The NFX podcast is about seeing what others do not and getting at true mechanisms behind people and companies than door change in the world. If you enjoyed this episode, let us know by leaving a rating and review and by sharing with friends you think should listen. You can also discover more content, like other episodes, transcripts, essays, and videos by following us on Twitter at n effects, and visiting effects.com. And now on to the show.
So today on the NFX podcast, we've got my Flint, Scott commoners, who's Professor Harvard Business School Scott, so glad you're joining us today. Thanks so much for having me. Yeah. So, look, not everybody who's listened to this knows knows are. And so I wanna explain why I'm having you on here. So I went to HPS years ago, and I didn't know you while you were there because you're younger than I am, but you are teaching the marketplace as course at HBS.
And you're also teaching a doctoral course in Morgan design. And you and I have had so much fun over the last few years talking and geeking out about attributes and tactics around marketplaces and whatnot. And when I look at the landscape of people who are aggregating and thinking deeply about how digital marketplaces work, you are one of the preeminent people who are doing that now. And last few weeks ago, you and I were in Cambridge and we went on a nice long walk we had so much fun talking.
And I just thought, well, I've got to get you in front of founders. You are really the up and coming person who's aggregating and disseminating and analyzing the tactics, the information that founders need to know in order to be more successful with their marketplaces. And you're just a treasure trove at this point and you're only just beginning. And we're gonna see you doing this for the next 40 years, and it's gonna great fun. So glad to have you. Thanks so much.
I don't know if I can live up to that intro, but let me just say it's 100% mutual. Right? Like, I have learned so much about marketplace review and and you've come and spoken to my students and stuff, and I tell them to read every NFX thing they can find because, you know, it's the center of the marketplace universe. Oh, thank you. Yeah. And we're trying to be that.
And you and I, I think, have similar approach to collecting and trying to figure it out because it's just delightful and and fun intellectually to figure this out. And then if you can help founders to do what they're doing. It's just gonna make a big difference in the world and we'll all move forward faster, and that's the joy of it. So I want people to have a chance to learn from you directly today. And last time I came and spoke at that you're class.
I brought my kids, my twin boys, and we got little blazers on them and little ties on them, and we had an absolute blast. So thank you for letting us do that. So let's dive in. So quickly, your Erdo's number is 3. Tell us about the Erdo's number and what's going on there. So the Erdo's number is sort of a historical mathematical amusement.
So there was this mathematician Erdish who was so incredibly prolific that, you know, he basically spent his entire life just like flying around the world collaborating with different mathematicians and writing papers, and he had so many different collaborators that they created, like, a sort of a, you know, a system. It was almost like a game. It's sort of like the Kevin Bacon number equivalent for math, where, you know, people kept track of how far they were in taper distance from Erdish.
And so if you would co author it with him, he was a 0, of course, right? Your Erdish number of 0, he has no distance from himself. If you co authored it with him, you gotta one. If you coauthored with somebody who had coauthored with him, you got a 2. And then if you coauthored with someone who's coauthored with someone who's coauthored with someone, and so forth. So I've coauthored with couple of different people who co authored with someone who co authored with Erdish who have 2.
So I'm a 3. The wacky thing, actually, it's like one of the papers is a number theory paper written with the, the professor who was my undergraduate, thesis advisor and now sort of like a, you know, a very good friend and long term collaborator. But another is with one of my economist co authors. Who through, like, economist channels had co authored with someone else who co authored with Erdish.
So, like, even, like, outside of field, outside of math, it's just like, you know, Kevin Bacon, but the mathematician's version to the stream. Got it. So Erdish is this center academic figure, and you are actually pretty close to him even though he probably passed before you even got going. I mean, it does show that you have really committed your off to this academic track.
You know, I think there are some people who might have decided to do academics for a while and then go into industry and get some equity and do some deals and then go back and that sort of thing. You are solidly on the academic track and in the world of publishing papers, and you've become prolific at You've been publishing lots of papers and you're how old are you? 34 as of recently. Wow. That's fantastic. How many papers have you published do you think? Is it 10? Is it 5? Is it 50?
So it's closer to, like, 80 or 85 or something? Unbelievably. 34 years old. Well done. Well done. So, Pete, this is it. This is to the listeners of the fuck is this is what we've got this guy on. He's, one of the bright lights. So there you go. So you're teaching this making markets class at Pete. Did you invent that? Yeah. Well, so I co launched the course with Tom Eisman, you know, who I know you know and have done work with in the past.
He's been on the unFX podcast before, and he's got this great book. Out recently about entrepreneurship and failure and how to avoid it. Excellent book incidentally. I'll second the recommendation. Excellent. But so he and I co launched the class. And, really, it had been a project we've been talking about co launching since I was in grad school. So I came to grad school, already excited about market design from the academic perspective.
I really, like, like, loved the idea of using economics as a lens for understanding how to actually improve market function in the world. You know, I did my PhD in this program at HBS, and they make you take some MBA courses even as a doctoral student. And so I took all these entrepreneurship classes and was just amazed by how many people in the world, you know, sort of how many entrepreneurs out there were doing market design in practice.
And, like, the academic field of market design has historically been you know, mostly focused on questions of, you know, public policy market designs or, like, you know, the the US spectrum auctions or systems like we used to match doctors to their residences But, like, by mass, it was all entrepreneurs. And so a big part of the project, like, you know, Tom and I took Tom as launching Tech Venture's class, you know, in 1st year, he taught it.
And, you know, we just started chatting about this for years, we were thinking like, so how would we take market design and teach it to MBAs? Yeah. And so many startups actually have marketplace dynamics embedded in them even they don't always think of them as marketplace. Exactly. Yeah. And so Tom had been teaching. What was the name of that course? So Tom launches a new course, like, every 2 years or whatever. That one was launching tech ventures.
So it was about how do you start a technology business? Got it. So you guys started talking about while you're in Grad school and then sort of matriculated. Now you're an associate professor at HPS. And so you said here's the raw outline of it. How did it start? You have, what, 14 classes, 14 cases in a semester. Is that how it works? 24, but close enough. It was really funny, actually. Like, forming a class is a lot like forming a startup. Right?
