How Digital Creates and Destroys Competitive Advantage. Digital Economics II. (Tech Strategy – Podcast 55)

This podcast is more about the strange but important economics of digital. And on how digital economics can impact well-established competitive advantages like economies of scale. This is a continuation of Podcast 53.

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

This is part of Learning Goals: Level 7, with a focus on:
  • #29: Digital Economics II

Concepts for this class.

  • Digital Information and Economics
  • Economies of Scale
  • Minimum Efficient Scale
  • 5 Scenarios Where Digital Impacts Economies of Scale
  • Network Effects
  • Standardization And Interconnections Network Effects

Companies for this class:

  • n/a


I write and speak about digital China and Asia’s latest tech trends.

I also teach Jeff’s Asia Tech Class, an online course and daily commentary for busy executives on Asia tech and China’s digital leaders.

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 topic for today, how digital creates and destroys competitive advantage. Digital economics two part two. Two weeks ago, I gave you a talk about sort of digital economics and at a little bit more of an advanced level. I’ve covered this topic a couple of times in podcast 15. which was sort of intro to digital economics, and then podcast 53, which was the one I just mentioned. And that was really focusing on why certain economic phenomenon mechanisms that are kind of secondary become much more powerful when the business or industry starts becoming more digital, more about information, software, things like that, like pricing, like bundling. like compliments, things like that. A lot of these relatively minor effects that no one talked about became much more important. And I gave you several versions of that, price discrimination, combinatorial innovation, switching costs, compliments, versioning and pricing, blah, blah, blah. I’ll summarize that a little bit. But today I wanted to kind of build on that and talk about, okay, what else does digital really sort of impact and… One area which is sort of the center of my research is competitive advantage. How does digital create or destroy competitive advantages? Many of which we’re very familiar with and suddenly we thought a company was terribly strong and entrenched and a little software enters the space and suddenly their defenses get wiped out. Newspapers are probably the most stark example of this because their competitive barriers were so strong for so long. And with a handful of years, you know, now newspapers, daily newspapers, really difficult business. But it used to be just about the best business model you could find anywhere, not that long ago. So today I’m gonna talk a little bit about economies of scale and network effects and sort of how the economics behind those is changing, mostly because more and more industries are getting impacted by software and data technology and all that. Okay, first before I go into that little housekeeping stuff, first of all, I had really great session yesterday here in Bangkok with, you know, members of this class, subscribers and about 15 of us met something like that in the boardroom here in Bangkok and started talking about you know, a couple of people, they presented their case on what they thought others argued with them. So it was really great. I mean, it was a lot of fun for one. But two, it was already, you know, it was really satisfying to shift away from me just talking, which is part of teaching, obviously. But, you know, the most important part is when people attempt to do stuff theirself. That’s when you really learn. So it was great to see people doing that and sort of arguing a little bit. And using some of my language to describe things, which was great. I really enjoy that part. So anyways, thank you to everyone who was here yesterday. That was real fun. I appreciate that and you know for those who weren’t there, which was the vast majority of people, really what we’re doing, what I’m doing, is I’m sort of trying to find an interactive forum that we can start to build into the class. Maybe we’ll start live streaming these, maybe we’ll start having people come in and log in and we’ll we’ll do this sort of session so it can be more of a community and more interactive and like that. And this was what I when I was sitting there yesterday sort of thinking about is, you know, how could we bring this to everybody, what’s going to work, what’s not working. So it was a little bit of experimentation. And it was hopefully I’ll be rolling out some of that to everybody because I think that process is really important. And plus it was a lot of fun. So thank you to everyone who came out yesterday. That was fantastic. And I’ll see you in two weeks when we’re talking about JD. So I don’t know if the team that’s presenting has been decided, but anyway. Okay, so that’s kind of housekeeping number one. Housekeeping number two, I’ve been talking to quite a few of you about what’s working, what’s not. And one of the common things I’m hearing is it’s just too easy to get lost in the content. A lot of ideas, a lot of concepts, a lot of companies. You know, and it’s just, it can be confusing. So as I mentioned before, I’ve been working on sort of an Uber graphic, not Uber, the company, but like a big graphic that