About 2 years ago, Intuit Founder Scott Cook came to NFX to meet with James Currier to discuss their shared interest in network effects. It was one of the most compelling conversations James has had with anyone on the topic, and so we recently visited the Intuit HQ in Mountain View, CA to continue the conversation.

Today we’re making it public, for the benefit of Founders everywhere. The discussion was wide-ranging around topics including:

• How eBay’s network effects saved it from Jeff Bezos
• Virality vs. network effects: A critical difference most conflate
• The mistake that trips up most 2-sided marketplaces
• How Microsoft orchestrated some of the world’s strongest nfx
• Why the old rules of defensibility no longer apply

And much more. Listen to the full podcast on SoundCloud, or download the NFX podcast on Apple or Spotify.

Transcript

James: Today, Scott Cook and I are talking through network effects. You know, the audience, Scott, is Founders, mostly early stage Founders. So they’re looking for practical tips, tools, stories, details, and things that they can use in their daily lives. Two years ago, you and I had a chance to discover that we’re both network effects wonks. We’re students of network effects and got a great talk about them in detail. There aren’t that many people, actually, who have spent the time, as you have, to figure out the nuances and how they work, and that sort of thing. So I just thought it would be great for us to almost recreate our discussion about network effects so that founders could benefit from the sharing of some of the details about this stuff, and some of the thinking. Just to set the stage so Founders understand, are you investing … Are you and your family still investing in companies?

Scott: We do some. In the Pre-K education field we do a lot. As much as we can find, as that’s our kind of philanthropic purpose. Then on the side for personal portfolio, yeah, once in a while.

James: Once in a while?

Scott: Mm-hmm.

James: Is there a particular type of thing you’re looking for? Companies with network effects, companies …

Scott: Oh, gosh, we’re always looking for companies with network effects. Yeah, generally, they won’t be the earliest stage because that’s pre-establishing of a network effect, but.. So they might be, when there’s some evidence of market traction and a team and business you can believe in.

James: Got it. You’ve talked a lot about the need for Founders to have their durable competitive advantages long term. What does that mean to you?

Scott: Maybe there was a time when you could start a company and not have to worry about competitors. But in the world we live in today, with giants that have such speed, and agility, and resources, whether it’s Facebook, Google, Amazon, whether it’s other startups, the vibrant startup infrastructure… You gotta assume that there are going to be other people trying to do the same thing as you. Either at the same time, or once they see that it’s successful. If you don’t have a source of durable advantage, well, you’re roadkill. All that work goes to nothing, and you get crushed. Then, not only to survive, but to thrive, the size of your profit stream is largely a function of your size of durable advantage. Because if you don’t have durable advantage, you’ll have 18 competitors, and they’ll compete it down to no economic return.

James: What forms does that durable advantage take? One of the things that we did, we published a perspective that in the digital age, there’s sort of four major things. If you discount IP, and particularly in software, it’s tough to use IP, but you’ve got brand, you’ve got embedding, where like an Oracle will embed their software in a company. You’ve got scale, like an Amazon just could scale so the prices get lower, and then you’ve got network effects. We kind of posit that network effects are the greatest of the four. Do you have any way of looking at durable advantage like that, that might be a framework that the Founders could learn from?

Scott: Yeah, it’s similar. The conclusion is exactly the same, that the one big, the giant in the room, are network effects. The others all have their flaws. So a traditional one was fixed costs scale leverage. That assumes that nobody else can put the fixed cost in. Well, today with the tech titans out there, they can put the fixed cost in a wink, in a blink. So fixed costs scale leverage assumes a capital-constrained world, and we live, at least in tech, with companies that are seemingly capital unconstrained. So that doesn’t have the power it used to.

One that gets some discussion is switching costs. There was a time when certain companies would have at have unique access to a source of supply, like De Beers with diamonds. But boy, you don’t tend to see that anymore.

James: The assets of the new economy is just data and software and all that moves very easily –

Scott: – moves easily, and brand I think is not a worthy defense at all. I’ve seen companies with phenomenal brands get crushed in a matter of years. I mean, go back to times that we remember in the early age of PC software, the best brands in the industry were Lotus 1-2-3, WordPerfect … and Microsoft just crushed them in a matter of years, and they had the brands. So I don’t think brand buys you very much. I mean, the best brand in search was … you name it, AltaVista, or maybe Yahoo!, now they’re roadkill. So I just don’t think brand matters that much anymore. If you go back if you look where brand used to matter a lot was in say consumer household products like those that may… appear on a grocery or drugstore shelf.

