, you know, let's call it a hiatus. So I'm very excited to be back. I think this is going to be another tremendous season and I definitely could not have wished for a stronger first episode, because today we're talking to the amazing Casey Winters, definitely one of the most knowledgeable and experienced two-sided platform people out there.

He started@apartments.com helped grow up to IPO, ready, grew painters to an enormous scale, and is now chief product officer at Eventbrite. And in the meantime has also advised a gazillion marketplace startups in between. Hip camp fair, Airbnb, Uber GoFund me. It's almost more difficult to name one that he hasn't advised.

This is not the typical founder interview that you're used to. I mean, I was also mostly listening and learning, so it's maybe more like a lecture and maybe lecture is the right word because Casey packed this chat full of knowledge, and you'll probably want to listen to. More than once for me, the most interesting things we covered toward the evolution of marketplaces over the last decade or so, what Casey sees is the different types of marketplace platform models, and which market dynamics drives the choice for each of them.

And then perhaps the coolest one is Evenbrite transitioned from a SaaS to a SaaS like network marketplace, and how they are now seeing a shift in what they offer to both sides of the. Finally another cool last point I think was how to capture liquidity on low frequency marketplaces. So I'd say empty your brain for a moment.

You know, whether you're walking or driving or whatever, and let your empty brain be filled with the marketplace wisdom of Casey winter.

Hi, Casey. Welcome to the podcast. Hey, thanks for having me. Thanks a lot for coming on. I've been a fan for many years already. I don't usually use the word fan for like business related stuff, but I think this is, this is fan level stuff. I obviously know who you are. And even though you're well known really the industry and probably many of the listeners do know us.

I think there's other, quite many who don't know you. So could you give us a little bit context about yourself, about who you are and then take us through your working history regarding marketplaces?

[00:03:22] Casey Winters: Yeah. Sure. Thanks. So I'm currently the chief product officer at. I've been working at Eventbrite for about three years and I manage the product management and product design research and growth marketing teams.

But I started my career as an analyst@apartments.com and it was my job to measure everything we did to drive demand for effectiveness. And it was a really good way to not only learn about marketplaces. But it's good way to learn about all these channels that you use to drive the traffic on the internet, which were, you know, very different from what I learned about in school at the time.

So I naturally started working on optimizing a lot of those channels once I understood how to measure them. These were things like search engine optimization, AdWords, affiliate marketing. I then started trying to work more on the actual product experience to improve conversion and understand what could make our experience stronger compared to competitors, going to focus groups, et cetera.

And it was@thattimethatapartments.com told me I was this marketing and product. And they didn't really know what to do with that. As those were entirely separate departments, things were pretty siloed back then. So I worked on a lot of intrepreneurial projects where that combo is a feature, not a bug.

And one of those was for real estate competitor to Zillow and Trulia, where I joined prelaunch. It went all the online marketing analytics and the. About a year and a half. And so that was 2008. It turned out to not be such a great time to work on real estate. So I joined grub hub after their 1 million series a as their first marketing person.

And did you

[00:04:53] Sjoerd Handgraaf: already sort of like@apartments.com and when you joined GrubHub, did you already sort of realize, Hey, this is like the same model, like, was that already a. Oh,

[00:05:02] Casey Winters: absolutely. So apartments.com was this a kind of crash course in these new internet business models and how, you know, they don't have inventory and they're extremely light on OPEX and, and super profitable.

So the founders grew up up, we're actually both engineers. At the company I was at before. I didn't know them, but you know, we had mutual friends, so they were, they were building something, you know, very similar. And it felt like even though it was a early stage company and you know, it probably start come, it was like 300 people that.

Oh, I recognize this model. I understand how it works. And I think I know how to grow it. Versus if I joined, you know, an enterprise SAS company that would've been a very different problem. I'd have had to learn a lot of things from scratch.

[00:05:44] Sjoerd Handgraaf: Got it. All right. So you joined GrubHub?

[00:05:47] Casey Winters: Yeah, so I was a employee number 15 first marketing person.

And unlike from stock. They didn't care. What I worked on, as long as I grew the demand side of the business. So grub hub has, you know, restaurants on the supply side and then people who order food from them on the demand side. So I could change the product. I can run marketing campaigns, whatever needed to drive growth.