kind of lays everything out in one picture. So it’s a little bit of a roadmap. So when if I cite something, you can look and say, oh, that’s what we’re talking about today. And here’s the parts I know. And it’s like filling in almost like a bingo card or something where you can slowly fill in the squares over time as you as you sort of build up your knowledge. And I think that’s about done. And so I’m going to be sending out. versions of that to people for feedback. You know, does this capture it? I think we’re about done with that. But you know, for those of you who might be frustrated like about this subject, I have heard you, I am working on it, almost done. Hang in there. It is a lot of content. I know that and we’ll try and have a nice clean framework so it’ll all sort of fit together and you can see where all these pieces fit. Anyways, someone brought that up to me yesterday and I, you know. I’m definitely, it’s priority number one is to get that done. Cause a lot of, you know, not a lot of people, but several people have mentioned it. Okay. That’s it for housekeeping. Let’s go into the content. And for those of you who aren’t subscribers, feel free. The free 30 day trial, try it out, see how you like it. And yeah, it’s, it’s, we’re getting there better and better, I think. Oh, that’s my feeling. Okay. Now, for those of you who are subscribers, this is all going to go under Learning Goal 29, Digital Economics 2. And the two ideas, well, three really, the three main ideas are economies of scale, network effects, and then standardization network effects. The first two we’ve talked about a bit, and I’m gonna sort of talk about how digital impacts those and then standardization network effects is actually kind of a big deal. And I haven’t touched on it much, so that’ll be the third one. Anyways, this is all in the show notes. In the last talk, which was podcast 53, I talked mostly about this idea of price discrimination and how as things get more granular, more data driven, our ability to price and the strategies behind that are naturally getting more complicated. It’s not like you just put one price on everything and you put it on the shelf anymore. You can have different versions of different types of software. You can have different prices for different people because you know a lot. I mean, we’re moving away from the mass market. and we’re moving away from mass market products to different versions, customization, personalization, market segmentation, all of that. You’d expect for pricing to get quite a bit more complicated and it has, and I talked about things like market skimming pricing and penetration pricing and all of that. It’s a big subject and it’s really important. And there were kind of two related ideas that also are becoming more important. One was switching costs, which is a type of competitive advantage, depending how strong it is. I mean, if you really lock someone in, that’s a tremendous competitive advantage. If it’s just switching your accountant, there are some basic switching costs there. It’s a pain, it adds a little bit of difficulty, but it’s not terribly strong. So these are all great agents. But again, really powerful switching costs tend to have a significant impact on pricing. because it’s almost like a monopoly. Like, okay, you fight for your customer and you’re fighting with your competitor for your customer. Pricing is a big part of that. That’s sort of round one. But then whoever wins round one, let’s say you’re buying an ERP system for your company and you choose Microsoft and that’s the win wins round one. Well, once you put them in, then it’s kind of a monopoly situation where they have you and you only have one person. So the pricing between sort of pre-decision and post-decision is actually very, very important. You might be incredibly aggressive on the pricing on the first step in order to win that and then get yourself into a monopoly position with regard to that one client where you can raise prices a lot later. So there’s a lot within switching costs and pricing. And one of the things software does very, very well is you can build in a lot of integration with companies, with… people, you can host all their files in online storage, you can run their ERP system or their CRM system for their small business, or you can be their payment processor and suddenly you’re handling. There’s a lot of ways with software to really sort of lock yourself in. So the pricing there, that was my point last two weeks ago was the pricing is very important in that area. And then bundling as well because it’s really easy to bundle digital goods. That’s what Spotify is, that’s what Netflix is, that’s, you know, these are all just bundles. It’s easy to put software and digital goods together, much easier than say physical products. So there’s a lot of important pricing stuff in all of these. That was the, that’s the quick summary of the last talk. Okay, the new stuff. Now, I mean, this is kind of my strike zone, which is, this is what I’m writing a book on if I ever get to it, which is how does digital create and destroy competitive advantage? This idea that you’re gonna have a barrier, you’re gonna have a protection. You know, how can you defend it against a digital phenomenon? How can you use digital tools to build one? How can you use it to attack someone who’s got one? There’s