Those are product categories where the consumer already has been using the product for years, just ran out, has to re-buy. The product only cost three bucks. They walk by that shelf twice a week. All you have to do to get them to try your brand is get them … they’re going to reach out toward the shelf and they just have to move their hand 20 centimeters to pick up your brand. Maybe brand image could make someone who’s going to buy anyway move their hand 20 centimeters to buy your brand. But we’re not talking about that world. The world we live in there’s … no one’s walking by a shelf, they didn’t just run out of your product. You’re trying to get them to do something new, generally that they’ve never done before. So I just don’t think brand is builds much advantage, nor is it much of a defense anymore. You need something more powerful.

James: Like these network effects.

Scott: Yes.

James: Are there other things other than network effects that you’d like to rely on, or really are we focused on that?

Scott: It’s not something you like to write. There are brief times when if your idea is sufficiently different, well … actually let me speak to one source of venture that still does work, but they’re rare, and that is, oh I think there’s an HBR piece that calls it multi-factor process advantage. So, this is something where you have a method of production that is very different from the rest of the industry and very hard to copy because it differs on so many attributes, such as Southwest Airlines.

Southwest Airlines has been the best performing US airline for 30 years. They use their labor different, they use capital different, they just … there are so many differences between how they operate versus a normal airline. And the other airlines are hidebound and union constrained and so they just can’t copy the superior method. So Southwest has retained a fairly durable advantage.Toyota production system is another. Toyota makes cars in the plant in a way that’s fundamentally different than the procedures used in normal auto plants.

James: Famously, they’ve allowed the other car manufacturers to come into their factories and videotape and spend as much time as they want and in the end, they’re incapable of copying.

Scott: GM had a Toyota factory in the GM footprint, in the United States. I had all the data and visited. I learned about this in part from a GM executive who was flying out from Detroit to visit the NUMMI plant that was here in the Bay Area, and the other automakers essentially haven’t copied well. You can see companies like Danaher and Ford who’ve copied the Toyota Production System and reapply them in their industry. I mean Danaher has a stock growth track record much better than SAP, on par with Intuit stock growth since we went public 25 years ago. Yet data Danaher works in industrial products and test equipment for factories, not sexy businesses at all. But they’ve applied this multi-factor process advantage to real advantage. So that-

James: We might even look at Apple and as Amazon might be-

Scott: I would say Amazon has built a multi-factor process advantage in software development. The speed with which they can move, and innovate in a large… It used to be, you build a large software company, and you get slow and stupid and not innovative. Amazon’s kept speed and innovation, rampant innovation, despite size. So I think there’s something going on there special, but those are rare. Few companies develop that.

James: That’s about culture, and about hiring, and real determination to do things different.

Scott: The real invention of a better way that works that’s generally totally against the way the rest of the world works. So that’s just not going to happen very often.

James: Yeah, that’s right.

Scott: But it’s powerful and those … that I think continues to work.

James: That’s right. So now you’re the chairman of Intuit. You founded it in 1983, a long time ago. It’s got a market capital of more than 70 billion, every year just keeps getting better. Did Intuit have network effects at the beginning, and does it have network effects now?

Scott: Yeah. For all of my work and study of them, we don’t have the strength in network effect position that I wish we had. So we’re a paradigm difference that allowed us to beat the competitors in the ’80s and early ’90s. Then we have enough scale … we have some scale advantages. Then in QuickBooks accounting software, we have developed a degree of I’d say a moderate network effect. In that, accountants know QuickBooks and prefer it so they recommend to their clients. There’s more bookkeepers who know QuickBooks, so they recommend to their clients. It’s easier for the client to work with an accountant if you’re on QuickBooks because they use QuickBooks. So we would have a moderate network effect advantage. I would qualify that as-

James: It’s probably not a protocol network advantage, which is a direct network. But it’s something like that, which is that there’s some sort of standardization through the labor pool that’s taking place.

Scott: This has a degree of switching costs. The switching costs here is the cost to learn a new bookkeeping system. Just like languages have high switching costs because once you learn a language, it’s so much easier, just use that one than learn a new one. This is not on the same scale of difficulty of learning. But once you have learned how a system works like QuickBooks, it’s so much easier to keep using it than to have to learn a rival.

James: So there’s that groove in. Then any new hiring manager is going to be taught by the marketplace that they should just hire QuickBooks person. That’s sort of a two-sided protocol.

Scott: Yeah. So we got a degree of what I’d call moderate effect of network effect there.

James: Have you tried any others in the business? Have you thought of adding any other like a platform? I mean, I know that Intuit has a platform.