You sort of have, like, this inkling of an idea, and you're looking for, product market Flint. And as you do so, you're looking for all of the sort of different supporting infrastructure that's gonna make it work. You know, the first challenges we're trying to understand what the real sort of practical problems that marketplace founders face are and then trying to sort of organize them into frameworks. And we came up with a course arc that loosely, you know, goes as follows.
You know, we start with sort of almost classical economics. Like, what are markets for? Like, why do they create value for people, for society. And then in what ways can markets break? Like, you know, how can there, you know, if people really want to transact, How is it possible that they might not or that they might participate in the wrong transactions or might be unhappy after they've done so?
That's markets to market failures, like, the way that markets can fail to achieve sort of their maximum social potential. Mhmm. And then we ask, okay. So if there's a market failure, How can you design a marketplace that addresses it? Right? So what are the ways in which marketplace businesses can mitigate different sources of Morgan Pete.
And then from there, you know, sort of transitions to thinking about, you know, there's like a, you know, discussion of mechanisms, like how you actually you know, do things like decide who should, you know, match with whom, and then how do you, you know, then operations and execution, right, how do you launch scale and manage on an ongoing basis to your marketplace? So it sort of takes you from the, like, foundational questions, like, what are the things that drive the need?
And where do the opportunities in marketplace business come from? How do you design the business to take advantage of those opportunities? And then very tactically, how do you actually run it? Got it.
Yeah. I mean, and then that arc of those 3 things, I mean, when we say marketplaces, in fact, if you wanna stretch it, I mean, all of society is sort of a marketplace, a marketplace of ideas, or you could you could look at the whole venture cap industry and all the startups as a marketplace, if you will, even though it's not centralized on one platform like an eBay or a Craigslist today, but it might be in the future.
But right now, it's not transparent and easily visible as one market, but it really is. And so when you're talking about market dynamics and designing marketplaces, you're really talking about how do you design society in Look, I totally agree. I was once asked, you know, what's the foundational research question of market design as you would describe it to a six year old? You know, I thought for a while, and I said, I think it's How should the world work?
You know, how do we signal to each other? How do we communicate? How do we transact? How do we progress? How do we get what we need? How do we make tradeoffs? Exactly. All those things are then embedded in these marketplace. And each node in a marketplace has to sort of come up with their own decisions but those decisions are based on whatever information is available to them and the marketplaces provide those informations and those opportunities or not.
And I can your second thing, which is, well, where's the failure? And this is where you and I debate a lot because I don't particularly like the word failure because you know, this goes back to my annoyance with a lot of VCs who say, well, your deck has to say what the problem is and what your solution is. And I say, well, that's not always true. Like, some businesses can be described that way as a failure or as a problem, but many others are just an opportunity.
You know, no one said there was a problem that Twitter was fixing or problem that Facebook was fixing. It just was, here's, wouldn't it be awesome if? And so I would kind of look at marketplaces as not necessarily failures, but as no one's thought of it Pete. Or here's something that you didn't even know you could transact. So there's no failure because you didn't wake up in the morning thinking you wanted to transact, and then there was a failure.
It just hadn't occurred to you that you could get somebody who lives down the street from you to give you a ride to work on an Uber or a Lyft. Right? It just hadn't occurred yet. We can debate that later. But so then you come into the 2nd phase about how are things not working as we want them too. And then you get to the point of, well, once you've got something going, how do you fly the plane?
That's kind of the word that I use about marketplaces is, you know, flying these things is really difficult. Because I've built gaming companies. I've built, you know, social networks. I've built, you know, SaaS software.
But man, when we built those marketplaces, the amount of data you had to calculate, the amount of changes you had to make was incredibly, voluminous on the order of running a, you know, a game as a service like a dragons of Atlantis where you have 45,000,000 people using it you've gotta decide you give people more shields or more spheres or you have to change those numbers on a daily basis. The same thing with marketplace is very difficult to fly.
And so that's the 3rd part course, which is what are the tactics you use to actually fly those things? Is that generally how we look at it? Exactly. To that last point for a second, it's sort of very natural with marketplaces that they're very hard to fly because at least in my view, even relative to other forms of entrepreneurship, you're envisioning like, a fundamentally different change in the way people behave. Like, all entrepreneurship changes the world. Right?
That's, you know, sort of a baseline requirement to be a successful entrepreneur. But marketplace entrepreneurship isn't just like exposing people to a new product, giving them a new thing that they that they sort of have an intuition for and wanna use. They're already doing something. They're sort of a current behavior. Or failing to do something. Right? They might just, like, be choosing not to buy a car or So if you're selling all bird shoes, it's got new materials. It's got a new line to it.
Pete like, oh, I'll just replace the old shoes with new shoes, but I don't change my behavior or change my self-concept. Exactly. Whereas a marketplace, you actually have to be envisioning a completely different equilibrium, like a different sort of behavioral environment. With new information inputs, new decision making processes, Exactly. New ways of making money. Or new ways of helping people achieve their needs and wants.
New ways of re ensuring your employees that the business is gonna be okay, even though you're now working in this new marketplace. I mean, all these things are very Morgan, and those sorts of changes are filled with fear for many people. And so having new marketplace dynamics brings all that up and creates all that resistance when something changes. It's exactly right. And the advantages have to be great enough to come that resistance. Right?
Like, I can swap in a new type of shoe because it's clear to me that that shoe is Beller. But if I'm going to, you know, Beller my car through a new platform, I have to believe that buyers will be convinced to join this platform. Right? So I have to sort of foresee this different behavior and imagine that the marketplace is creating enough value for those buyers to bring them on. And they're meanwhile wondering the same thing.
This is sort of like the chicken and egg problem, but rephrased as question of human behavior. Mhmm. And so for teaching this course, you go out and you talk to entrepreneurs and you'll actually write cases on Their particular business. How many cases would you write a year? Varies a lot. In the 1st year, I think Tom and I, between the two of us, we wrote something like 10 or 12. In a given year, I write, you know, 3 or 4 and then, you know, sort of a couple of similar sorts of things.
Like, I recently wrote a note that's of lays out my framework for how you think about mapping your marketplace solution to the, you know, specific Morgan, challenge your opportunity you're trying to address. But it's a lot of fun, and it keeps you on your toes. Right?
