Is Baidu the New AT&T? The Basics of Physical vs. Virtual Networks. (Tech Strategy – Podcast 26)

In this class, I talk about networks vs. ecosystems. And how networks have evolved from physical to virtual / digital.

You can listen here or at iTunes, Google Podcasts and Himalaya.

This is part of Learning Goals: Level 4-5, with a focus on:
  • #17 Basics of Baidu and Physical and Virtual Networks

A lot of the AT&T information comes from the book The Business of Platforms.

Concepts for this class:

  • Network Effects
  • Networks and Ecosystems
  • Complementary Products, Services and Platforms

Companies for this class:

  • Baidu
  • AT&T


I write, speak and consult about how to win (and not lose) in digital strategy and transformation.

I am the founder of TechMoat Consulting, a boutique consulting firm that helps retailers, brands, and technology companies exploit digital change to grow faster, innovate better and build digital moats. Get in touch here.

My book series Moats and Marathons is one-of-a-kind framework for building and measuring competitive advantages in digital businesses.

Note: This content (articles, podcasts, website info) is not investment advice. The information and opinions from me and any guests may be incorrect. The numbers and information may be wrong. The views expressed may no longer be relevant or accurate. Investing is risky. Do your own research.

—–transcription below

Welcome, welcome everybody. My name is Jeff Towson and this is Tech Strategy. And the question for today’s class is Baidu, the new AT&T. And this is really an introduction to the idea of physical versus virtual networks, how they’re different, how they’re evolving. And it turns out both of these companies are a pretty good example of that. But first, if you haven’t subscribed, please do so. You can go over to and sign up there. You get a free 30 day trial, so there’s not a big obligation. See how you feel about it. And there’s a ton more content that comes out mostly by email about all of these subjects that only goes to subscribers. So please do that. It’s great, and I really would appreciate it. All right, let’s get into the subject. So networks are kind of a complicated idea, mainly because it’s a bit of a fuzzy concept. So people don’t know. mean a lot of things by it. It’s one of these terms you hear used all the time, network this, network that, social network, physical network, network effects, network externalities, all of that. It’s actually a bit of a not terribly well thought out idea, so I think it’s kind of fuzzy. I don’t tend to talk in terms of networks very much. I usually talk in terms of platform business models, which leverage in. ecosystems, businesses, types of activity that are inherently networked, but aren’t necessarily networks themselves. So I usually talk about platform business models today. I’m going to jump into a little bit of this idea of networks, but I warn you, it’s a little bit of fuzzy. You ask 10 different people what they mean by network. You’re going to get 10 different answers. Now, for those of you who are subscribers, this is going to go under learning goal number 17. So, you know, there’s four levels I’ve laid out to sort of progressing your knowledge, getting better and better. Easy first steps, level two, level three, level four. And within that, there’s usually five to six learning goals per step that take you some time to get through. And this would be learning goal number 17, which goes under level five. So just to know where you are in that, you’ll see that up there. But within this learning goal, I really want you to think about two ideas. So if… If you forget this a year from now, try and remember these two ideas, which are network effects and complementary product services and platforms. Those are kind of the two big ideas under this learning goal. And then also sort of getting the basics down for Baidu, which is an important company. Obviously, it’s a search engine for China. So that’s how to think about that. OK. Now. The two words that kind of get fuzzily messed together are ecosystems and networks. I mean, what is an ecosystem? I don’t know. It could be anything. It could be all the partners, customers, supply chain within a particular ecosystem. Hollywood, you could call Hollywood an ecosystem, not just the movie studios, but the actors and the directors and the tech companies and the distributors. That’s kind of one ecosystem. You could talk about the advertising world as another ecosystem. You could talk about semiconductors. You can kind of talk about it industry by industry. You could consider a suburb an ecosystem. It’s got the shopping malls, it’s got street. streets and roads and maybe a metro and people and businesses. That’s kind of nature is an ecosystem with lots of animals, predator, prey, food, water, resources. I mean, the ecosystem just kind of all depends on how you cut it up and look at it, but it’s a lot of interconnecting pieces, parts, and I generally consider ecosystems just sort of the landscape we’re playing on. And within that landscape, We might build a business that is a traditional business that sells a product or service, or we might build a platform business, which is more in the business of connecting other parts of the ecosystem. I kind of view business as things we are building that live within an ecosystem that was already sort of there, although it can be created by other businesses. And within these sort of ecosystems, depending on how you want to define it, that’s where I kind of say networks are kind of things that exist within ecosystems. That it’s when certain things within the ecosystem really start to connect with each other. Communication is a little bit, it’s an activity that’s sort of inherently about connecting one person to another or one business to another. Roads and transportation are a bit of a network that live within a city because that’s how things move around and go from one location to another. So I always kind of view networks as just sort of part of an ecosystem. And certain ecosystems have a lot of them, like communications and other ecosystems like fast food. don’t really have networks involved in them very