Scott: Yes. We do have a developer platform again built around QuickBooks. Again, that’s moderate. There’s more systems that hook into QuickBooks. Thus, QuickBooks, If you adopt it, gives you more data access and interaction that’s automated with other systems. So I’d say again, that’s a moderate level. We’ve tried to make that stronger. I think maybe we’ll discover out how to make that stronger. I want to ask how you first discovered network effects and became aware of them. Then I’ll describe it wasn’t at Intuit that I discovered the power of network effects, It was elsewhere. So first, how did you get?

James: Yeah, so we started a company in ’99, that did not have network effects. In the end, we sold it to a company that did have network effects. When we got over there and saw how bad their management was, and they still had a $7 billion market cap, and this is monster.com… The fact they still had $7 billion market cap and they were running it so poorly, and no matter how they try to screw the company up, it’s still continue to chug along and produce a billion dollars of revenue and, and be worth so much that they were buying us, not us buying them. It was a real shock to us.

That was 2004, when we sort of said, “Whoa, this is a whole different paradigm. This is really durable. You can make all sorts of mistakes and go to sleep and wake up in the morning and your company is just bigger than it was before more powerful than was before.” So everything we want to do going forward is got to have these things. So we started becoming students at that time. It’s taken probably a decade to start to tease out all the different types because we’ve heard about Metcalfe’s law and that only relates to direct network effects. But then there are these two sided network effects like the one you described that Intuit has with the labor market, or a marketplace, or a platform network effect, or an asymptoting marketplace network There’s a several of these two-sided network effects that are different from Metcalfe. There’s there’s a whole region around data network effects that just haven’t been measured or discussed yet. So it’s taken us all this time to sort of tease all that out. But that’s how we came to it.

Scott: That’s an okay way to do it.

James: Now we’re investors. That’s why we’re helping others to do it. How about you? When did you first discover this?

Scott: It was in the late 90s. I … some crazy reasons managed to be on the board of Amazon and eBay at the same time. eBay was … it just defied explanation. I remember I first came across eBay whenever they published books of … This was the early days of the web. They published books that showed the amount of traffic and screens… pages per user for thousands of websites. I think that book was online. I was just perusing to see which websites were getting the most activity in particular, which we’re getting the heaviest usage.

I was just scrolling down these pages online looking. Most of the pages per visit, were in the two or three range. So people barely used websites back then. Then I ran across this thing called eBay, E-B-A-Y. Then I get an email from Meg Whitman and she said, “Hey, Scott, I’ve just joined this company as CEO of eBay. I’d love to come talk to you about you joining our board.” I said, “Well, I don’t know what eBay does, but I really want to learn.” So I found out this amazing thing. They opened up a retail store, Pierre did, and he left it empty. He put nothing on the shelves, and the world came and filled it up with merchandise.

Because it was kind of the original two-sided network executed marketplace done in software. It’s just amazing how the thing grew. John Doerr and I were skiing in Aspen over Christmas, and Jeff Bezos emailed us saying, “Hey, I’m in town. I’d like to take you guys to dinner.” So we left our families, we had dinner with Jeff. Partway through, we said, “Jeff, so why don’t you invite us to dinner?” He said, “Well, I want you both to know that we have been secretly building our eBay killer. We didn’t want you to know, Scott, because you’d have a conflict with that. But we will launch it on Tuesday. I’m sure you’ll want to, of course, leave the eBay board because they’re going to be toast. But so I want to give you that heads up.”

James: Nice. Very kind of him.

Scott: Yeah, very much. So I said, “Well, Jeff you know … you know how I love what you’re doing Amazon’s phenomenal, and you have truly vanquished a number of competitors surround you. Around you, 359 degrees, you have mediocre companies as potential places, categories you could enter, except one. We have a company with a powerful network effect. This is not going to end very well.” But he was undaunted by my advice. Then I watched what happened. It turns out that within a matter of a couple of months, the two most powerful companies in the digital world, which at the time where Yahoo! and Amazon, both decided to attack eBay with clones.

They had done excellent clones of eBay. In fact, the software was better, the sites looked better. eBay at the time looked kind of more like Craigslist, and these two competitors from … had really slick working sites plus they were being advertised from the most powerful home pages in the business for several months, Yahoo’s homepage, the most powerful page in the world, did nothing but advertise Yahoo! auctions plus a major pricing advantage. eBay at the time … take rate was around 6% of every transaction. The take rates at Amazon and Yahoo were zero, it was free.