You do sort of, like, dive into a company and just start talking to them about how their marketplace works and how they see it and what sort of, you know, what the market dynamics are and how they see their business operations in response to those dynamics, And first of all, like, you know, I mentioned you're sort of pulling frameworks. Frameworks pop out of that. Right?
So one thing that we noticed over and over and over again was that Marketplaces that provide as one of their sort of principle benefits, a quality guarantee, often can provide that you know, much more cheaply than you would think because somehow the presence of the guarantee changes user behavior in a way that means that the guarantee doesn't have to, like, you know, be used very often. Right? Somehow, like, the availability of insurance is causing people to not need it.
And we saw this in the very first case I wrote for which was on GoFundMe. Are there a crowdfunding platform for causes? I know you know it, but, you know, just in case any of the listeners don't, so they're sort of like Kickstarter, but for, you know, raising money for a cause, And unlike Kickstarter, you don't get anything back. Right?
You might Pete, like, a nice thank you note or an update on what's going on with the cause or something of the sort, but there's not a product or something you're expecting to receive. Which creates all sorts of strange potential. Right? You know, you you you could naturally worry that people are gonna have scone for the money or misuse it, but GoFundMe implements a strong guarantee.
They say, like, if we ever discover or anybody discovers that the money and some campaign was misused or not used as stated in the campaign sort of offering and guidance, then we're just gonna refund all of the donors up to some very high threshold. No questions asked, even if the money's already left the door. Right.
If someone's got a campaign for girl who needs treatment for cancer and she doesn't have the money and you give money and then it turns out that they went to Vegas and went on a Beller, then go fund me would give you the money Exactly. Even if it's already been spent. I asked my students right here to pose the question that this came out of the conversation with GoFundMe. Like, I didn't know even to how to ask this question beforehand.
I ask my students, so how much do you think it costs to maintain that guarantee? You know, eventually, what they realize is that, you know, it sounds like very expensive, right? Like, it sounds like Gofoni's got a super carefully vet every individual campaign, and they, you know, run tons of them. You start adding it up and you realize, like, some, first of all, lots of campaigns don't get very big. And so for them, like, you just sort of have, like, a diversified insurance portfolio.
And besides in the smaller campaigns, most of the donors or friends and family or friends and friends and family, so there's a lot of social proof moreover. And this is like the the big behavioral change, you know, it turns out that if there are lots of ways you can raise money on the internet, you know, there's an indiegogo version, you know, for causes as well. And the fact that GoFundMe has this guarantee means they have a really strong incentive to police malfeasance. Right?
It's really important for them to vet campaigns. And so if you were planning to try and defraud your campaign donors, that's not necessarily the place you wanna be. Right? Your optimal strategy to pick a platform that doesn't have the guarantee. And so as a result, go fund me, like, you know, sort of has this behavioral deterrence, right, that the fraudulent campaigns on average actually would prefer to be elsewhere.
So you don't have to do that as many first because of social proof second because a lot are staying away. And then third, so that's the guarantee the insurance value. Like, somehow it turns out that you don't actually have to use the insurance that much. Right? Just having the guarantee reduces your need is that second point. And then there's actually a third one, which is even wackier, which is complementary. Go fund me, actually, to vet the campaigns, doesn't send a group of lawyers. Right?
What they do is they send team, or what are the campaigns you actually need to vet? They're the ones that are getting really, really big and super popular. They go viral. You know, go fund me says, hey, here's our media team. Mhmm. You know, we've got a bunch of experts in promoting campaigns. We notice your campaign is going viral. We wanna double that.
Mhmm. And, you know, they ask, you know, by the way, can we, you know, speak to, you know, the sick individual, like, you know, take some really good photos. And so they have this sort of, like, opportunity to vet and observe for the campaigns that are the biggest insurance risk. Along the way, which again helps deter, you know, the potential malfeasance. So recently, you coauthored, an Pete article, Harvard Business Review article called what makes an online marketplace disruptive.
And you ask this great question. What are industry changing marketplaces doing that others aren't and can those practices be replicated? So What were you saying in this article? Because some marketplaces are disruptive and some are not. Talk us through that. Cool. So first, let me quickly acknowledge my co author, Cliff Maxwell, who is this absolutely brilliant student, you know, this year's MBA class and was previously chief of staff to Clay Christensen.
So he really came into this with a, you know, lens of having worked on disruptive innovation, like, you know, directly with the, the the literal originator of the term. And what we look at in this article is, you know, sort of an attempt to understand when marketplaces have the potential to, you know, create value and, and transactions among other non consumers and non producers. Right?
Like, many marketplaces sort of just like superimpose some form of infrastructure on an existing market and sort of help the people who are doing transactions already, you know, do them more efficiently or better or with, like, you know, more assurance that they're gonna go Beller. Reduce friction or increase transparency or increase insurance or that sort of thing, digitize it. Make it doable from your phone, but it's the same transactions that have already been being done.
You've digitizing the existing transactions. That's one way of going. Totally. And you can create a lot of value there. Right? Like, there are a lot of great businesses built on that. But what they don't unlock is mostly they don't bring in sort of new consumers who would never have been able to participate in the transactions before or new producers, same thing who, like, for whatever reason, couldn't have been sort of in the Morgan, whereas disruptive marketplaces achieve this.
And this really goes back to Clay Christensen's classic theory of disruptive innovation, right? What is disruption? It's building a new solution for non consumers that is sort of good enough for them. People are, you know, aren't buying a product for whatever reason. They weren't buying it because they didn't want it or it wasn't the thing for them at that price. You're saying non consumers, but you don't mean consumers. You mean participants. Non participants. Non participants.
Yeah. Yeah. Sorry. Clayton Christensen's term was non utilizers. Non utilizers. Yes. That one. It's about enabling participation among non participants. And, you know, to do that, Cliff and I argue, a marketplace really has to somehow create a new type of transaction. And, you know, in some sense, it's all it's almost tautological, right? You know, why do you have to create a new type of transaction?