So I worked on, you know, building out the SEO strategy, improving conversion rates, you know, new features for the core product. Ad-words email Realty. Out-of-home television, you know, you name it. I tried it to get more people to order food online and we're very successful. We grew it from, you know, a couple of cities when I joined to pretty much the entirety of the U S and I left shortly before growing up, actually iPad.

So it was really great experience. And then I joined Pinterest and by the time I joined Pinterest, they had a name for the. Work I did, which was called growth. So I ended up leading the growth products team at Pinterest. And while I was there, we were able to make, you know, SEO a primary growth channel. We were able to improve conversion rate by five X and double the activation rate to really set us up for a sustainable growth strategy.

And it was around that time that I started advising other companies on growth and scaling. Airbnb and pocket where my first two, two clients. So when I left Pinterest, I just became an executive residence at Greylock, a venture firm, you know, here in the valley and just started advising a lot of their companies on growth and scaling.

And you know, what I realized there is. For most of these startups, you know, it's their first time building a company. It's their first time working on a marketplace and they're not sure what's normal. What's totally broken that they have to fix. Now it's totally broken that they can wait five years to fix.

So it's about that time. I started taking advising a lot more seriously and took on a bunch of my own. Clients have been great, some tack and Tinder or my first couple of clients. And then I added, you know, Reddit and Canva and, you know, Maccari hip camp, uh, mostly marketplace companies. And I just realized that, you know, being lucky enough to have gone through two startups that have gone on to scale and be worth billions and IPO, that I could help them.

You know, no matter where their problem was, you know, if it's around hiring, if it's around fundraising, if it's around, you know, just growth, you know, obviously. So I not only started doing that, but I started working with this company called Reforge to build out, you know, teaching programs that can help scale kind of the teaching to their teams on, on things like products and growth, et cetera.

So I was doing that for a couple of years and I was really happy doing. I really enjoyed the culture and the problems that have been bright. So when they opened up the chief product officer role, I took it and yeah, now it's been three years. All right.

[00:08:26] Sjoerd Handgraaf: Yeah. That is a fantastic journey. Like you're one of the few people who, I don't know how many people there are actually, who've been able to, like you said, like, get marketplace are two sided platform companies, two of them actually to this amazing, amazing size.

Because like often I'm trying to look for success. There is. And even though there's a lot of like interested in marketplaces, there are actually like in terms of huge ones, right? So many that like get up and then also stay up for a really long

[00:08:49] Casey Winters: time. I think Richard Barton has like the peak of the industry, their founder at multiple of them.

[00:08:55] Sjoerd Handgraaf: Okay. I should get him on next on. Yeah. Alright, thanks. So that's a huge experience in marketplaces. How have you seen them evolve over the past? Like 10 to 15 years in terms of business models, monetization models, and then markets, they are.

[00:09:10] Casey Winters: Yeah. So, you know, when I started working in tech marketplaces where this ultimate lightweight business model, you know, I'd studied a lot of retail in college.

So the idea of like going to apartments.com and not having any inventory and having so little OPEX was just like revolutionary to me. So a lot of these early marketplaces were super lightweight. You know, they connect buyers and sellers and get out of the way, like many of them didn't even manage the payment.

Like apartments.com. We didn't manage the payment flow and neither did our biggest competitor at the time, you know, Craigslist. So, you know, once these companies were created, these network effects, right, where they get stronger over time is to add, you know, more suppliers and more people buy. Had made them super sustainable and really hard to compete with.

Like it felt like they were entrenched. So I think the next generation of marketplaces felt like in order to be 10 X better and supplant that network effect, these marketplaces had to build out a lot more product and service to win. So at first it started by owning the transaction, then, you know, managing risk, which is what we did at grub hub.

And now you've seen that evolve further where, you know, the door dash managed to supplant grub hub and market share by building out a fairly unprofitable delivery network actually facilitate people, getting their food at which meant they could offer it in many more markets, like the suburbs and in the United States.

And now that facilitation of the purchase is, is common in most dominant marketplaces, where it definitely was not, you know, when I started out, but that of course comes with increased costs. So many of these marketplaces have to build out additional business models outside of just that transactional fee to be able to support them, whether that's ads or SAS fees.