So you would think that a free competitor, advertised and delivered by the most powerful entrants in the world would crush little eBay. But in fact, the reverse happened, they bounced off, nothing happened. eBay kept growing, those guys couldn’t get traction. Even though they were free, and eBay charged 600 basis points. I said, “That got my attention.” If there’s something more powerful than I’ve ever seen in business and then you started seeing the … Intel microprocessor, the Microsoft Operating System, Word and Microsoft Office, you started seeing these… all of a sudden the picture came together: shoot, there’s a number of companies that have built some phenomenal competitive advantages. Where no attempt to disrupt them has had any effect and where their share positions get closer and closer to 100. This is an effect the world has never seen before. Except in a government granted monopoly, which is … are increasingly rare nowadays. So that’s what got my attention. Then I started studying them in earnest.

James: That was sort of late ’90s, early 2000.

Scott: Yeah, ’98.

James: Did you stay on the Amazon board, or the eBay board?

Scott: Ultimate I had to choose. Because the company’s stayed rivalrous. I elected to stay on the eBay board.

James: Nice. With the network effect. Now eventually, Amazon has gotten very large and does have a big marketplace. More than 50% of their transactions are now in their two-sided marketplace, but it just took 20 years.

Scott: Yes. They decided to really embrace the two-sided market. The first merchant in the world ever to do that. To really embrace after having been emerged … where they put the merchandise on the shelf. They embraced the two-sided where other people put merchandise on.

James: So it was a mental shift that they had to make first. Then when they came at it again in the late 2000s it started to work a little better.

Scott: The whole move into Amazon auctions was a move there. So Jeff got them. Jeff is so smart. He got the mind shift early. He just had to work. Initially, they couldn’t win by going head headfirst into eBay. They had to find a way to work around.

James: I see. They’d had to have a different chain of supply. Yeah. Interesting. So, what was the most surprising thing you’ve ever learned about network effects? These are great founding stories.

Scott: I’m still learning, there’s so much. I think most of what we will know, we don’t yet know. I’d say the thing that’s been surprising is how small the founding teams are of some of the great network effect businesses today. Pierre Omidyar worked alone to write… and launched eBay after what he says was a three day weekend of work. So what’s stunning is how tiny the teams were, who built, and launched the first versions of what became powerful network effects. That just wow. The world’s never seen something, where so few can produce something such a global strength, reach and strength.

James: On the backs of this internet thing that someone else built for all of us.

Scott: But you can have small teams on the internet that don’t amount to much. You can have big teams. That’s been the biggest surprise to me.

James: Which also speaks to the capital efficiency you can get from these companies, which implies greater returns for the employees, for their stock options, which implies greater returns for the investors.

Scott: Absolutely. Pierre, up till that time the investor … the investment that Benchmark made into eBay was the single best venture investment ever made in world history. But also think of Pierre’s thing. He… Not only did he code it himself, he launched it himself. He was the only employee for many months. Only when so many checks came in that he needed help cashing the checks did he finally hire somebody. So remarkable efficiency of both labor and capital.

James: Yeah. With these marketplace network effects, these 2-sided ones… A lot of people have talked about the Twitter, and the Facebook, and the WhatsApp. These are direct network effects where everyone’s coming for the same thing. One Twitter account is the same as another, one Facebook account is the same as another, but when I go to an eBay, when I go to buy something on Craigslist or sell something on Craigslist, I’m … those are two different types of people with two different types of experiences.

Have you thought a bunch about … I’d love to dive into a discussion of how some of these two-sided network effects work in more detail, because you’ve got one going here with Intuit. You’ve been on the board of eBay a long time, and you got to watch that grow up from the beginning. I think the one-sided network effects, or the direct network effect has been sort of overdone, but the two sided hasn’t been nearly as much discussed. I think that might be helpful to people to hear some of your thoughts about how you see those things playing out.

Scott: Well, lets exchange thoughts. One part is that two-sided, they’re also n-sided where there’s more than two sides. So the classic one here is the Microsoft and OS network effect. So say Microsoft Windows, where you had hardware makers, so makers of PCs, independent software developers, consumers, people who buy the machines, corporations and consumers, and then we had Microsoft. So they were actually orchestrating three sides. The thing the consumer chose, then drove hardware makers, which drove the software, and so they were able to orchestrate a three way. So, two sides is one part of a … what could be more complex. So that’s the first observation, maybe more than two sides. So what have you learned about two-sided?