Well, because these people weren't participating in the old transactions So for whatever reason, those transactions must not have been right for them. So at that level, it's a tautology. But if you think about it as a way of understanding what your marketplace could do, there's actually a lot of structure to it. Right? You sort of ask, like, what types of new transactions are there? And So so we have we have sort of 4 categories.
Maybe the easiest one to think about is what we call a, like, a smaller supply unit or or smaller unit transaction. So this thing like Airbnb, right, what was the early Airbnb product? The thing that appealed to the people who were previously not participating in the existing market for travel housing, you know, which was hotels, you know, that was like, I have a room.
You can stay in my room instead of, like, renting an entire room all to yourself in a hotel, you know, with all the services, like, I'm a random person in the city I've got an extra room in my apartment. You know, there's gonna be minimal services. I'm gonna give you a towel, but I'm not gonna give you tours. There won't be a concierge. You know, there won't be housekeeping or whatever. But also, you know, the space is shared.
All of these things that, you know, weren't what a hotel con you know, a hotel user, you know, someone who participated in transactions with hotels was looking for, but when a lot of people who, for whatever reason, couldn't afford to stay in hotels or, like, didn't like the style of them, whatever the case, This was exactly what they wanted. And it's a unit that is smaller than what a hotel could provide, right?
It's sort of carving things that were previously non performing assets extra rooms Pete people's apartments and houses and turning them into rental assets to participate in the same type of transaction that people have been doing, but in a completely new Right? Like, people were looking for travel housing, but a lot of people couldn't get it because the existing option was sort of, like, too big a unit for them to buy. Interesting. But that was such a small market, Scott.
That wasn't really a place because that product wasn't really consumed. What Airbnb found was a new supply. Not a new demand. It's a very similar demand. You're paying 60 to 80%. In some cases, a 140% of what you pay for a hotel room because you want the experience, but what we found was new supply. Wasn't that the breakthrough? So to be fair, you know, another one of our 4 categories is, you know, sort of new supply opportunities.
So it's possible that, you know, and and these things are not exclusive. So it's possible that, like, the smaller unit, like, thing wasn't as the, your view, the 1st order component of this, like, new supply opportunity, I still think there are many ways in which this is a different supply unit from what had previously existed? Well, it's certainly a different supply unit.
I just don't know that we should set the history of Airbnb taking off as an room, because it, it very clearly did not take off as an extra room with an air Beller on in the extra room. It took off as a rent, the whole unit you have key. You're the only one in there. But again, it's a lot easier than running a hotel, and it's like a different, unit context and footnote also in sort of the theory of disruptive innovation like, what is disruption?
Usually, you're targeting, you know, non participants on some dimension the existing market wasn't serving. Right? And so when you speak about, like, you know, the new experience of living, like, in the community or, like, sort of, in the city in whatever way through an Airbnb, that's very much also in line with the disruptive idea. And so Airbnb did multiple types of disruption.
They sort of like started by creating something that an absolutely basic nonconsumer could use, then very quickly moved up Morgan to this, like, much bigger disruptive opportunity, which was bringing on this supply that's easier to provide and and smaller than a hotel and, like, still less overall infrastructure. There's still no concierge typically, although I've stated some pretty nice Airbnb is on occasion.
And then Morgan importantly, and and this is what you're highlighting, I think, they differentiated the experience in a way that in and of itself was appealing, and that suddenly allowed to start sucking in. Right? So in disruptive innovation, like, how is the the disruption actually happens? It's when the product becomes so good that the rest of the market wants it too. Right?
Something that starts out being, you know, a play for the current non participants that's not good enough for the existing participants. Suddenly gets good enough that the existing participants want it also. And that's the moment I think of that you're describing where it takes off. We're like, now this is a substitute for a hotel. Yeah. I mean, with Airbnb, the non participants or all of us who have an apartment and we're not staying in it while we're in the Bahamas. Exactly. Nonproducers.
Nonproducers. We're not participants. We didn't even know that you could take the back house in your yard and rent it out during the Super Bowl in the San Francisco Bay wasn't a thing until Airbnb because you hadn't gotten on VRBO yet. Right. Exactly. And VR VO didn't provide the infrastructure that was needed for that to happen. Got it. So what other examples do we have that were disruptive? You've got new supply. You've got smaller units of supply. So breaking things up into smaller pieces.
Then there's bundles. So strategies where you're taking a bunch of units, you know, from different suppliers or units from different, you know, units of demand together and, like, rolling them into a bundle that is worth enough to sell separately. So here, think about something like class pass, right, you have all these gyms with extra Pete and courses, But their business model is sort of built around selling bigger subscriptions and member, you know, memberships.
So it's not worth it to them to sell the individual units because they risk cannibalizing their own you know, sort of total demand, you know, their main business of memberships. ClassPass sort of tries to take all those units and, like, roll them into a bundle across the businesses, and sell that separately in a way that is sort of non cannibalizing. Right? It's targeting customers who don't actually have demand for the main gym product. Now not cannibalizing.
It turns out it does, in fact, cannibalize. And this is, again, you know, sort of as disruptive innovations move up market, they often do that. Right? They often sort of start displacing the original sort of environment, but that was at least the idea. Like a little like hot wire.
Yeah. Which was from 15 years ago where they would take the receipts on airplanes and say, you can buy it on our website, but you don't know which airline it's on, which allowed those airlines to sell the receipts without cannibalizing their main business on Expedia or through their own websites. Right.
You have to sort of provide, you know, a bundled unit in some form that's not good enough to directly cannibalize the, the main Morgan, at least at the outset, but that is valuable enough money for the, you know, for the operators that they're actually willing to join the platform. Mhmm. Right. To increase their utilization. Exactly. Other fixed assets like an airplane or a class in a gym. And then what's the 4th category? 4th category trust wrappers.
So this is a little bit different from the others, right, although it it plays in with new supply or demand a lot.
The idea here is when there's sort of a a pure failure of trust, you know, for whatever reason, like you don't know who your counterparty is or, you know, you're worried about privacy of the transaction or people learning information about you through the transaction or something, trust wrappers sort of create, like, a certification or, you know, a safe harbor that makes it possible to transact with people sort of in a way that has sort of new forms of trust.