And that's certainly something we're doing at Eventbrite. Like we just launched, you know, our SAS products last year and that's going pretty.

[00:10:54] Sjoerd Handgraaf: What SAS product is that actually, cause I tried to look for it, but I couldn't find it. Yeah.

[00:10:57] Casey Winters: It's called Eventbrite boost. So unlike a lot of other marketplaces event, creators are still doing a lot of their own marketing to drive people, to Eventbrite to transact.

And what we realized is, you know, these are small businesses, they're wearing lots of hats, but they're not exactly expert marketers. And we kind of are. So we built a tool just to help them audit. That are marketing to drive people to their events. And that's something that's really resonated with the market and that's something we continue to invest in.

[00:11:24] Sjoerd Handgraaf: So is it things like, can you buy like paid marketing for example?

[00:11:27] Casey Winters: Yeah. So you know, everything from like, you know, Facebook at Facebook, Instagram advertising to also email marketing, which are the two main channels that event creators used. Got

[00:11:36] Sjoerd Handgraaf: it. Got it. Cool. Maybe I'll cut this out. But like, I remember that we discussed this in the reforged class.

That's like, oh, one of the things like, what could Eventbrite do to service them better out? I was like, I dunno, do they have that? I think that was part of discussion. And then you vaguely alluded that this might be something we do in the future and probably it was being developed at the same time. So that is cool.

[00:11:54] Casey Winters: Yeah, we built it and yeah, it's, it seems to be working so far. Excellent.

[00:11:59] Sjoerd Handgraaf: Yeah. And so how about market? So I'd like to discuss like the, you know, the type of business model. Like you said, like moving up more from just facilitating sections are actually handling the transaction, owning the experience. So how have you seen like the markets that, you know, marketplaces are seen sometimes as, as huge disruptive for, so how have you seen that

[00:12:16] Casey Winters: develop?

Yeah, I think on the market side, probably the most exciting transition we've seen more recently is the ability to make the marketplace model working. B2B transactions, fairs company. I advise that's a notable example on the wholesale side of things, but in order to make those work, they're usually much more complex where workflow is as important as a feature as demand, but we're really starting to see a lot more.

These marketplace models be attached to these more complex B2B transactions than just consumer, which I think, you know, for the first decade to, to, to of, of marketplaces with. Yeah.

[00:12:52] Sjoerd Handgraaf: Yeah. Got it. And done it like the future of business models, like, cause you have a great piece on your blog, which you can find it KC accidental.com by the way, for the listeners at home about the sort of the three stages or three types of online marketplaces, which basically ends with owning the supply completely.

And also more reasonably, you wrote this great article together. I hope I say it, right? Like de LA Torre F about this vertical integration as a sort of final stage of marketplaces. Like, could you tell us a little bit about, through your thinking there for those of us who have an expert about the art.

Yeah.

[00:13:23] Casey Winters: Sure. So, you know, as we talked about a little bit earlier, there's a difference between a product that just connects buyers and sellers, like say the Facebook marketplace and is otherwise, you know, buyer beware and a product that processes, transactions and manages risk and fraud and trust like say an Airbnb versus a marketplace that does a lot of work and facilitating the value, like a door dash handling the delivery themselves, or, you know, fair, creating more favorable payment terms and offering free returns.

And then you see some companies like an open door that appear to operate like a marketplace to, to buyers and sellers, but they actually are managing inventory. So like the core elements of a marketplace, which is like, oh, you know, the supply has the inventory. Isn't really true. They're buying the homes directly and then finding people who they can sell them to.

So that obviously has a very different operating expense, a look than a lot of these other marketplace models. I don't think it's that one of them, one of those models is better than another. And that like vertically integrated is like the final form that every market needs to go to. I think there are dynamics of the market that kind of dictate, you know, what the equilibrium of the market should be and what's most sustainable.

So. Food delivery as an example, because it's something I've spent a lot of time on. There are basically. You know, food delivery companies that have been started and scaled in each one of those models. Right. The, you know, the first model. Right. Which is more of just like the listings was like a Yelp, right?