much. There’s not a lot of network activity going on in buying, you know, a KFC or McDonald’s supply chain a little bit, it’s not a huge, you know, massive impact on the operations of the ecosystem. So that’s kind of how I just sort of viewed sort of stuff outside the window, ecosystems, networks, they just exist out there. Some of them were built by companies recently, some were built over hundreds of years. It’s just the landscape we play in. Now within this, you can start to build business models that leverage in networks. And platform business model, as I’ve gone on and on about, is a sort of network-centric business model where the primary goal of the business model is to enable interactions between various entities, people, information, computer networks, businesses, whatever. But it’s sort of a network-centric business model as opposed to selling chicken at KFC, you’re selling a product or a service. So I’ll give you a couple of definitions I’ve read. the book Platform Revolution, which is a great book. Here’s a quote from them. They say, platforms are enabling value, creating interactions between external producers and consumers, which I would then add to that. That exists within the ecosystem or network. the platform’s overarching purpose to consummate matches between users and facilitate the exchange of goods, services, or social currency, therefore enabling value creation. So that all kind of assumes that there’s a network somewhere within there that you’re enabling interactions. Networks are pretty much everywhere. Generally, .. the more fragmented a situation is, the better the network’s gonna be. If you’re in a town with only five big businesses, there’s not gonna be a big network between those five businesses. There’s not a lot of network activity between the airlines, between them. I mean, well, I take that back actually because they are kind of transportation networks. Okay, forget that comment. But generally it works better when you have lots and lots of people or lots of lots of smaller businesses or lots and lots of content. A lot of fragmented entities, activity, can connect them a little bit more powerfully. Now, why do people talk about networks? Why does this come up so often? It’s really three things. Number one, networks are about connecting users and their activity. Let’s connect everyone with telephones and then they can call each other. Let’s connect businesses with consumers and they can start shipping them things. Let’s connect cities with roads and then trucks can go back and forth. So there’s this idea of we’re going to connect things with the network. That’s kind of what people focus on. Then they get very excited about the idea of network effects, which is one of the big ideas for today. A network effect. It’s basically a feedback loop. I think that’s a better term for it, but nobody uses it. Where the more you do something, the more it improves, the more it increases, the more the value of it improves, the volume of it increases, something like that. Where it’s basically a positive feedback loop. And the way usually people talk about this, let’s say communications, because I’m going to talk about AT&T. The more people that have a telephone, the better the service is for everybody. If 10 people get telephones, it makes the service better for everyone who already has a telephone, because there’s 10 more people they can call. If 10 more people go to the KFC, your chicken doesn’t taste any better. It’s still the same chicken. A network effect means the product or service itself inherently increases in its value with more users and or more activity. Now I actually think there’s another part to that that people don’t talk about. A network effect can increase with users and activity on the network. It can also create… increase in value with more complementary products, services, and platforms. And that’s the other big idea for today. Idea number one, network effects. Idea number two, complementary services, products, and platforms. And I’ll talk about that. If either of those things happen within a network, the value to the user goes up and it can go up very, very quickly. And the complementary innovations and products people usually talk about are content and software apps. Don’t worry about that. I’m going to go into that in great detail in a couple of minutes. The other aspect to this is… Networks can be physical and I’m going to talk about this why I sort of want to talk about AT&T Because AT&T was a physical network. It was laying down copper lines all over the United States and every other country AT&T would be the US Over literally decades to do this. It was a physical network the same way a railroad is a physical network and Electricity was a physical network where you lay the wires and suddenly every house has activity. You know a lot of the network and network effects and complements, the two ideas for today, they started out a hundred years ago as discussions of physical networks. Now in our day and age we are mostly talking about virtual networks. You know Facebook doesn’t build anything in the real world. It’s all servers and data and connections Ironically a lot of the stuff AT&T built, but their you know their network is virtual PayPal WeChat pay You know, money is not actually being sent, it’s virtual currency. So a lot of the networks we talk about now are virtual networks. And then some companies are both physical and virtual. So that’s kind of why I teed this class up as we’ll start with AT&T, which was a physical network, and we will move to Baidu, which is a virtual network. Okay. Now let’s talk a little bit about the history here. Let’s say we’re talking about railroads. That’s a physical network. And the way railroads evolved were you had a lot of local railroads. You had certain people were building trains, cars, rail lines in Boston. Some were building them in New York. Some were building them in California. And you had sort of these local networks where you could ship things around within the city. And then what happened that was important is somewhere along the way, I think it was the government mandated, they said, look, we got to standardize the rail size here. are five feet