Blockchain do this a lot, right, by enabling you to transact securely with a party you don't necessarily know or to see how all of your information was used. Like, you know, there are all these, experiments with data marketplaces where, you know, every access shows up and is sort of recorded.
And so you can, you know, know who's using your data and how as a result, again, just like in the GoFundMe case we talked about, you know, that's a deterrent to malfeasance because the malfeasance is sort of, like, observed and directly. We also see this. I did a case on this group in Israel. It's the innovation arm of 1 of, Israel's biggest healthcare institutions, like Kabi Healthcare, the Innovation called MacKabitEC, and they've built a virtual research room.
So they have this incredibly rich health data. And we're trying to figure out how can you open this up to researchers to enable people to, you know, both do scientific research on it, you know, learn about, you know, better information about causes of disease and opportunities for treatment but also to entrepreneurs.
How can you make it possible for different people who might have ideas for healthcare startups to work on their data Pete to, you know, develop the technology and sort of identify the opportunities.
And so they built sort of like a virtual research room with privacy technology and sort of synthetic data for experimentation, all of this infrastructure to support people coming in and using the data in a way that wouldn't leak sensitive information and wouldn't, you know, somehow harm that the patients whose data was being used and so forth.
And that's enabled tremendous value on on all of those dimensions, right, there've been a ton of research papers written on this data set and now several successful startups.
And what they needed to do was somehow create a way of accessing the data in a way that all the parties could trust was going to be safe, right, since all sides, it's the researcher's information isn't somehow going to be, you know, ideas somehow aren't gonna stolen, but even more importantly, it's the patient data is not gonna be misused or somehow, like, you know, leak or result in harm. And so I think you use out schools as an example as well. And then Angie's list is not actually disruptive.
Let's compare those 2. Yeah. Okay. So Unfortunately, I tend to think of Angie's list as not disruptive. And again, it doesn't mean it's not super valuable. Right? But what are they doing there? They're like, you know, enabling a much more efficient Morgan for home services. And and that's really good because it's really hard to navigate home services contracting. Right? Like, you know, it's super hard to find somebody to repair some part of your house.
You know, like, who is gonna do a good job and, like, be trustworthy and so forth. You know, creating a system for managing that market is really, really valuable because a lot of people waste a lot of time on it. Right? Now both sides, right, if you're a provider, you spend a lot of time trying to find customers and, like, you know, sort of figure out how to credential yourself.
And and if you're a customer, you spend a lot of time trying to find a provider who got the right skills and the right, you know, sort of, you know, the the right abilities and you can trust to do the job well and so forth. So there's a lot of friction But they're not enabling new participation. Right? They're not somehow bringing in new people who could do home services contracting by and large. Right? Like, you know, is there a little bit of that? Sure. Potentially. Right?
If the main reason you stayed out of this space was because it was, you know, you were afraid of the marketing challenges and building reputation than in a bit, but mostly it's sort of enabling this very, very frictional, like, you know, highly inefficient market to move much more effective I think their market caps around 7,000,000,000, so it's clearly valuable to do that. Oh, yeah. No. No. No. Like, even the extremely successful marketplaces, you know, don't have to be disrupted. Right.
So this is an example of out school being the disruptive one. They've got new supply, new demand, new transactions that just weren't taking place Morgan, and Pete are cited in spreading the word and, and suddenly you have this pretty, like, like a fiverr maybe, where they started out saying everything for $5. And everybody said, well, that's a stupid idea. That's never gonna get big. You know, how many things can you buy for $5? But at least it gave people a focus. It was disruptive.
People didn't know they needed to buy things for $5 getting your name written on a grain of rice for $5. And then as they got bigger, they just said everything $5 and up. They just expanded their Morgan. And now they've also got a $7,000,000,000 market cap. Have you, at this point, got this framework for best practices for what in changing marketplaces are doing and that others aren't.
And can entrepreneurs, can the startup founders that are are listening to us use the framework for idea generation because, look, one of the things we say then affects is the biggest waste in our ecosystem is great people working on bad ideas. And, you know, having been an investor now and been an entrepreneur now for 20 plus years, I've seen 1000 tens of 1000 of ideas, I can categorize your business into this is gonna waste your life's energies.
It's not a good idea, or the smaller group is this is definitely fertile soil. And if you go hard at it, you will have success. And then there's a middle group, which is, I don't know. And I would have put Airbnb in that. Like, I don't know. Will people sleep on an air Beller in someone's room? Maybe. And if and if they do, then that could be an interesting business, but I can't tell you that it's a fertile soil.
Like, I would have gotten that one kind of wrong But if we can move people from bad ideas to, maybe they're good ideas to, they're definitely good ideas. That would be very helpful because then people would waste fewer years working on ideas that are gonna go nowhere. So how does this help founders with idea generation or company pivoting? Cool. Alright. So first, just one footnote on that point about formulating ideas.
I should say that this is particularly important in the marketplace environment because, as I was saying earlier, right, you're envisioning, like, a fundamentally different change in behavior, that often means that, like, the normal, like, testing things we do to find product market Flint, harder to make work, especially really early on. You sort of need a theory of change. You have to sort of Have a model in your head of how what you do is going to change the way the market works.
Yeah. And we we often will see that founders will be working on a marketplace for 4 years before stuff starts to happen before those changes start to make place. And you really start to see Flint off in these marketplaces in year 7 and 8. Exactly. And so You know, the types of time scales you're thinking about mean that you really have to road test the idea in theory before putting it into practice, or you're at or rather as you're putting it into practice, I suppose.
But, yeah, so how does the framework help you spot sort of marketplace opportunities? First of all, it draws attention to this idea of non participants. Right? And it pushes you to get a sense in your head of whether you're building a marketplace for existing participants or non participants because these are really different, right? Like disruptive innovations are about enabling non participation to turn into participation.
Whereas lots of other businesses are about reducing friction, and, you know, sort of enabling the existing transactions to work much better. And these imply, like, really different launch strategies. And you sometimes see founders get it backwards or over So I, like, I think about food delivery marketplaces. Like food delivery, to my mind, is really like in the Angie's list category. It's not in principle, a disruptive innovation, because food is a luxury service.