Like, Hey, you can browse, you can find things. And then if you want to actually like go to that restaurant or order food from that restaurant, like, Hey, here's the address. Yeah. And that's how GrubHub started as well before it added on like, And then GrubHub is the first to scale, really with that online ordering functionality, which means we process the transaction.

You know, if something goes wrong with the order, like we take care of it, all that kind of stuff. And then, you know, what Postmates and DoorDash started to do was actually. We actually make the deliveries happen. Like a lot of people don't know this, but like when you order from GrubHub, most of the time, like the restaurant on its own processes, that order and has a delivery driver that they work with that manages it.

Whereas, you know, DoorDash. Postmates. We're like, no, we have the delivery driver and, you know, we'll make sure the food gets there and that meant they didn't necessarily need to work with the restaurant directly. So that allowed them to scale to more and more markets, but obviously a lot more expensive to manage a fleet of, you know, delivery drivers.

And then there were companies, this has kind of died off in the delivery space, but there were companies like. You know, sprayed for example. And they had all of the elements of a marketplace to the consumer, but the reality is like they were just making the food in like a dark kitchen somewhere and they also manage the delivery drivers.

So they didn't really have as much variety on the supply side, but otherwise. The competitor in that market. And I think what we've seen is like the Springs of the world, they haven't really been able to sustain because the market has like this triangle, right. Between price time and quality. Right. So pretty was trying to have a, a good price was trying to get it to you in 10 minutes and have it be quality food.

And the reality is like, you can do all three of those at a profitable level. Right? So they were losing dollars per order, which meant they had to either trade off on quality or trade-off on price or trade off on time. And ultimately those trade-offs meant that they couldn't really find product market fit with consumers.

So what we've seen as the dominant model is not a vertically integrated model, like a Sprig, the dominant model seems to be a, uh, you know, heavily managed marketplace. That's facilitating the delivery of the value, like, like a door to Ashton or an angry.

[00:16:57] Sjoerd Handgraaf: Yeah, that makes sense. Thanks. That was excellent now.

So I never looked at it like that, but the triangle is kind of indeed like the secret on breakable space. You can probably not own, you know, it's like other things would like, you can have something fast, cheap, or good quality like this just, that never gets broken. Well, you started with some marketplaces of considerable size and you send it to advise Uber and Airbnb.

Like, do you see any, you know, not only have seen this sort of like a Margaret place takeover of the world, do you think any, are there any opportunities for these major global marketplace that's still in the size of Uber and Airbnb?

[00:17:30] Casey Winters: Well, I think specifically in those cases with Airbnb and Uber, they pointed out opportunities where supply was artificially constraint due to regulation, and they leveraged consumer sentiment to effectively change regulations.

And that allowed them to unlock these, these fairly large markets. So I think to try to find these major gold marketplace opportunities you were looking for, something in the market could be regulation, could be, could be elsewhere. That's artificially constraining the supply and therefore could be, you know, raising the price.

And you know, where else could you do this? Well, I think the first place I look is anywhere where there is regulation in that cap supply. This could be like licenses in areas like, you know, doctors, you know, countries that have like very high minimum wage laws that might be, you know, taking people out of the market that would be willing to work for cheaper.

But, you know, I have no idea of any of those specific ideas that would work otherwise, you know, I would probably start at the more. Yeah, probably.

[00:18:24] Sjoerd Handgraaf: All right. Yeah. Thanks. That's that's I never thought about those, but indeed like regulation as a sort of constraint indicator is precisely what Uber has been, always like lobby again.

So that makes total sense. And from a more, you know, if we look at a smaller, like non venture-backed marketplace from a more lean standpoint, do you see any opportunities there for marketplaces?

[00:18:42] Casey Winters: I think there's a lot of opportunity there because while we think of these marketplaces as like massive lead global scale, not raising venture, it gives you a lot of opportunities.

You otherwise might not consider it because you don't have to prove how large the addressable market is to anyone to get money and non venture size outcomes can still create a ton of value in these markets and life-changing money, you know, for the employees. So, you know, one of my favorite marketplaces is Diskox, which I don't think has ever really.

Money, but it's not only like the true catalog of like music releases on the internet, but also monetize this through largely like selling vinyl between buyers and sellers. So I think about like, what are those things that might appear too small to be considered a venture backed market? So like, for example, In college, I built a product around the Tony Hawk's pro skater video game.