apart in Boston but they’re three feet apart in New York these things can’t connect. And this led to one of the first network effects when everyone standardized their rail size and suddenly all the tracks could connect you basically had a massive jump in the number of places you could send your car. You could send your car from New York to California. and people and you had a massive jump in locations but you also had a massive jump in users and activity because suddenly everyone could send their cars everywhere more locations that’s increasing the size of the network and you got a direct network effect which is the more locations you could go to the more valuable the network was to everybody who was using it and therefore more people joined because it was really really valuable and then it turns out the more people that joined or carts you could put your coal in or whatever. So there’s a lot more activity. But that was kind of the first direct network effect. And… it would be what we call a one-sided network effect. The more users and activity, the more valuable. And it also was one of the first good examples of compliments because… It turns out you could add services and products that made the experience for the user more valuable. A compliment is hot dog buns. You go to buy your hot dog. OK, the hot dog is good, but it’s better if you can buy a bun with it because then you can put the hot dog in the bun. And then mustard is another compliment. You know, mustard and buns are compliments to hot dogs that make the whole thing more valuable to the consumer. Well, this was a network that had a lot of value to the user, the shipper, rail rider, passenger, cargo shipper. They started to add compliments which were like, okay, if you’re going to stop at a train stop in Omaha. We should have a local store so you can get off the train and go to the little store and buy some snacks. That made the ride better for the passengers. And we should have a gas station and we should have maintenance people and a repair shop such that if your train becomes damaged you can get the service there and that’s better for the, you know, the B2B, the business sending the trains. So there’s all these complimentary services that were built up along the rail lines that began to crisscross the United States. And it turned out it was banks, service centers, food, hotels, the local bar. Maybe you go into Omaha and you stay overnight, you stay in the local hotel, go to the bar, then next one you get on your train and keep going. So there’s all these sort of, people, they owned all of that. They owned the trains, they owned the services that supported them, maybe they offered, you know, they owned the local consumer services. Those were all those compliments made the train service better. So that’s the idea. Network effect plus compliments. And trains are pretty good example of those I think. And that brings us to AT&T. Okay, so trains would be a physical network, network effect clearly, plus compliments. Pretty easy to visualize. You could say electricity’s the same thing. The electricity companies ran wires to everybody’s house. That’s a physical network. Electricity came in. But it turns out electricity in your house wasn’t that useful until people started making washing machines. Well, that came later probably. Lights, fixtures, appliances, things that you could use that made your electricity service more valuable. So they were complements to the electricity service. Radio stations physical network, you know, you put up the broadband or the towers all over the place TV stations They broadcast television again. Those are physical networks Became much when they had network effects for sure But they didn’t become really valuable until the compliments came along and people started putting out television shows to play on the network So these things all kind of go together network network effect Compliments and compliments can be product services or even additional platforms, which we’ll talk about Okay, so that brings us to AT&T and how this sort of over time began to switch from what we call a physical network to a more virtual network and AT&T really was the sort of midwife to this phenomenon. So 1876 Alexander Graham Bell patents the telephone. copper wires you speak into it, it translates the vibration into an electrical current, transmit goes down the wire, on the other end is translated back to a vibration and you can hear it. Okay that’s 1876. Within a couple years, because he was a pretty good businessman, Alexander Cranbell, he founds a company which was later renamed AT&T in Boston, Massachusetts, and starts to lay down wires that lets you call other people. And the joke is, I mean, this is classic network effect, right? If you call one person, that’s great. If you can call two people, that’s even better. Three people, four people. Arguably the greatest salesperson in history was whoever sold the first telephone, because the first telephone couldn’t call anybody. So I don’t know who that was. We should really figure out who that person was. That’s really impressive. But in the early days, what you saw was a real sort of direct network effect, where there was really one type of user, which was your sort of standard home, and you would call your friends and you would call your family. And that’s pretty much it. This was before there were businesses. This was before there was long distance phone calling. This was before all of that. It was just like, look, you live in Boston, you know, people, family in Boston, you can call your family. That’s kind of the first simple user group. And you have obviously a direct network effect there based on the connection, the more family members and friends that you’re connected to, the more valuable the service is to you. Now here’s where I’m going to start to diverge from maybe some stuff you’ve heard. People talk about communications networks as a direct network effect. Zoom, which we talked about, is a direct network effect. You can call more people. It’s more valuable. I think the language here is and thinking is sloppy. It’s in the early days of a network like this. It is the number of connections that increases the