You just say restaurant food is a luxury. Having drunk food delivered to you sort of at a specific moment in time is a luxury on top of a luxury. Right? And and sorry, at least on the consumer side, it's like it's incredibly sort of like luxury format business, at least in the abstract, there is a little bit of disruptive potential on the supply side I'll talk about in a moment. But if you were operating that business, right, if you know it's a luxury, like, what's your goal?
Your goal is to take the consumers who are already doing the most of it and help them do it better and more cheaply, you know, for you. Right?
You're you're trying to create efficiencies in the Morgan, so it's less expensive to provide delivery services and give those, like, you know, power consumers, the opportunity to consume, like, more, make it easier, make it sort of better for them so that they engage in this luxury more and more off But in practice the way all these delivery service marketplaces launched, I shouldn't say all, many of them launched, was to try super hard to convince everybody to buy delivery food, right?
They like through voucher after voucher, discount after discount, trying to get people hooked on the idea of buying this luxury product, who were not previously consuming. And, like, maybe they were willing to consume at the price point of the discount. But, you know, once the price goes back up, you know, you could only afford to buy delivery food as much as you can afford to buy it.
And, like, you know, you whatever however bad your sushi habit is, right, you know, the cost Pete up and you notice.
And so they had this view going in, I think, that they were trying to that their real goal was to get the people who were not big delivery customers to become big delivery And so this, like, led to this launch strategy where they, like, burned tons of capital on trying to recruit those customers, even though they turned out to not be Meanwhile, on the other side, you know, that capital's gonna comfort some, like, where are you making the money?
They tried to make the money off of the supply side and the restaurants through very high fees. If you think about it, the restaurants were the potential nonproducers Pete, right, like lots of restaurants had no delivery at all because it was too expensive for them to run their own delivery in structure.
If you give them a net of delivery, you know, of existing drivers, technology infrastructure they can use and invest in making that a core part of their business, then suddenly you've brought all of the supply online that didn't exist Morgan, and your value to those consumers, the customers is even higher. And so the disruptive potential in food delivery was really, in my view, on the supply side, and they went the other direction.
They sort of thought they were trying to disrupt on the demand side. And as a result, like, stymied their ability to, you know, sort of really take advantage of the disruptive potential on supply. You really should be attentive to who are the non consumer and nonproducers. Again, consumer meaning everyone, who are your non participants, really? You're saying go after the non participants because they're gonna go from making no revenue. To making some revenue, and that's going to draw them in.
They're gonna have lower friction. They're gonna learn to do their business on your platform. And be much more attached to it and be more loyal to it. And you're gonna get a lot more rapid change from them and rapid growth. Is that sort of one of the things that you're suggesting? Exactly. Or at least that's one strategy. And know that if that's your strategy, that would that's what you should be doing. If your strategy is just, I'm going to be reducing friction.
I'm gonna make this market run better. Adopt that as a strategy you know, but focus then on the customers for whom those improvements are most valuable, who are probably the people who are doing the the transactions the most right now. Right? These are like fundamentally Right? Like, if your goal is just to reduce friction, you wanna focus on the power users, the higher frequency. If your goal is really about enabling non participant to participate, then you're focusing on the non users.
And so there's this big gap and, like, lots of founders sort of try and do the middle at the start. In either case, you hope to move into that middle you're coming from different sites. Yeah. We always say try to find the white hot center. Totally. The white hot center. So as we look at ways of helping founders think about idea generation or helping them move to the white hot centers of the fast moving water so that they can grow more quickly. One thing is looking at non participants.
Another thing is looking at new things that have never been in a marketplace before, like NFTs or throwing parties in second life. If that's 2 frameworks, what other frameworks could we give them? So in a second, we should talk about, you know, compare to advantage and your opportunities as the marketplace operator, because those actually also generate their own set of ideas. But one more thing in this, like, how do you identify your sort of pure opportunities?
You know, my PhD advisor, Al Roth, used to say, a market where you see rules that should never have to be rules has some problems in it. And I'm gonna generalize that to, you know, a market where people are trying to transact. Like, going far out of their way to participate in a transaction that's really, really difficult to do, indicates, you know, real potential value. Crawling on your hands and knees over broken glass. If you see people doing that Exactly. There's something there.
Cause again, right, you know, we talk about, like, lowering the cost of participating in a given type of transaction, right, if there's so much friction that it's unimaginable that anyone would do it in the first place and they're doing it anyway, that suggests there's a real valid reason to help them do it more easily. So then on the comparative advantage side, right, so I think of marketplace design as having 3 principle dimensions. There's curation matching and support.
And curation is you know, structuring the transactions that can happen in your Morgan, matching is, you know, helping determine which ones actually do happen, And support is helping them to happen. Right? You know, so a curation intervention might be, you know, vetting all the suppliers on your platform. And a result of that is that if someone knows if they show up on the demand side, all the suppliers have been vetted because you have done that, and so they can expect high quality.
Curation generally, like, raises the average trans action quality and also the worst case scenario in some sense. You know, James the the, you know, absolute transaction quality higher matching is like helping people decide who to transact with or under which terms they should transact. So that could be like, you know, they're tons of suppliers, we're gonna recommend 5.
It doesn't have to be saying you're gonna transact with this one person, but it's like cutting down the search costs and all the complexity of, like, figuring out who a potential partner could be. And then supporting services, these are things like, you know, payment processing, standard contracts, communications tools, And also things like value added services. Right?
Like, you know, if you're gonna have sort of a an insurance option or a guarantee, like, we were talking with about with GoFund may your support make the transaction sort of happen more effectively, you as the marketplace organizer could have particular skill and advantage in any one of these dimensions, and that might also guide, like, how you try and fix a market failure. You know, so there are a couple of different ways, for example, you could try and improve average transaction quality.
Right? If you're really, really good at vetting suppliers, right, you just happen to know the entire Morgan, you know what makes a good supplier, and you have a really efficient way of of figuring out who's high quality, then maybe you go for curation. You try and build a super narrow heavily curated platform with only the best supply, and then your advantage to demand is that, like, they know every transaction is super high quality.