That's probably too small to be a venture market, but like you could probably build a profitable business there. Now I built a product once around obscure electronic music. Like you can make things like that into successful businesses. Now with the scale of the internet, even if they seem like extremely niche university, like more of a venture back to them.

[00:19:50] Sjoerd Handgraaf: Yeah, that's yet. Or we see loads of dos, like super-nice models. Like I think I had a talk with picking those Boris where to set that, like there, you know, there's a huge, huge class of dose type of ideas that don't cap the, the requirements for venture capital, but are still like, indeed, like life-changing businesses, like in terms of size and this cog.

And I always forget about this because that is technically a, you know, like that has been around for so long that it's just like, no, it's just Discogs. It's its own thing, but yeah, it's like totally a marketplace. Like, yeah, let's talk a little bit about event. So, so we're talking about, you know, Margaret place doesn't may or may not be considered marketplaces if anybody's in marketplace by some definition, but kind of a special one.

How do you currently divine event?

[00:20:32] Casey Winters: So a lot of people would see Eventbrite from the outside and consider it a traditional marketplace with a cross side network effects, just like any other. And that's not really how we see it internally. We think Eventbrite is something we call a SAS like network.

We have buyers and sellers, but we sell a monetization value prop and an efficiency value prop to our supply who are event creators, not an acquisition value profits. You know, we get you more tickets sold. It's usually an acquisition value prop that you see drives those cross site network effects be traditionally associated with marketplaces.

And on the demand side, we mostly historically have sold an ease of use value prop. Like you can buy the ticket easily, and we're not going to charge you an insane amount of fees. And usually it's a discovery value prop on the demand side that drives those, those cross site network effects. So. We've seen a number of these companies emerging last decade that connects buyers and sellers like Eventbrite, but the seller is primarily there for monetization and the seller feels like it's responsible to bring the buyers directly to the marketplace instead of the marketplace doing that themselves.

So you can think about, you know, square Patrion, sub stack, GoFund me Kickstarter. They're all examples of. And the root dynamics are very different from these other marketplace models. You know, even if they look similar from the outside. So what's unique about Eventbrite now, 15 years into the business is we actually are starting to drive a lot of acquisition to our supply.

And that means we have the potential to change our messaging, our pricing, et cetera, over time to incorporate that acquisition value prop to the supply side and message that network effect, we are in the process of building, you know, more than we could in the past. So we're in this state of transition from a satellite network to something that, you know, we still are really good at that, but we also are now driving an incredible amount of, uh, acquisition.

And how does that change the business model over time? So it's really exciting. Yeah, that's

[00:22:26] Sjoerd Handgraaf: really cool because like, you know, last season when I talked to Lanny, Rybicki, I'm saying his full name, but I know that you know him Lenny, but I'm like, he has this model where it's like, here's this great blog post, how to kickstart and scale of marketplace, where he goes through the steps and then.

Which side are you constrained on? And like, I think he mentioned that like sometimes it's changes and you know, it could potentially change on Eventbrite that at some point you might be supply constraint if like you become such the dominating way to buy, like, you know, like the people look to you to what to do on the weekend that actually you might need more supply.

Like you might need more events or. Exactly. That's cool. That's a really cool, you know, evolution talking about evolution. Like, so like you said, like if, if I didn't really start as a marketplace, we brought her as a sort of SaaS tool. Could you tell us a little bit about the transition? Because like, well, maybe transition.

It's not like, I think you rightfully pointed out earlier already, or like in your articles, I'd like, it's more like layering something on top, so it's not like, oh, we leave one behind, but could you tell us about, well, let's still call it.

[00:23:25] Casey Winters: Yeah. Sure. So as we think about this transition, like, how is that occurring?

Well, well, the original idea for a membrane was Kevin Hart's. One of the founders was early at PayPal and he wanted to explore other industries where they could make payments easier on the web and they found events, ticketing. So as an event creator, the idea was you sign up, you set up your bank account, create a listing page.

And then you send your audience to Eventbrite, to transact, and then we deposit the money in your bank account minus our fee. You know, how many people are coming, you know, how much to plan, you know, all that kind of stuff. And two amazing things happened from that initial core product experience. One was many of the people who bought tickets ended up creating events on Eventbrite themselves later on.