James: The two-sided network effects are more prevalent than three-sided. In fact, I think the three-sided are less common than the n-sided, where you might have a market network where there’s a whole bunch of different types of people that are inside of an existing offline marketplace, that we then digitize through a website and mobile apps. The three-sided is so difficult to get because they all have such strong independent ideas about what they want out of the marketplace. Typically, when you’re building these two-sided marketplaces, you basically are building two different companies at the same time. You have to satisfy the demand, and you have to satisfy the supply at the same time, in this … in the appropriate volumes, so that liquidity is achieved for both sides and people don’t defect to someplace else. It’s a very difficult plane to fly. So flying that with three becomes astronomically even more difficult.

So we’ve seen very few of these three-sided marketplaces, as you described with Microsoft. But there you go, I mean, they’ve been around since ’76, ’78, and they’re still worth $700 billion. It’s very, very durable once you get it, but hard to do. We’re now seeing more of these n-sided marketplaces where you could have, let’s say, an event planner with a photographer, with a venue, with-

Scott: Caterer.

James: … caterer. So they can all coordinate and collaborate, using the software in ways that they couldn’t have done before. We’re seeing that go in the travel industry, we’re seeing that in the consulting industries, et cetera. Because these are marketplaces where there’s lots of independent workers, and obviously Intuit and the software you have services a lot of these SMBs. They are all in their own small networks that are offline today, and we could help bring them online. So we’re seeing that. The other challenge with these two-sided marketplace is again, gets back to the issue of liquidity, where if I come, and I list an object that I want to sell, and it doesn’t sell, then I might never come back. Or if I come to find something, and it’s not there, It’s not in quantities, or prices that I want, then I won’t come back.

James: Balancing that and then paying both sides essentially with your equity or with your time to balance those out as you grow, has proven really difficult. So, as a result, we’ve seen a lot of success with companies like OpenTable or Salesforce, who immediately put out a SaaS tool, like you’ve done to put out SaaS for a whole group of users, and then you build up the platform. Then on top of the platform, you can stack other applications and that builds you a network effect that starts with a SaaS company first. But again, it takes a mental shift to then move into this new network effect type of the business.

Scott: Yeah, let me pick up on two of the things you talked about, because I think there’s so right on. The first is you mentioned, it’s like running two companies at once. I think that’s quite true. I see companies, when I coach CEOs and our internal startups, they’re often much more comfortable at looking at one side of the market, and they tend to forget the other.

James: Because of personality?

Scott: Yeah… just they’re more familiar with one side. They’re trying to solve the problem on one side by using the other side. But in fact, you have to solve some large problem for each side. The problem will probably be different. For PayPal, a buyer wants to be able to buy the item, a seller wants to be able to collect money. So there’s … but you have to solve each … a big problem on each side. Otherwise, you won’t get people in-

James: In your own personal biases, or your own affinities end up coloring you balance.

Scott: I find teams will be so focused on one side and just make assumptions about the other side. No you gotta understand, what’s their biggest problem? And how are you going to solve it? Because if you don’t solve it, they’re not going to come.

So I think that’s … on each side, there has to be a large problem that you can solve well by participation in the network. Then secondly, there’s the thing that you mentioned about the chicken and egg problem, that all networks have a chicken and egg problem, has to be solved first. If you can’t start getting scale, you’ll never get off the ground because these will only work when you start getting volume. I’ve not seen good work written outside the company on the ways to crack the chicken and egg problem. But our strategy team here at Intuit, there the best work I’ve seen. There’s four or five different ways that we’ve observed people solve the chicken and egg problem. One is virality… LinkedIn did this where members invited other members, another is incentives, PayPal did this where they paid participants five dollars if they brought in a new participant and pay the new participant five dollars.

So those are a couple of methods that have been used. Another is the one you mentioned, which is build something with standalone value that works standalone where you don’t need the other side. Like Microsoft Word was a very good word processor. You didn’t need other Word users because you could use printing as the interface. That’s what we do with QuickBooks, OpenTable did that by first selling a software platform to restaurants for managing reservations and you didn’t have to be part of their network for that to be of value.

James: Yeah, we’ve actually got an article called 19 ways of solving the chicken or egg problem. We should send that to your team.

Scott: The world’s needing someone to talk about that.

James: Yeah. We’ve got some good videos on it as well. It’s definitely a big challenge. The point you make about needing to run two different businesses at the same time… you know, often these marketplaces one side is more important than the other. So it’s actually easier to get one side than the other. In fact, the founder, I found needs to develop more affinity for the harder to get one because as soon as they get them, the other will typically show up.

Scott: There are cases where the.. one side will show up as soon as you get the first side. You’ve got to figure out how to get the first side.