If that's really hard, then, like, maybe you instead go for a support strategy, like, where you get really good at providing trust guarantees and sort of, like, down you're sort of insuring against downside risk, or you take on a lot of the transaction yourself, right, like you sort of, you know, instead of having the supplier transact directly, maybe you, like, sort of fully partition the Morgan, you, like, buy from the supplier, like, improve
anything, you know, any goods that are lower quality and then sell them directly to the customer. But whenever the case, You might have different sort of ability and comparative advantage in these different dimensions of the transactions, and that also guides what you actually work on.
Interesting. You know, I agree with you that you need to look at what you're good at already, but I gotta say in coaching teams, often they keep doing the thing they're good at even though the market doesn't really value that piece of it. For instance, they get really good at getting supply, and they keep reporting to me on how great they're getting supply, but in fact, it doesn't matter if they don't have any demand. Demand is the harder side. We call it demand side marketplace.
If you get the demand, the supply will show up, and they say, look. We've got all the and look at how great our supply is growing. And then and I'm like, I don't care. And they're like, but we're doing so great. It's our comparative advantage. I'm like, the market doesn't value that comparative advantage. So sorry that you developed something that the market doesn't value and won't pay for, but you've got to change that. And it's a very hard emotional shift for them to make.
So I would caution against saying just focus on what you're good at because it's often the case that what you need to do in these marketplaces is figure out where's the greatest point of leverage? Is it the vendor financing? Is it the insurer And if it is, then that's where you need to be. And if you're going to make a success in this sector that you've chosen. Yeah. You're totally right. I should have done that mapping you know, again, market failure. I talk about sources of market failure.
Like, what is it in the market that's stopping the transactions from happening? Is it that people don't trust each other? It that they don't have a way to communicate? Is it that the search cost to find a good transaction partner are too high? The specific design intervention you choose has to address whatever is actually driving the lack of value or the loss of value. And so you could be, like, the world's best, exactly to say.
You could be, like, the world's best like recruiting suppliers because you built a platform that's super easy for them to use. It's like a really nice SaaS product. And so every single one of them wants to sign on. But if the problem is demand side search costs are too high or something, or, like, you know, demand side trust is low. Whatever it is, it's keeping the demand side away. Bringing in the suppliers doesn't get you there. And so it's definitely true. Right?
Like, you wanna be in this can we repurpose the white hot center term? We wanna be in the, like, the white hot optimum. Right? What is it something that is addressing a principal source of value loss in the market that you are yourself really good at or have some special advantage in fixing. Got it. So you keep talking about market loss.
If a market for NFTs doesn't exist and it's just growing, Using your word about market failure or market loss, where is there a market failure or loss in the NFT market today? Cause I would argue there isn't yet because we're just inventing like, there's no way of understanding loss if you don't have a concept of what already is. How does it help you? Cause you're really into this term market failure. Is that because it's really dramatic? It gets everyone's attention, or is it useful?
Is that language helpful for founders in order to think things through? Because I'd love to use market failure in the places where sense to help people see stuff they haven't seen, but I'm not sure that it's helping me right now. I wanna understand that better. Yeah. 1st of all, where does the term come from?
So it really comes from, again, like, classical economic, We have this idea that markets somehow, you know, push towards better outcomes where better is actually, like, interpreted very broadly. Like, the most sort of classic economic way of thinking about it is we just make outcomes more efficient. You find people who want to, you know, have things that they want to trade and you have people Flint people who want them and you enable trade. Market sort of Move toward pareto optimality. Exactly.
They're moving towards pareto optimality. Pareto optimality, meaning there's no trade left that would make everybody better Exactly. They're pushing in a direction that, you know, sort of maximizes, you know, sort of total trade value, you know, relative to what the market will possibly bear. And we have a more nuanced view of this. You know, it's not just, you know, transaction sort of like maximizing the total productive power of the market, but it's you know, sort of the value of the Morgan.
Right? Like, we care about fairness. We care about equity of outcomes. Like, all of these things together is sort of I like to use the word social potential, right? Like, how much value for society in aggregate could a market create? And so when I say market failure, All I mean is that the market and, you know, whenever it's boundaries or sort of which are sort of abstract boundaries are is not achieving the maximum social potential.
And the cool thing about, you know, market failures, like, suck. Right? That's like a circumstance where we're not achieving maximum social potential. Right? We'd always rather achieve maximum social potential because that's, like, good society. But the cool thing about it from the perspective of entrepreneurship is any time a market is not achieving its maximum social potential.
If you manage to achieve more of that or if somehow help it, like, do more, you know, sort of do more, it doesn't even have to get to its maximum. You're creating value literally by definition, right, you know, when I try and sort of understand a marketplace opportunity, I look for places where markets aren't achieving their maximum possible value. And, like, to that to me, But that's every marketplace, Scott.
Sure. But some are doing it more than others, and sometimes the sources are more addressable than others. And so, like, you ask about the NFT, Mark, it. And, of course, yes. Right? Like, you know, I spent 2 years teaching at Chicago, but I still completely believe that, like, essentially every market does not achieve its maximum social value. Sorry. That had a lot of negatives in it.
I still believe that essentially all markets have some degree of market failure, but some are much more efficient than others. You asked about the NFT context. Right? Like, to me, you're right. I think we haven't yet seen sort of the structure of failures that might arise in the NFT market itself. But we are seeing NFTs addressing market failures in other existing markets. Right? Like NFTs to me, like, right now, there's sort of a technology that It's unusual. Right?
It's a technology that may also become, like, a market object in and of itself. But as I said earlier, like, they're creating property rights. That address what was a market failure, say, in the art market or in in a sort of other creator markets where people were creating digital content, and there was no way for them to sort of successfully, like, trade that value to the people who valued it the most or, you know, and so in that sense, that was the market failure.
It was, you have people who have content, you have people with, like, really high demand for content and who want to support these creators, but there was no framework. There was no technology for doing that. You've also been thinking a little about composite marketplaces, which are multiple marketplaces glued together with network effects that flow from one to another. How are you thinking about this? Oh, so these are super cool.
So I've been thinking about these ever since I came across a business called Rocksell, which, r a k s u l. But it's a Japanese business. So that's the English transliteration of the name. It's 3 marketplaces together. So they started as a printing services marketplace, and Japan has this massive small business culture.