So this of them probably grew virally through our user generated content, which was events, which I don't think was exactly planned. And the second was that all of our event creators were doing marketing. That meant they were sending people directly to Eventbrite, which means they were sending like links directly to Eventbrite.

And if you know anything about the Google algorithm, you know, that quantity and quality of external links is one of the key ranking signals. So we started getting all of these links from our creators. So then we just started creating pages, showing the inventory of our creators, all these cool events to consume.

So, this is like a San Francisco events page. And these pages started ranking really highly when consumers search for Google for, you know, events or things to do. And this started driving incremental demands to event creators beyond their own marketing efforts. And then, you know, as more and more people started buying tickets, we then had their emails.

We could tell them about other events they might be interested in. We then started building like search and browse functionality. So if you're like, and then bright, like I want to do a cool music show this weekend, we could show you a bunch of other. We then layered in, you know, partnerships with Facebook and Spotify to get more exposure for our inventory.

And all of these just kept increasing the percentage of ticket sales that Eventbrite drove directly versus were coming from the event, creators own marketing efforts. And I'd say we're still in the beginning phases of understanding how this new core competency we built and driving demand works and how we fully leveraged it as a company.

But it's very exciting.

[00:25:40] Sjoerd Handgraaf: Yeah, that does sound very exciting. And it's a great segue into my next question, because, you know, we can't really talk about marketplaces and not mention liquidity or chicken and egg problem demand, supply problems. So, first of all, like how do you usually determine with the marketplace or how would you usually determine what side to focus.

[00:25:57] Casey Winters: Yeah. So I usually try to determine which side with some latency can attract the other side, easiest nine times out of 10. That means you start with attracting supply. First. It's hard to have a good discovery product, which as I said is usually the primary value prop for them. If there's no supply to show them.

So the exception to that rule is when you can show a good discovery experience without that direct relationship with supply, like we talked about, you know, with, with DoorDash and Postmates and the previous example, this is actually how grandpa started to, we grew up on dataset, aggregated every delivery menu in the city, scanned them and put them online.

Thumbtack is another similar example, but even at GrubHub, we started demand first and use that to build an audience to then attract restaurants directly. We eventually figured out. Well, starting with demand. We're starting with supply was actually more efficient and would work faster. So for all of our new markets, we switched to starting with set up lifers.

[00:26:54] Sjoerd Handgraaf: All right. So when, when you do supply, like what are some of the best tactics strategies to succeed?

[00:26:59] Casey Winters: Yeah. So this is an area where venture capital can be very helpful. Like, you know, Uber and Lyft, for example, you know, just pay their drivers, whether they got rides or not. And they're really days to make sure they have supply for writers at grandpa.

We didn't quite have that venture capital footprint that they did. Right. We raised a 1 million series a, which is now a small seed round, but we did find is that we went to restaurants directly. Um, and signed them up. They would give us about four months before they really expected to see a volume of incremental orders.

So this kind of started the clock on my job, you know, as the demand side to really show them incremental orders. So when we started doing this, we clustered where we would sign up supplies so that we can have. And a continental local area to have great SEO and SEM landing pages within a variety to drive decent conversion rate.

And that was our main way to drive demand. So, you know, primarily ad words and SEO, and we found that to be very scalable. So, you know, start with supply, you know, get 50 restaurants in a very dense area. Then go start building, you know, SEO and SEM expertise to make sure those places got orders before the restaurants wanted to.

[00:28:05] Sjoerd Handgraaf: Yeah. And then when you look at demand, right, like you said, it's probably in rare cases, the first one to look at, but if you wouldn't need to look at it, how would you go about that?

[00:28:14] Casey Winters: Yeah, I think you see, in most cases you're either building a listing style products like GrubHub did in its first market or you're scraping or faking that you have supply.

And then when you get demand, you basically send it as a. To try to get that supply, to sign up when there's actually a booking. So, you know, GrubHub built a list of all these delivery restaurants that before it worked with restaurants directly, and it just went to restaurants and said, we have people looking for food, you should sign up for online ordering, you know, Thumbtack had this lead gen flow for demand where he would then just, you know, take any lead of God and then email local suppliers at new in the area.