value. But that fades away really quickly because you don’t really call thousands of people in life. What tends to be more valuable than the number of connections is the type of interaction you’re doing once the connection is in place Calling your mother is more valuable to you than calling someone in another country. You don’t know So the activity on the network the interactions that happen on the network That is where most of the value comes from when you start to think about these things not the number of Connections which is how people always think about it. It’s important in the early days by other effects very very quickly. Now the reason I bring this up is because there’s a famous thing called Metcalfe’s Law. You’ll hear about this all the time Metcalfe. And what Metcalfe’s law says is the value of a network increases at about basically what we’ll call N squared. N would be the number of nodes. If you draw out a network, you put nodes, which are like little dots on a piece of paper, and that could be people. It could be phone landline connections. It could be railway stops. The nodes, the N, you could count those up. And if you have, let’s say 10 of them, the value of the network would be 10. squared. Now that’s an approximation. It’s actually n times n minus 1 divided by 2. The argument is if you have 10 people with phones in their house and you add an 11th person, so the number of nodes in the network goes up by 1, the value of the network doesn’t increase by just 1. because that new person can call every single person on that network and every single person on that network can call that one person. So you actually increase the value of the network by N minus one divided by two. If you draw this on a piece of paper, draw a bunch of dots, those are your nodes. Draw lines that connect all the dots and all the possible combinations you can think of. So every little dot connects with every other little dot. And you realize the number of lines between the nodes, the linkages, goes up dramatically faster than the number of nodes. Now he argued it was n times n minus 1 over 2, which is basically about n squared, almost exponential, sort of grows quadratic to exponential, but more than linear. So the value of the network increases far faster than the number of people that join. You’ll hear this cited all the time. I basically think it’s wrong. You know, one, first of all, when they call it a law, Like this is not a scientific law like Newton’s laws where it was tested in the lab and retested. This is just a business idea. They called it a law so people thought it was a lot more tested than it was. You can immediately start to punch holes in this idea. And I think it’s true some of the time, but most of the time it’s actually false. Now the guy who did this, Robert Metcalf, I’m pretty sure, thanks Robert. He was a Harvard PhD. He worked at Xerox PARC in Palo Alto, and he was sort of early involvement in the ethernet, which is when we started to connect computers to each other, which is another type of network. We’ll call that a protocol network as opposed to a physical network or a virtual network. Then he founded 3Com, which is a massive company. list and he basically focused on communications networks with a lot to do with computers and you know made a huge amount of money. Phenomenal. He coined this phrase. I just don’t think it’s right. I think it falls apart almost immediately. First of all, it’s basically an exponential. Anytime someone tells you that something increases exponentially. Okay, so if the value of the network increases exponentially with the number of nodes or users, you immediately ask, okay, until when? Well, in Metcalfe’s law, there’s no… it doesn’t flatline, it doesn’t stop, it keeps going. That immediately tells you this is wrong. Nothing in the world increases exponentially forever. There is no example of anything. You know, if you… Viruses like okay. We’ve got the virus situation the virus turns out It’s very powerful because it grows exponentially one person gets it they give it to two people Those two people give it to two people each then you’ve got four then you’ve got eight then you’ve got 16 exponential growth It doesn’t go on forever because you run out of human beings to infect You know, there’s a rule that if you like, if you take a piece of paper and you fold it in half and then you fold it in half again and you fold it in half again, so it doubles the width of the paper. you’ll basically reach the moon, I forget the exact number, around the 45th time you’ve folded the paper. Okay, why doesn’t that happen? Because nothing grows exponentially forever. So anytime someone tells you they have a law that’s exponential, you immediately say, no, no, no, exponential is a situation. It lasts for a while in some scenarios, which is awesome. Nothing goes on exponential forever, and then it flatlines. So that’s a problem. The other problem is the value part is bad terminology. The value of the network. Okay, what does that mean? Value to who? What are we talking, what does that? If you do finance, when we say value, we’re talking about the financial valuation of a company. So we look at the cashflow. We don’t care how many nodes in the network there are. We look at the cashflow this business generates. discounted to zero at the risk adjusted cost of capital, that’s the value of the company. Well, that’s not what he’s talking about. So that’s not value what he’s talking about. What he’s talking about is perceived value to the user that when someone goes on this service or product, the value to them increases dramatically with the number of people they can call. Okay, I think that’s clearly false. the value to the user, I don’t need to contact 2.7 billion people on Facebook. Somewhere around 500 friends, the value to me as a Facebook user declined dramatically. And it’s not about connections, it’s probably more about activity. It’s not even the 500 people I’m connected with. It’s probably the 50 people I actually stay in touch with and talk to. After about 50 people I’m chatting with every week, after that, the value flatlined. So I think the value to anyone, whether it’s a business or a communications network