James: Correct. We’ve seen a lot of failure in the marketplaces when they have an affinity for the weak side. We’ll actually call them demand-side marketplaces meaning the demand is what’s important on this side or will call them supply-side marketplaces meaning if you get the supply, the demand will come. A good example of that would a company like Outdoorsy, where there’s 35 million people trying to rent an RV, but there’re only 50,000 RV is to be rented in the United States. So if you can just get more supply on that marketplace, you are going to end up with a lot of the demand just finding you because they’re also looking for it.

Scott: I think job boards are similar way. If you have the jobs, job seekers will find you.

James: That’s right. So figuring that out and getting people focused on that is one of the challenges. So who do you turn to for knowledge around network effects? Who do you and your teams turn to? I mean, Tom Eisenmann has written some great things in HBR out of Harvard Business School, but not since 2004, 2007. Are there other folks that you can remember reading that you thought were great?

Scott: There are some academics who are reasonable. Andrew Hague, who’s at BU and Marshall Van Alstyne, also at BU, have written and got a book. But again, that’s a decade old, I think. So I think there’s good material there. I think the stuff that you guys are producing maybe the most current, and most inclusive, even beyond what I knew. So that’s … I think where I’d send people first.

James: We found that when the word network effects comes up, we first have to unpack viral effects from network effects. Certainly viral effects, they’re completely different playbook, they can often be better used on top of things that have network effects simply because of the palette of semantics. A palette of language you can use to develop features and pathways for people to get viral, but it really is a separate thing. Then once you’ve just been able to identify the look network effects about defensability, and about value creation in your product, then you can start to break it down into its multiple components.

Scott: Yeah, well said. It’s often conflated with something unrelated which is virality. It’s great if they happen together, but they don’t have to. They’re separate effects.

James: That’s right. You could actually buy a network effect on a two-sided marketplace. If you spend $1,000 to get enough supply, times 1000 people, that might be enough to get you there. It doesn’t have to be viral at all and have a great network effect. So what advice Scott, would you give to Founders as they’re starting their companies, particularly around network effects and their impacts?

Scott: One would be to figure out early on, is it really a candidate as a network effect or not? Because just a two-sided business is not a network effect. So there’s another confusion. A business with two sides may or may not be … I mean, at a typical magazine was a two-sided business. You’ve got readers who read the editorial and advertisers who pay for the advertising. So newspapers, magazines are two-sided, but they’re not network effects. There was a little bit of a transportation scale advantage for newspapers at one time, but magazines never had that which is why there’s so many magazines. You don’t at all see the strength and power and position that nobody in the magazine industry achieved what modern network effects do.

Scott: So just being two-sided does not mean you’re a network effect. I think a key to that is understanding multihoming. Is there a natural incentive for the participant, or the participants on each side, to deal with only one platform? That’s single-homing. Then there’s a much stronger case that you can build a network. On the other hand, if the participants on one or both sides can play the field easily, then you’re not going to have that durable advantage. You see, some of this with Lyft and Uber where it’s very easy for drivers. In fact, drivers usually run both and travelers, passengers often have both. Which is why you have two of them there losing money instead of one that’s successful, or profitable, at least.

James: Multihoming, sometimes called multi-tenanting as well.

Scott: Yes. So I think that’s an early thing to figure out to run the test of whether this really turn out to be a network effect or not.

James: To do that, you would have to read up on them, you’d have to think about it, you’d have to study them, you’d have to see what has produced network effects in the past, if you’re a founder, to try to get a better sense of whether you really got one or not. You got to find people who are knowledgeable about it.

Scott: Then managing to crack the chicken and egg thing. So reading your article on the 19 ways to crack the chicken egg problem.

James: When you think about these network effect businesses, I mean, eBay obviously, is now what? 24 years old or something. What are some of the exemplars you’re seeing even more recently? I mean, clearly, Slack has got a great direct network effect, I’ve got some embedding network … I’ve got an embedding defensability, where they’ve embedded as enterprise software into these companies. What are some of the companies you look at say, “Yeah, boy, those guys are doing it right recently?”

Scott: I’m not sure I can tell you who’s doing it great. I can tell you who seems to be delivering results. So GitHub would be another. Of course, the ones that are on people’s lips would be Lyft, and Uber, and Airbnb. What other names come to mind one of the more recent crop? Well, Instagram, WhatsApp, to be sure, in China a number of you know, Tencent.

James: Companies like Dropbox just went public. They’ve got a nice network effect there. Between all the sharing that that goes on, or you’re looking at a Poshmark will probably go public next year in two-sided marketplace, maybe even a three-sided marketplace, into… As they add stylists into their fashion marketplace.