a lot of interesting dynamics there. And economies of scale, okay, that’s a good first bucket to talk about, because that’s something people have been talking about economies of scale for 100 years. You know, if you have a factory, you wanna be twice as big as the other person, because that gets you lower per unit cost for every widget, bigger in terms of volume. You know, part of your business has a fixed cost. And so if you do more volume, the per unit fixed cost is less. And that’s factories and things like that. Okay, that’s a very common competitive dynamic, competitive advantage. How does digital impact that? Well, I think there’s four or five ways that digital can sort of impact that situation. Now, when you’re talking about economies of scale, The standard explanation you’re gonna hear about this is you have sort of a big fixed cost and usually a big upfront cost and then low marginal cost. So you have to build the factory, that’s a big fixed cost. And I’m sorry, big upfront cost. And then when you have the factory, a lot of the operations, staffing it up, having maintenance people, all of that is sort of a fixed cost. And then as you increase the volume you go through, the marginal cost per unit drops. And the more volume you push through, the marginal cost in theory should drop because you have more fixed costs than variable costs. That’s kind of the standard definition. People have been talking about this in software companies forever because they realized companies like, I don’t know, Microsoft, Adobe. had a lot of upfront costs because you have to code the whole system. But once you did that, then you had very little marginal costs. Like once you have Adobe, you can just give away copies for free. I mean, it was a very extreme version of that same factory example I just gave you. And you can get a tremendous amount of fighting in that first phase because everybody knows whoever gets to scale first, especially in things like software, even something like semiconductor foundries and you know when you’re making electronic parts oftentimes the marginal production costs are quite low even though it’s a physical good. Most of the costs are in that upfront development design phase and the fixed cost of building you know a foundry or whatever. So first movers big deal getting the market shares a big deal. So software is just a more extreme version of this economies of scale argument we’ve been hearing for a century. Now, I don’t think that that’s the standard explanation. Oh, let me cite some of this. A lot of what I’m talking about today, this is not my thinking. I don’t want it to come off like this is my thinking. I’m citing other people. This is their work. The person I tend to follow is Hal Varian, who I’ve mentioned in a previous podcast. He’s kind of a guru on the economics of information. I mean, he literally is the guru. He’s an economist. He focuses on microeconomics and information economics. He was the chief economist at Google. And he was the founding dean of the School of Information at Berkeley. I mean, he is the guru. And 60% of what I’m saying today is coming from him and some other people. So I don’t wanna take credit for that. Okay. Although when we get into the competitive stuff, it tends to be more me. Okay. I think that the explanation I gave you on economies of scale there is not quite right. I think it’s fine. I think it’s first past thinking. I think there’s some subtlety there that are important. The first bit is when you think about having economies of scale as a competitive advantage, there’s actually two different things going on. And Michael Porter, I think, nailed this when he said there’s five forces, one of them is rivalry. competitive rivalry, this is you fighting against your existing competitors, and then sort of market entry, barrier to entry. How easy is it for someone to jump into your business? And those are actually two things. Now the explanation I just gave you kind of lumped those all together and you really have to do think about them separately. You really do have to think about them separately. There are certain businesses where it’s actually quite easy to jump in. making a cola. That’s not hard at all. You buy some cans, you buy some water, you pump it full of sugar and caffeine, you get your buddy down at the grocery store to put it on the shelf, bam, you’re a cola company. Very easy. You don’t have to do it global, you don’t have to do it nationally, you can just do it in your neighborhood. Very low barrier to entry. But once you get in there, the competitive rivalry, the fact that you’re against Coca-Cola, they have very powerful advantages there, but not on the entry side. A counter example to that would be, I’m gonna start a railroad. Now starting a railroad, there are some economies of scale, but not really actually, it’s network effects more than anything. But really the biggest problem with starting a railroad is you have to get the land and lay the tracks. It’s not even that it’s expensive, it’s just difficult, because you can’t get the land to build a railroad network because the cities are all built up and the roads are in place and there’s homes built. If you didn’t get your railroad tracks in place 50 years ago or 100 years ago in the US, it becomes almost impossible. The difficulty of entry can sometimes be about cost. If