or a payment network, all that, it flatlines very, very quickly. It is not exponential. And it’s definitely not about the number of connections. So anyways, this thing’s got problems all in it. It’s worth knowing what it is, because people cite it. Don’t pay attention to it. just to continue ragging on other things. There’s another law which came out from David Sarnoff, who was an RCA executive called Sarnoff’s Law, which was about broadcast network because RCA was a broadcast. That basically the value of a broadcast network. grows linearly. So if you’re broadcasting from a tower, your TV show, okay, that’s one to many. The more people that listen to it, it goes up linearly with the number of users that are part of your network. More true, but the truth is, no, not really, because people don’t, you know… The number of active viewers is what you want. You don’t really want user number in general. You want engagement, you want participation. You know, in all of these business models, these digital platforms I’ve been talking about, I keep saying there’s three assets you have to, you watch. Number of users, their engagement and participation and data. Those are the three intangible assets that build a digital platform. It’s really that second one, engagement and participation that tends to be the most important. One last one here, Reeds Law, David P. Reed, he basically argued that the value, and here’s that MIT guy, he said Metcalfe’s law actually understates the value of networks because within a large network, like a Facebook, you actually get smaller networks, clusters, and this is actually a really important idea. You get clusters of people that you actually talk to. When you go to high school, you don’t actually hang out with everyone at high school or college, but you might be very active on the football team. That would be your network cluster within the larger network. And the value to someone actually follows more from the clustering than it does from the overall network size. I think that’s actually a very interesting idea. So chat worms, WhatsApp basically was a cluster phenomenon. OK. So that’s just a little bit of theory. I don’t want to go into that, but you should know what they are. All right, let’s do AT&T. Now AT&T, so I said 1876 and Alexander Graham Bell creates this thing, patents it. They begin laying copper wire down all through city by city by city and start to get a local direct network effect. Lots of people sign up. The only activity they’re really doing at this point is phone calling. So the number of nodes, the connections would be landlines into houses, but the activity, the engagement would be phone calls. connect a bunch of people but people aren’t making phone calls the value of the network is not high regardless of what any law says Metcalfe’s law or whatever. This is why if you join Facebook the first thing they’re trying to get you do on Facebook is to start posting as fast as possible to connect with at least five people and then post. They got to get you connected they got to get you engaged in participating. All right what you know what AT&T did early on was they plus local entrepreneurs, companies with their own investment groups all across the country. So it was a little bit like the railroads that like, you know, you didn’t have one group building a whole national railroad. You had lots of sort of regional railroads being built and then they were ultimately all connected. Well, in this case, you had lots of, you know, 4,000 plus groups putting in the copper wires in Florida and California and whatnot. And they had their own local investment because this actually took a long, you know, took a lot of money to lay all this stuff down. And what AT&T did was they were sort of the long distance connection. So if you’re a user in Boston, you’re probably mostly calling people in Boston in 1890. People in Boston didn’t know people in New York in 1890 that much. I mean you were a local person, so you were mostly doing local phone calling, which was the AT&T product they were selling, was local phone calls. That would be one service fee. I don’t know if that’s how they did it back bill you had local service and internet or long-distance service and then maybe international service. Now AT&T because they were wicked smart you know they controlled the connections between the local groups. They were the long-distance service and it turns out that’s the hardest part of this to replicate was the connections between the local networks. Turns out that’s pretty much the same thing for FedEx and SF Express and DHL. You know if you’re doing SF Express in Shanghai there’s a million groups that can ship a package from Pudong to Puxi. But if you want to ship from Pudong to Chongqing you have to use these sort of between city shipping services. That’s SF Express. That’s actually where the power is. It’s between the cities not the local shipping. It’s kind of the same with these early phone calls. The power the local plan. But their goal really on this first bit was just to get usage, to get people signed up, to get the lines into their house and get people making phone calls. And so, you know, they would basically price their service below the cost of providing the service. And they had to get permission from the government to do this because it’s kind of anti-competitive behavior. But the argument to the government, as far as I can tell, was basically one of network effects. Look, we should be able to price this below our cost. because if more people are on this everyone will benefit because the service will be better for everybody which is a pretty good argument. Now this is back in the 1910-1920 they got about 10% of the United States on their platform over 20 to 30 years. I mean, that’s how long this took to stitch together. And to get 40 to 50% of the US on this system took about 50 years. So it was a long, long haul. and eventually they ended up with a hundred percent of the market. They were the only company that offered phone service in the United States and they had a total monopoly for almost a hundred years which was pretty amazing. Now there’s a lot of reasons for that. They had a lot of government