Scott: I think… is Rent the Runway turning into a network effect, or is that …?

James: Doesn’t feel like it to me. It feels like more of a sort of eCommerce, direct to consumer. They’ve got a nice spin, they’ve got a nice pitch, but ultimately, they are the ones who are owning the objects and then renting them out.

Scott: Certainly, if you expand recent to include the last 12 years, you’d have iOS and Android operating systems with their associated app stores.

James: Absolutely. I mean, before Apple had a market cap of 40 billion, they then added iTunes and jumped to about 60, which iTunes had a two sided network effect for a while there, and then they added iOS, and they topped a trillion. As soon as they finally added network effects to their business, they multiplied the value of the company by 10x. What are the impacts of these network effects on society? Because, like you said, a few people can create so much value, a few people can be in control of so much communication between others. Many of these companies are based in the Bay Area, and then they’re servicing the world from here.

James: Have you thought a little bit about what we might expect to see with society, and what government needs to do or the broader impacts of the fact that these things are much easier to build and they were 30 years ago before the internet, and then the incredible power that they create once they are built?

Scott: I think the longer view would be that there have been in history… this is not unique to our moment. Situations like this have existed for 100 – 150 years. There are more common, more walks of life today, but the basic reality… You had the Bell Telephone Network, which is where the concept of network effect was named. Where one company eventually built a communication network, which had all the characteristics of the modern network effect. You had great concentration of wealth in the oil industry and in railroads. Railroads were monopolizing … because of the capital requirements, people didn’t put tracks. Left other tracks so you tended to have a high share in a form of communication / transportation –

James: Direct, physical, network effects. The core, most defensible.

Scott: And these were generally built by very small teams. The companies might be large, but the owners were concentrated in the original railroad barons, the oil barons, you know, like Rockefeller. AT&T was for most of my life, the largest market cap company in the world. So this phenomena has existed for quite a while. I think for tremendous societal benefit, I had a guy saying… What’s the thing I read? That they promised us flying cars, but they only gave us Facebook.

James: Yeah. Or, or 140 characters, I think that’s the quote.

Scott: Yes. That’s one way to view it, but here’s another way. So my family and I were vacationing three … four years ago in New Zealand, and we took a boat ride for a while, and we’ve done a hike and then we took the boat back, and I roamed around and met some people on the boat including four or five Brits who relocated from the UK to New Zealand about a decade before. And I asked them all, “How do you like life in New Zealand?” They described all the things they liked. So I then said, “So what do you miss about the UK?”

One said, “Oh, there was this pub we used to go to. We really love that pub.” Another said, “Oh, there was a shopping street. We will really love to go to the shopping street.” Somebody said something else that was physical, and I said, “Well, you know what’s odd here is you never mentioned any people, that you miss any people.” They said, “Well, of course not.” Pulled out their phone, “They’re right here, they’re on Facebook. In fact I see them more now than I did when I lived there. I see them every hour, practically, on Facebook every day when I wake up. No, I’m closer to my buds in communication and more frequently than I was when I lived in the UK, thanks to this thing called Facebook.” When you think about it, their predecessors who moved from the UK to New Zealand, say 50 years before would most likely never, ever see their friends and relatives in the UK ever again.

Scott: They would never talk with them, because long distance was so expensive. If they ever talked with them, it would be when somebody died, and then for three minutes. They would write letters painfully by hand, and then they would take weeks to get there. They would never physically see them again. But now we have worldwide free video calling. This ongoing register of life events, through Instagram, through Facebook, for free, for free. Oh, my God, what a change this makes to maybe one of the most fundamental human desires, the ability to stay close to those that you know and love. All facilitated by having these global networks on a backbone of free digital bits. Now, if we were divided into each country’s network, and they didn’t work together, or you have stuff like the way telecoms used to be regulated, It wouldn’t be, but because these are kind of national organizations, which can take advantage of the power of free.

Scott: It’s like science. When you get things to absolute zero, the rules of science change. The rules of physics change at … close to absolute zero. Similarly, in economics, the rules of economics are going out the window, when you get to free, and you start getting unlimited supply of free things. Thanks to Skype, global worldwide free video calling, Instagram, Facebook, Twitter, the ability to stay close. Closer than even when you are in the same city. Without these technologies is a change. We take it for granted now. But if you go back 50 years and share what we take for granted and don’t even think about to people who are living then, they would say: “The world could never get that good. That’s so far beyond any belief, only a god could deliver the thing you just described.”