Many, many industries, and printings, one of them, aggregated up, you know, a bunch of the best printers and provided an electronic commerce front end to, you know, do your printing transactions. And so, you know, what was going on before, right, like, you'd call your local printer. They'd say, great.
They'd take the job, then they'd realize they didn't have quite the right size printing so they'd call one of their friends who's like, you know, half the city over and, like, and as a result, like, your job ends up being done very far from you with a lot of, like, sort of individual sort of subcontracting along the way, and that might also take a long time super inefficient.
Yasu County Matsumoto, incredible founder incidentally, who was, like, basically my age, who was, like, the coolest thing ever talking with him about this, like, he sees this market and says, no, wait a minute.
Like, we can find the best printers and actually invest in them, make them even more efficient, right, you know, lease them higher quality printing equipment, like, just sort of lend them some some engineers, you know, to make their operations Beller, and then provide, like, sort of a single electronic commerce front end that's gonna take care of all of the matching and also provide a bunch of support, like quality guarantees and so forth. So they're doing, you know, a lot of direct matching.
Like, you just tell them, here's what I wanna print. They choose a printer, call them up, and send them the job, and then some other supporting things like a quality guarantee to make sure that you understand that the, opportunity is high quality. They did that. It succeeds. They go off and they were raising, I think, a series C round. They they raised a bunch of money, and then the founder announces, okay, great. Now we're gonna launch 2nd marketplace. Right?
They took all this money that's sort of, like, you would nor normally think, like, okay, they're this is now they're they've succeeded. They found product market fit. They're just in scaling like, okay, great. We're gonna launch a second business, which is a ground shipping marketplace. So they operate this they start operating the second business where very similar.
So Japan has all of these small business shipping companies, and some of them are like really small, like 1 or 2 trucks, and they're taking like, slack capacity and gradually, like, sort of, like, expanding. They partner with 1 of the big shipping companies, you know, to sort of both provide, like, deal flow and to, you know, sort of try and flow opportunities to some of these smaller businesses. And then, you know, so we Pete writing a case study.
Like, I flew out to do a case on these two businesses And I get there, you know, sort of my friend and collaborator, Nogo Asato, who's the director of the Japan research center says you're not gonna believe that's just launched a 3rd marketplace, which is a television advertising marketplace. And so television ads turns out James business structure. Again, there are like tons of small individual channels that have trouble selling ads.
And, you know, so they built an electronic commerce front end for the TV ads. So they bought the same business 3 times, right? It's an electronic commerce front end Morgan very fragmented supply side where they've also done some things to invest in making sure liesize really high quality. But the question is, like, why these 3 businesses, like, poses to my students? Like, I'm so confused. Right? Like, could you imagine telling your investors who just gave you a bunch of money?
Like, we're about to launch like this totally unrelated seeming second marketplace and then a third one? What's the trick? The biggest cost point in printing is moving all the paper around. It's shipping. And they discovered that if they can make the shipping market more efficient, not only do they make money there, but they lower the costs on the printing side pretty dramatically. And meanwhile, like, who's the customer for the TV ads?
Well, first, you ask, who's the customer for the prints goods? It's mostly a bunch of small businesses that printing menus or advertising or marketing materials, if you can take those marketing materials and turn them into TV commercials, now you can own like 20% of your customer's marketing budget instead of 5%. If you get internet ads, maybe like 50%.
And so they built an applet for taking your print materials and turning them into a 30 second TV commercial basically, you know, sort of entirely automated. And they can say, look, we just offer you this additional service. You can get on channels near where your customers are that previously couldn't sell their ads.
And so it's an unreal paradox, but when they bring on, like, more TV supply, like, more ad space supply, That raises the value for the customers of the TV market who are also the customers for the printing market, which raises the value for printers to be on the form, which raises the value for shippers to be on the platform. So there's this network effect that flows, like, you know, bring on a supply side participant on 1 in one sub marketplace. Drives supply to marketplaces over.
And what's the vision for the business? Right? There's gonna be like the back office for every small They do all of these different operations that they need in this, like, very high quality, like, guaranteed, organized way through a single central platform. Sure. Yeah. I'll tell you my reaction would be fan freaking tastic. Sounds awesome. More network effects.
And if the CEO's dynamite, then you just the CEO and say, I mean, you know, Bezos is selling books and then suddenly he's doing AWS and suddenly he's doing Alexa. It's like, yeah, let him go. Spot on, right? Like, and indeed it turned out it worked out really Beller. Right? Like, the market suddenly realized that Rockwell wasn't a printing company. Like, their valuation wasn't relative to other printers. It was a major tech platform. You're exactly right. You have to see it that way.
Yeah. It's not like general electric who had experience in manufacturing, and they were doing refrigerators, and then they were doing jet engines, and they're in manufacturing allowed them to scale up in both things, but there was really no synergy between them except the manufacturing expertise. That's sort of a conglomerate if you will.
These guys are actually describing a situation where it's a quasi conglomerate, but there's also now scale effects and network effects that allow you to become one thing. We call that, of course, reinforcement. We've got an article called Reinforcement that addresses how you take 1 network effect and add other network effects on top of it once you get 1. Well, that's a great case study for us to think about with composite marketplaces.
Not only do you want to start with the white hot center in one market and then expand to own that whole printing market, but then you start to see other reinforcements and take more and more market share and then use those network effects to take much. Great. So it's the way you build decacorns and companies larger than that. So Scott Commerce, As always, my friend, so fun to talk to you. We will have you back sometime when we're both more rested, and we'll cover some more stuff.
I think all of these very tactical things are very good for founders to hear. And I know I love riffing with you on them. I always learn something every time I chat, and thank you for doing your work. Thank you for doing the work you Thank you. I always learn from you. And, you know, if there weren't people like you backing and helping grow these fantastic companies, I have nothing to write about. Thanks. Alright, buddy. We'll talk soon.
At NFX, we believe creating something of true significance starts with seeing what others do not. Send this episode to any friends that may need these insights and frameworks, and feel free to rate and review us your favorite podcast platform. Thanks for listening to the NFX podcast.