support, regulation, what we call regulatory capture was a big part of it. There weren’t any real substitutes for what they were offering. They had a lot of supply-side advantages, economies of scale, very difficult to replicate assets, things like that. So they had a whole ton of advantages. But the most interesting one that everyone talks about is the network effect. Okay, so I said there was three ideas. This idea of physical versus virtual networks. There’s this idea of network effects, just talked about that. And then there’s this idea of complements. What are the complements to AT&T’s phone network? Well, this is where it gets really, really interesting. because what they would offer you is they would give you things like a phone book. Okay, you sign up for AT&T and you get a phone book and that gives you a list of people you can call. That makes the phone service much more valuable to you. That’s great. And they would give it to you for free, but in truth it was a matter of bundling. you were actually buying the service plus the phone book, but they bundled it together into one price. And we went through the economics of bundling, which is very powerful. And that’s what you were getting, but it felt free. Hey, here’s a free book. Everyone gets it. It made the service more valuable to you. And actually, let me make a quick qualifier here that this AT&T story, a lot of the information coming from this is from a book, which I should give credit to. It’s called The Business of Platforms, Strategy in the Age of Digital Competition, Innovation and Power. like that sub headline that much. It’s a really good book. It’s written by Michael Cusumano, Annabel Gower and David Yoffie. I’m gonna mispronounce his name. I think it’s Yoffie. I will put the link in the notes but this AT&T, a lot of that is from them. So I definitely checked that out. Very good book. Some very interesting thinking on this subject. Okay so white pages, these phone books they give you, you know that started early on. It was a compliment and it made it more valuable. And then the famous story is eventually they decided well we should we should have a book that also lets businesses communicate you know their phone number and their information to people and the person doing that to R.H. Donnelly was printing this and I guess he ran out of white paper one day so he put it on yellow paper and that became the yellow pages. And this is kind of the important thing because White pages, the white phone book that everyone used to get, that’s a compliment. It’s a complimentary product. It makes the service more value. The yellow pages are a complimentary platform. The yellow pages were a platform business model because what did the yellow pages do? It connected people who have phones, consumers, with businesses who want to reach those people. So it’s a two-sided platform. and you know the businesses they would basically charge the businesses. You can put in your advertisement or your listing within the phone book. You know we charge you and then we give it to consumers for free and you can reach them. I mean it’s a two-sided business. It’s a two-sided platform. And because they were charging one of the benefits of platform business models is you can charge one side and give it to the other side for free, which is what they did. So they charged the business, gave it to consumers for free. And this thing was ridiculously profitable. According to this book, The Business of Platforms, 50% profit margins on this product for 100 plus years. And this is the really kind of important point about compliments, that every now and then when you have a platform, like a physical network like a phone line. Or let’s say Alibaba Taobao, which is a marketplace platform. And you create compliments that make it more valuable. The same way a network effect makes it more value every now and then one of those becomes its own new platform And we call that a complementary platform The yellow pages were a complementary platform To the phone network the same way ant financial is a complementary platform to alibaba taobao Which is why it’s one of the reasons it’s so powerful. That’s amazing. This is the um, you know, I called this the uh, the velociraptor You know the that super dinosaur in Jurassic Park where you have the half T-Rex half Philopsoas rapper who is the Indominus Rex and used to kill everyone. That’s a digital platform. A complementary platform is like having two of those hunting together. It’s just a lot more powerful. Okay now let’s transition over to Baidu. I mean the AT&T story which is why I think it’s a great story is it shows kind of the main points but it began as a physical network. phone lines. But then what happened? Eventually, people started to run data. through these same lines and it wasn’t just voice, it was data packaging going back and forth. Once that began to happen, this physical network that had been synonymous with voice calling was able to carry virtual networks where suddenly you’re just sending information back on the same lines, but you’re not really dependent on the physical network as much. And this is when these phone companies started to get hammered because, They used to be the kings. The phone network was the most powerful network you’d ever seen. But now suddenly, fax machines are sending across, Facebook is sending connectivity messages to your friends across, WhatsApp is sending messaging across for free. And suddenly the phone lines, their greatest fear sort of became, are we just gonna be the pipes? And we don’t really have the network anymore. We just sort of provide the infrastructure and all the value has moved up into virtual networks that sit on what we built, but we don’t control those. I mean, when is the last I mean, how often do you make a real phone call using your mobile network anymore? I don’t make them that often. I’m more likely to use WhatsApp or WeChat than to actually use their service at all. The only thing I’m really using is their data network. And that, you know, most of the values in the virtual networks that these other digital giants control. So they were kind of the midwife from physical networks to the world we live in today, which