I want to replicate an operating system or building a hospital, there’s a barrier to entry which is about cost. That first step is expensive. whereas the first step with launching a cola is actually not expensive at all. But it’s not just that it’s expensive, sometimes it’s just difficult and can’t be done. Like it’s very hard to lay railroad tracks. It wouldn’t be even be about the money. It would just, you couldn’t get everyone to give you permission county by county, street by street to build a railroad network now. So when I look at barrier to entry, that first step, I think about cost and or difficulty. You know, building a company like Pixar, where we’re gonna make animated movies, like Frozen or Toy Story or whatever. Yeah, there’s a lot of costs in that, but it’s also a very rare skill. It’s actually kind of hard to do that. It requires a lot of talent. So there’s a difficulty in doing things that is often not just about, oh, it’s expensive. So that there’s a barrier there you have to think about. And then if you do get into the business, sometimes, okay, I’m in the business, but I can’t beat Coca-Cola because they have economies of scale in marketing and distribution, and they just keep pounding me month after month, and they drive me out of business slowly. So they have economies of scale there. So there’s different phenomenon, and some companies have both. Some companies have one or the other, depending what you’re talking about. building a hospital in a small town, let’s say 10,000 people, in the mountains, they don’t actually have any competitive advantages in terms of ongoing rivalry. The problem is there’s just not enough people there. There’s an upfront cost, there’s a barrier cost. You gotta spend a lot of money to build a hospital. And then when I do start the hospital, there’s not enough people there to support two hospitals. So I have to steal the patients from the other hospital. So there’s a lot going on here, but one of the factors to think about is. Is the market circumscribed and is it growing? Economies of scale usually require a circumscribed market where I’ve got the hospital in the small town in the mountains. It’s a very demarcated circumscribed market where if I have economies of scale there, a dominant position, there’s not enough extra room in that market for someone else to get in there and get to my scale. Now, if I’m in a hospital in downtown Atlanta, that doesn’t work because the market is quite big. There’s room for quite a few large size hospitals and the market is growing. So you want a circumscribed market and you want no growth. That’s a really powerful scenario to have economies of scale. If there’s growth, there’s often room for someone else to get in there. You gotta have both. Economies of scale is relative. to a market size, not just bigger in general, but bigger than my competitor, relative to a clearly demarcated market. You gotta have all of that. And the other thing you kinda need is you need a management that is actively defending. You can’t just be a big hospital in the mountains, and if someone starts a small little one, you have to respond. You have to drop your prices. You use your advantages of scale to drive them out. If you just sort of turn a blind eye, they can fight their way in slowly and then you turn around one day and you realize they’re 80% of your size and your ability to beat them is gone. Because it’s a relative power. If someone gets to my size, my ability to beat them with economy scales goes away. So it requires management who actively defends. Okay. And one last point on this economies of scale. Economies of scale don’t go on forever. As you get bigger and bigger, you don’t get more and more powerful. It’s a great mechanism, but trees don’t grow to the sky. Every good mechanism eventually sort of gets some negative feedback, some negative issues, and it starts, nothing works forever, right? So. You know, economies of scale usually can get a, in an industrial age, economies of scale might get you 50% of the market, they might get you 70% of the market. We rarely see economies of scale business on the production side getting you 95%. It doesn’t work beyond a certain point, and I’ve talked about this before, that one of the things that happen is as you get bigger and you get more scale, your business gets more complex. You have a lot more people. Your complexity goes up. You get bureaucracy. You get political infighting. And that seems to be something just naturally that happens in human organizations. There aren’t really businesses with 10 million employees anywhere in the world. There seems to be some limit of how many people we can put into a company before the dysfunction grows. and overshadows the advantages of scale. So the disadvantages of scale at a certain point overtake the advantages of scale. And I brought this up when I was talking about AI and machine learning as an operational dimension, a competitive dimension. It was one of my smile marathon dimensions. And maybe the gold standard there is what we’d call zero human operations, which is what Ant Group is trying to build. Where the core processes of your business don’t have any people involved. It’s all software and they might well achieve a situation where they have all the advantages of scale, but none of the disadvantages because there’s no people. I’m