And that would get them to sign up for Thumbtack initially. I'd say a key thing to remember about ease seeding strategies, whether it's supply or its demand is this is worth, that is not sustainable. And that's okay. Like you're doing a bunch of unsustainable. But the goal is to do the unsustainable work, to unlock liquidity, which is the network effect.

And then the network effect is the more sustainable growth strategy for your business. And you're going to have to feed that network effect as well, but that's really what product market fit is. So you're willing to do all this non-scalable stuff to get to that liquidity or product market fit. And then at that point, you switched to some of the things we talked about is more of the sustainable ProTech.

Yeah, that's

[00:29:30] Sjoerd Handgraaf: great because actually I want to talk about, well, I want us to hook onto what you said earlier about like started small, right? Like you mentioned stomach a lot, get 50 pastry restaurants in the, um, in the region. And that's something that we've heard like a lot actually on the podcast. I think also I haven't had the honor yet to email her, but I think Sarah Tavel has this white hot center metaphor.

And like, um, you, when already told us a little bit about the grovel, but could you tell a little bit more about how you, you know, cause it grow up, started Chicago. Yeah.

[00:29:58] Casey Winters: So really in one neighborhood in Chicago. All right.

[00:30:00] Sjoerd Handgraaf: Yeah. So could you tell us how that goes? And then also, like, how do you make that decision dental?

Like, okay, now, you know, now we move to the next one.

[00:30:07] Casey Winters: Yeah. So I think the main goal, and there's not a lot online about this, but we were trying to understand is the shape of the network effect that you're trying to build. So a ritual is one of these companies I advise and their network of. Is how long you're willing to walk for a coffee or lunch.

That is a very small distance. Whereas compare that to like Etsy, right? As this network effect is wherever it can ship. So like that's pretty global, you know, rituals is hyper-local and then a lot of the companies we've talked about, you know, are more in between like, You know, mainly it's about cars in the area that you are, but you could also use it when you travel.

So there are some more global elements, you know, Airbnb, I mainly care about, does it have places where I'm traveling today, but if it's adding supply in places, I may travel in the future. I'll I'll at some point, you know, get, get a benefit from that. But your only goal is to get an initial network effects to work.

And that's what we call liquidity or product market fit. So you don't want to expand into incremental markets until you have liquidity. In the, you know, the network effect that you decided is the appropriate shape. It could be one market, it could be one category, whatever. And then if you manage to unlock that, which is usually the hardest part, you're trying to understand what created that so that you can replicate it in these other markets and in order for it to be possible for a market or a category to get to liquidity.

Yeah, my belief is the frequency of transaction needs to be somewhat high. So a lot of the examples we've used are in food or transportation, things that you're going to need to do at least once daily. Right? Yeah. You can make up for that. If the value of that transaction is incredible compared to the status quo, like, you know, Airbnb was both cheaper and in a cooler part of town than hotels, right.

And that allowed them to build a successful marketplace despite being low frequency, the Zillow managed to take a low frequency habit of buying a home and turned it into. Keeping track of the value of your home. So, so there are ways to get around the frequency issue, but the companies that can focus their marketplace on one category typically have that high frequency like food, you know, for GrubHub or transport for Uber.

And those that don't frequently need to be category agnostic to find that initial liquidity like Thumbtack has in a ton of categories, Poshmark, isn't a ton of categories, Upworks and a ton of categories in order to be able to get those network effects to really work. Great.

[00:32:24] Sjoerd Handgraaf: Yeah, that makes a lot of sense.

So shape. What do you mean with shape in this way? Like if that, do you mean

[00:32:29] Casey Winters: like how global versus local the network effect is and what is enough inventory to create a high-frequency experience so that he can get via buyers and sellers coming back on a, on a regular basis, you know, They're not going to remember you the next time they have that, that need, right.

So in the case of Thumbtack, well, if you hopefully only get a wedding DJ, once in your lifetime, you're not going to be building like a high-frequency relationship and Thumbtack, but if they also do you know, everything around your house, or, you know, lots of random chores, you can build up a frequency of using Thumbtack for lots of different things.