is mostly about virtual or digital networks. And, you know, that sort of brings us to the story of Google and Baidu, which are search engines. which if you look at them, they look a lot like the yellow pages. They look like a digital version or a virtual network version of the yellow pages. I mean, what did the phone company do back in 1900? You know, they connected everyone fine. Everyone’s got a line in their house. They did that. And then you had to find people. But how do you find people? I don’t know, you remember all their phone numbers? You can’t do that. How do you look up someone’s phone number? Well, you use the white pages or you use the yellow pages. I mean, those were kind of physical search engines where you flip through and find the number and then you call them. Well. Once web pages started to emerge, and it turns out there are billions and billions of web pages that are getting created, and you’re sitting on your laptop, this is like, you know, 2003, 2004, how do you find anything? Well, you kind of need a virtual phone book. And it turns out most of the connections are not people to people. That’s more like email. Most of the connections are more like the yellow pages. It’s I’ve got to find a business or an institute or something or information or a page. So it’s just like the yellow pages. It’s just, you know, a million fold larger. And that jumping into that problem. is Google and Baidu, Baidu in China and Google in the US, where they start to do the same connection that the Yellow Pages did. And they basically adopt the same business model early on, which is we will be a two-sided network. that connects businesses or really anyone who’s put up a web page on one side and then users which were mostly PC users at that point. Those will be the two user groups on our digital platform and we will, the first thing we will do is we’ll handle search because that is their biggest problem is finding each other and we’ll build all the algorithms that do all of that. And then second to that we will put an ad network on it and we will start to charge not to be listed like a yellow pages, but to be found. Because you search for something, it shows up, then you want to be in the top, or you want to be listed, or you want to be advertised, or whatever. So they varied it a little bit, where they focused a lot more on search, obviously. But the original business model of Baidu was pretty similar, and you know, the difference between Baidu and Google. is really language. I mean that’s why these things diverged early on was one was for the Chinese language and one was for English and then other languages but that’s where Baidu really dominated was just Chinese language search and Chinese language web pages. Now that was I would call that Baidu 1.0. And do they have a network effect? Well, obviously, we call this an indirect network effect. The more webpages that standardize their formatting so that they’re easily searchable and get listed and shown in ByDo, the more valuable that is to people who are doing searches and vice versa. If you’re gonna standardize your webpage and make it fit the parameters that ByDo wants, the more consumers that use ByDo, the more valuable that is to you. So obviously, definitely helping other because that’s the whole purpose of search. And it also turns out there’s a lot of long-tail effects that are pretty interesting. The bigger the search engine you are, the more you have lots of tiny niche subjects. It’s actually pretty easy to replicate a search engine for very commonly searched items. That’s not that hard to build what ByDo and Google Do Well is they give you great results for all these tiny little niche searches that nobody does. But they have such scale that they can do well on all the long tail searches as well as the common searches. So we’d say they would OK, they have a network effect that sort of ByDo 1.0. And then it gets a little more interesting because it turns out they have at least two other network effects. And these are kind of important ideas we’re going to get into in later topics. The first is they have what we call a standardization network effect, which is, you know, this is, we talked about this a bit with coordination platforms, which is this idea that the PDF format or the word doc format or the MPEG format. When more people use these formats where it becomes a standard, it makes it more valuable for everyone. If we’re all saving our documents as PDFs, that makes it more valuable because then I can open your documents and you can open mine we can send to each other. So when you have a standardization effect, it can often create a network effect. The more people that use PDF, the more valuable PDF is to everybody. If we were all using different formats, it would be a real pain and we’d have to convert all the time. Well, that’s kind of what happens on the webpage side of Baidu. that as web pages start to do their SEO, their search engine optimization, when if you have a web page, you optimize the way you lay it out, the way you tag it, so that it is easily searchable by, in mine’s in English, so it’s Google. But if you’re in China, you’re standardizing the format of your web page and the way it’s tagged and categorized and the way you put in information to make it easily searchable for Baidu. So there’s a standardization effect web pages that are standardizing themselves to what Baidu looks for. The better Baidu works. And then also it’s easier for those web pages to work with each other. Suddenly there’s consultants who are very good at SEO optimization. I mean, SEO stands for optimization. But there’s more there’s consultants, the more there’s tools that help you do this. The more we all agree that this is the way we’re gonna set up our webpages and standardize, the more valuable it is to lots and lots of people in lots of ways. So there’s a standardization effect going on on that side of the platform with webpages. LinkedIn actually has the same effect going, which is kind of interesting because nobody ever talks about it. LinkedIn has several network effects. They’re a pretty powerful platform. And one of them that they have is, if you’re a business person or a