keeping my eye on that one. Okay, that’s economies of scale. Now, how does digital impact this? I think five ways. Some of these are from Halvarian, some of them are mine. First one is the one I mentioned is economies of scale in a business that has zero human operations where the people, I mean, there’s gonna be people designing the system, monitoring the system. But when you go on YouTube and watch videos and you click one after the next, after the next, there’s no people involved in that. It’s just happening by software. There are YouTubers who are uploading videos, they go live pretty much on their own. unless someone flags them for moderation check or something. And then you watch them. So the core part of the operation doesn’t have human involvement too much, although content moderation is a big problem. Things like that, I mean, payment processing, maybe, although fraud is another issue there. And what Ant Group is trying to do is to be able to offer loans to small businesses and consumers within the Alibaba ecosystem. where the decision is made without any human involvement. And in theory, they could offer millions of loans every day with no loan officers, no branches, and no traditional bank could do that. So there’s this idea of, could digital sort of turn economies of scale into something new we haven’t seen, which is… If we get to zero human operations, will we see these advantages of economies of scale can grow further than we’ve ever seen because the disadvantages we’re so used to in the industrial age don’t kick in? Could we see businesses that are much larger than anything we’ve ever seen before because they’re not based on adding more and more people? Maybe. That’s one of my five sort of scenarios I’m looking at. Another scenario, scenario two, is this idea of minimum efficient scale. One of the reasons economy of scale doesn’t work beyond a certain point is, if I’m a factory and there’s another company that’s a factory and I’m twice as big as they are, or three times as big as they are, and then someone else enters the business. my relative size advantage doesn’t go on forever. At a certain point, a new entrant or a smaller competitor can get big enough to sort of what we call minimum efficient scale, where once they get to that point, they’ve pretty much captured all the advantages I have. And then even if I’m bigger than them, it doesn’t really matter anymore. Once they get a factory of a certain size, their cost structure looks pretty similar to mine, even if I’m five times bigger. there’s a minimum threshold where you’ve kind of hit the efficiency level. Now, one of the things digital does is it is very good at lowering the minimum efficient scale to compete in this way. So the famous example is laptops. There used to be sort of economies of scale and publishing. You’d have a publishing house, you’d have a local newspaper, you’d have to build the printers, you’d have to, you know, all of that stuff. Well, there was a certain fixed cost to doing that where you could do it efficiently. You had to get to a certain scale, but suddenly if you give people laptops, they can start writing books on their laptop with Word and they didn’t need all the typesetters, typesetters and editors and printing and all of that. So the minimum viable scale dropped dramatically. And then, you know, I can sit at a laptop and write articles, and I can pretty much compete with a newspaper. I mean, I can compete with major newspapers because the minimum efficient scale of operation has dropped dramatically because of the digital tools. So that’s kind of important. One that people talk to talk about a lot in Silicon Valley is Amazon Web Services, that it used to be if you were going to start. you know, a new tech company or a software company, you had to raise a couple million dollars just to buy the servers and to get the hardware in place where then you could start to code and try your idea. Oh, we’ve got a or whatever. Well, you don’t have to do that anymore because you can just create an account at Amazon Web Service and buy whatever hardware access or storage capability or computing capability you need. as you need it. I mean, it really wiped that out. And now, you know, it used to meet, let’s say early rounds for startups used to be one to 2 million, and now it’s like 100,000 because you don’t need that money anymore. You know, I get a bill from Amazon Web Services every month. I think my last bill was 36 cents. I mean, it’s ridiculous. And whatever I use for hosting videos or podcasts or traffic or whatever, it just scales up automatically. I mean, it’s really great. And so when you think about minimum efficient scale and if digital is gonna change that. Think about my earlier point of this advantage has two components. It’s got the barrier to entry and the ongoing rivalry. Amazon Web Service really wiped out that first bit, which is that getting into the business, that upfront cost got wiped out. It didn’t necessarily hit the ongoing cost. So this phenomenon can hit either side of that or it could hit both, but you kind of want to know what it’s doing. Okay, so that’s kind of scenario number two. Scenario number three where digital impacts economies of scale. Let’s say we have both of the previous things I mentioned. Excuse me. Let’s say we have a growing market because maybe digital phenomenon