Then you look at the ability to get a ride wherever you need it, the ability to find apartments and for people who have the spare apartments to take advantage of those to turn those into income. That’s allowing people to stay in their homes longer, they’re giving people second incomes. There’s so much unused assets in the world, whether they’re second bedrooms, whether they’re vehicles. Most vehicles sit idle 90 plus percent of the time. Whether it’s RVs who it idle most of the time, enabling assets bought and paid for, to turn into value for the user and for the owner.

This is a miraculous world we get to live in. I think of Wikipedia. You know a destitute girl in a slum in Bangladesh whose family has a phone and a data plan, has more information at her fingertips for free, than Bill Clinton had as President. Now you get Bill Clinton’s guys and gals two days, they can get a lot of information, but at her fingertips, and for essentially only the cost of the data plan, which particularly like in India, now they have the lowest data plan costs in the world, and that’s just any random kid whose parents have a phone, smartphone. This is unbelievable power, thanks to vast contribution network effect around Wikipedia.

James: 700,000 people knew where to go, which was to wikipedia.org. They knew where to put their time, and they knew it was going to be valuable to other people, and therefore it was valuable to them to spend those hours because they had confidence that, that would be the one place that the information would aggregate. Now the woman in Bangladesh can also know that, that’s where the information is aggregated. And that value to her, that value to the writers aggregates because we have this concentration. So yes, people get worried that we have this concentration of power, this concentration of influence. We also have this, flourishing of value.

Scott: Again, Wikipedia contributors aren’t going to contribute to some fractionated thing that nobody comes to. They want to contribute where the world can see and benefit from their contribution. What this has produced is something I wrote about an article I wrote in 2008. This volunteerism, where now, millions of people volunteer to help others through these platforms. So the Wikipedia editors and authors don’t get paid. It’s all work for free. That they do for a sense of social standing, for a sense of contribution, for a sense of benefiting others. We run help boards inside Intuit for our customers, where customers can ask questions and other customers can answer them, including in areas like tax and small business. Some of our customers will answer thousands of questions from other of our customers.

James: For free.

Scott: For free. We got some of our super users to come to one of our leadership meetings with our execs and three of them on stage. All of them had answered over 10,000 questions, all three. One, a school teacher, she gets home, she does her school work and grading, and then into the night she’s answering tax questions. Another guy retired, answers tax questions all day long, tax questions. Some of them they have to go research, they have to go into the IRS documents, and that’s no fun, to research, to get the answers.

Scott: Finally, one of our execs asked, “Well, why do you do this? Is this the points, the scoreboard, the medals, the badges?” They looked at each other and say, “Yeah, we’ve seen those badges and scoreboard, but no, this is how I help people. I know about this, I like being able to help people, I can now help thousands of people in a way I could never do, just helping my neighbors because of the size of these platforms.” It’s uncorked the degree of volunteerism unprecedented in world history.

James: Yeah. Where they’re giving people opportunity to feel meaning.

Scott: Yes.

James: Value in themselves and self-respect through these platforms as well. So all those benefits come to counter some of the concerns that people have politically about some of these really powerful concentrations. I agree with you, It’s amazing. Both is true.

Scott: What a world we get to live in.

James: It’s true, all true. All driven by network effect businesses.

Scott: Where you compare back to 50 years ago, so that’d be 1968, one … a phone wired to the wall, one phone in the household, cars that broke all the time, one TV, big hunk and box, three channels, you took what they gave you. If you stepped away, you lost it forever. Oh, boy, a phone call long distance so expensive that, practically, someone had to die or get married before there’d be a call. Wow. That’s just that’s just 50 years ago in the lives of us, or our parents. How the world has changed. One encyclopedia, now getting out of date every year, sitting on the shelf, painfully slow, no links, so no way to link… Between you’re interested in something you couldn’t follow because there were no links. Wow. No YouTube if you want to learn something, I mean, no Khan Academy. So you had to sign up to school or something. You couldn’t just learn on the spot. Now you got learning at your fingertips, 24 hours a day, for free.

James: We all know to go to YouTube. All the people who make that stuff, know to put it on YouTube, because that’s where it will be.

Scott: Yes. Again, if there was no YouTube, it wouldn’t be there, it wouldn’t be a place for it. So what a world we get to live in.

James: Scott, this has been great. Great to catch up with you.

Scott: Thanks. This is great.

James: Good to see you. Thank you for your time.

Scott: Absolutely.

James: Look forward to more conversations like this.

Scott: I’ll keep learning from you.

James: Thanks.