professional, you kind of have to have a LinkedIn page now. Like, you don’t really send CVs and resumes anymore. People look up on LinkedIn to check you out. So it’s kind of become the de facto standard for your professional profile. So you kind of got to have to have to have it and then because everyone has it everyone knows that’s where you go to look and it becomes more valuable because you know people will go look so it’s kind of another standardization network effect. So Baidu 2.0 has you know this matching function this searching function between searchers and web pages it also has this additional network effect on standardization side for web pages and there’s actually at least a third one there’s several because Baidu’s got other businesses but in their search engine the other one they have is they have what we call a sort of a learning effect or people call this a data competitive advantage or a data network effect. Some people call it a learning platform. It’s basically this idea that if you go and search on ByDo it will recognize it’s you by your IP and the more you search the better results it will give you because it learns from your activity. So the more I search on Baidu, the better my results get, personally. But at the same time, the more people that are using Baidu collectively in the aggregate, that also makes the platform smarter. And so it’s starting to give better results to everybody. So there’s a bit of a, I mean, this is kind of a fuzzy idea. And I think the thinking on this is not great yet. Turns out this is a big deal in AI learning platforms. Some people call this a learning platform. Some people call it a data network advantage or a data network effect. Some people call it a learning or a data competitive advantage. There’s a lot of weirdo terms, but you definitely see it in search engines. So, ByDoo 2.0, least three network effects going at the same time. And it’s completely virtual as a network, although the core thing is basically a digital version of the Yellow Pages. Alright, I think that’s probably enough theory for today. The goal of this, this is learning goal 18 for those of you who are subscribers and I really wanted to take the first steps into this, but not go too far. We’re gonna hit this again, we’re gonna go to more complicated examples, but I wanted to kind of cover the basics of Baidu. but put it in a framework that makes sense. So I wanted to go through a little bit of the basics of physical networks and how they’ve transitioned into more virtual digital networks over time. Now we’re gonna go into a lot more of this because there’s a whole lot here and Baidu is actually a lot more complicated than I just laid out. But I think that’s probably a pretty good pass for today because I’ve been talking at you for about 45 minutes now. So just to reiterate the key learning goals for today, the basics of Baidu and physical versus virtual networks and the two main ideas to walk away with this, network effects and complimentary products, services and platforms. And complimentary platforms are really pretty awesome. They’re rare, but you should always keep an eye out for them. And I guess a couple of fun things just to add up, it’s Sunday night here in Bangkok and… I think we’re finally coming out of it here. It looks like the shopping malls might be opening up in the next week or so. I sure hope we’re coming out of this in some form. I’m dying to get back to work and get going. I am now like a master of driving the Batmobile in Arkham Knight. Like… This is not a skill set I needed at this point in my life. Like, it’s ridiculous. Like, the stuff I’m wasting time on. Oh, here’s a terrible idea. Here’s an absolutely terrible idea. There was a new show that started streaming on Netflix over the last couple of days called Too Hot to Handle. And I turned this thing on and watched four episodes. And man, I don’t know whether to tell you to watch this thing or like run screaming in the other direction because it is both like awful and awesome at the same time. It’s like I stopped when I finished watching these episodes. It was like I spent two hours just eating bags of potato chips. Like there was some level of enjoyment in eating the potato chips. But when it was over, I just felt awful. Like I don’t I don’t even know what to tell you about this thing. Like you got to. check it out a little bit. But for those of you who don’t know this show it’s a reality TV show and boy I hope this is not true. I hope it’s something they’ve made up in life but it basically took about 10 men and women who are apparently super attractive and very active on Instagram and all these dating sites and whatever and are just getting a ridiculous amount of romantic attraction from the opposite sex, you know, just overwhelming because I guess they’re really good looking and I guess if you’re really good looking and you’re on the site, it’s just like a tidal wave of, I don’t know, getting hit on or something. And so there’s these people that just do this a lot. They took 10 of these people, they put them on an island and then they let them all just like go crazy and then sort of the twist on it, not to give it away, which it’s in the title of the show, is they forbid all romantic interactions and these super attractive people that are like doing this all the time have to actually get to know each other and like talk to each other. And it’s just awful on every level. Like the discussions are awful, the people are awful, but it’s also like, it’s hard not to watch it. Anyways, I really don’t know what to make of this. I’m sort of hoping that this is something the TV show made up and this phenomenon does not really exist online, but I suspect it does. Anyways, I leave it to your discretion. This is both like a, it’s both very interesting and very awful at the same time. So you know, be warned. I’m not watching the other episodes. I stopped. I’m like, no, this is bad for me. I gotta stop watching this. Anyways, that’s it for me. Have a great week everybody. Stay safe, mask up, wash your hands, all that good stuff. And hopefully we’re almost through this thing. But thank you for listening and I will talk to you next week.

Leave a Reply