Yeah, that's

[00:33:08] Sjoerd Handgraaf: actually talking about frequency just on it to go on a slight sidetrack. How do you think about like, low-frequency marketplaces stem, you know, like while you mentioned CLL earlier, I believe that like, you know, like you're not going to buy a house every, like how do you go about determining liquidity then?

Yeah.

[00:33:24] Casey Winters: So the main issue you have with. Frequency marketplaces is says it's hard for you to own demand, which is typically the dominant strategy of marketplace. If you own, the demand supply will always come to you to get access to those bookings or orders or whatever. So it's really hard to, it's really hard to own demand.

As a low-frequency marketplace, you could build liquidity, but a lot of times you have to frequently keep buying the demand, which is obviously a lot less profitable. So the strategies that low-frequency marketplaces use to get around that constraint are either they have dominant distribution with where their demand is.

So even if they don't own it, they're very well-represented with whatever owns demand. So commonly that's Google, right? So take apartments.com. We're super, super low frequency marketplace. Dominant SEO. So if we know you're going to go to Google when you have the need, and we're always going to rank, number one, it's almost as good as us owning demand, you know, directly, not quite but close, right.

But it doesn't have to be Google, right? Like your demand might be on Reddit and you have a great presence there, or your demand might be, you know, in discord, right. As long as you have access to that demand that can make up for the fact that you don't really own it. And they think of you directly. So I'd say that's the most common strategy, low frequency marketplaces.

And then we talked about the Airbnb example earlier. Even though they're low frequency, they were just like 10 X better than everything else you could do. And that created a really strong brand perception, which created loyalty even with the frequency. And Airbnb's been able to maintain that over time, even as it isn't so much cheaper than a hotel anymore, you know, things like that.

It started to degrade. Now the Grant's been able to, uh, to keep those. Loyal and engaged. So that's, that's a common one. You see, I think some others that are less frequent, but definitely work when you can make them work is the Zillow example is they took a low frequency transaction and they sequenced a higher frequency engagement product on top of it in their case, this estimates.

And you've seen other marketplaces do this where if the transaction is. The way that they can guarantee they'll get the transaction, which needs to be, you know, higher in value to make up for the lack of frequency it's that they have some sort of way to keep you in network so that when you need to transact, you're definitely going to do it with them.

And I think, you know, Zillow is of course the best example of that. And then the last is crafting your products as a bit of an insurance plan. So that even if the need for it is rare, the need for it is super intense when it occurs. So you want to make sure, you know, you have the product when it means, so this is how most insurance products works hotel.

Tonight's a good example of this outside of insurance, where yeah. I would keep it on my phone in case I got stuffed somewhere in business travel, right. Then I hope to not get stuck. I hope to never use it, but like, it's going to stay on my phone and it's going to stay top of mind for when that a use case occurs and that use case has occurred and I've used HotelTonight and in those instances, so those are some four that I feel like are really good strategies to kind of keep a close relationship with demand.

Whereas usually because of low frequency, you can own.

[00:36:31] Sjoerd Handgraaf: Yeah, thanks. That was a great answer for it. Like low frequency marketplace. Like we see that more and more, especially like B2B where there's a lot more of those kind of things coming, but those four points are great. Micah point. If I remember correctly, you have a, also a blog post on that and that caseyaccidental.com.

I could pick your brain for another hour about marketplaces and we could go down many, many rabbit holes, but I want to thank you for your time. And thanks for coming on as the last thing. Is there anything, uh, you know, how can people follow? You read what you're writing, anything you would like to. Yeah.

[00:37:00] Casey Winters: Sure. So I write a lot frequently about marketplaces@kcaccidental.com or tweet about them at one case fan and, uh, build programs on product and growth at Reforge. And of course, if you're looking for something fun to do this weekend, there's always a been bright. All right. Thanks a

[00:37:17] Sjoerd Handgraaf: lot, Casey.

[00:37:20] Casey Winters: Thank you for listening to two-sided the marketplace podcast. If you enjoyed today's show, don't forget to subscribe. If you listen on iTunes, we'd also love for you to rate and give us a review. If you got inspired to build your own marketplace, go visit www.sharetribe.com. It's the fastest way to build a successful online marketplace business until next time.