have taken a stagnant or slowly growing market and started to expand it. Shopify has done this. Actually, Ant Group is doing this. They’re going into a business like credit, which has traditionally, at least in China, been between major state-owned banks and large, mostly state-owned enterprises. They’re coming in with credit technology that is expanding the market by letting SMEs and small companies access credit in China, which has been a very big limitation of the economy for a long time. So… The digital tool is growing the business, growing the market. And as I kind of said, like if your advantage is economies of scale, you really want a circumscribed market without rapid growth, because rapid growth gives the opportunity for someone else to jump in. Okay, well now we’re using digital to turn a slow growing market into a fast growing market. And then in addition, we might have this phenomenon where it’s also dropping. the cost structure, maybe it’s dropping the minimum efficient scale required. So we could be seeing digital doing both of those things at once where it’s starting to grow a market in a way it hasn’t before. And it’s also starting to drop the cost or maybe the minimum efficient scale. And the net result is that’s a great opportunity for you to jump into a business with an incumbent with economies of scale and just leapfrog right past them in one move. That’s a pretty cool scenario. And I think we’re seeing this in ERP and we’re seeing it in companies like Dropbox and and Stripe and even Square and Ant Group where they’re both dropping the cost structure and growing the market at the same time using digital tools and then they’re jumping in. Pretty great move actually. Okay, scenario number four. What happens when we have… a business that has durable goods, which is like, you know, I’m selling you a sofa. Now one of the problems with selling sofas is, you know, people buy it, it’s a durable good, they keep it for years. So there’s a lot of intense competition to get the, it’s not like the person’s going to come back next week and buy something every week. You know, it’s a one shot deal. And so there’s a lot of fighting to get that good. Okay, that’s pretty common. We’ve seen that in a lot of things, motorcycles, Harley Davidson, furniture, a lot of durable goods have this sort of difficult economics, or not economics, but difficult buying behavior, purchasing behavior. Okay, what happens when we have a business like that and digital starts to make it more innovative or starts to make the cost structure drop? And that could be something where a traditional good, like buying a motorcycle or a scooter, which is kind of a hardware business, suddenly you start infusing that hardware with software and it changes the nature of things. And we’re seeing this a lot in China with smart devices popping up all over the place. Okay, if we infuse software into a durable good and make it more digital, could that make it suddenly? more innovative, where we’re seeing ongoing innovation and upgrades, which is kind of what Tesla is doing in cars. Cars used to be this bad scenario where you sell someone a car and then hope they come back in five years and buy another one from you. Tough business. Kind of a hardware business. But I mean, the Teslas are sort of hardware meets product. So yeah, you’re selling them the Tesla, but then suddenly you’re giving them software updates like all the time and you’re adding new features all the time. So you’re really sort of changing the way durable goods have traditionally been purchased, which had a lot to do with like a furniture business had a lot to do with economies of scale in the showroom. that people buy furniture very rarely, they wanna see it, they wanna sit in it, so you have to have a massive showroom that costs a lot of money because the inventory, in this case, happens to be big and expensive. So really big furniture stores tended to dominate cities, which is one of the reasons Warren Buffett bought furniture stores. Big competitive edge. Well, now you’re kinda changing those dynamics, you’re making it more innovative. And so suddenly, if you’re selling a more digitized durable good, let’s say like a smart car, your problem is you wanna sell them another car before the current one is obsolete. Your biggest competitor is not another car company, it’s your current customer that they already own your car and you’re trying to get them to sell that one and buy your new one because you’re advancing faster than the durable good sort of depreciates away. So it’s almost like you end up competing with yourself. as a company. It has to do with like this idea of durable goods versus how rapidly the products and features change and some of them you can update. Or often you can release a version of the good that’s cheaper. Okay, is that in my interest? If I’m selling someone a durable good, do I wanna come back a year later, even though they’re gonna keep this thing for five years and say, oh, by the way, here’s a better version of what we just sold you a year ago and it’s cheaper. There’s a lot of interesting dynamics in this idea of durable goods as they become either more innovative or declining cost and you end up kind of competing against yourself. It’s really kind of fun to watch that play out. Anyways, that’s a whole bucket of scenarios, but keep it in mind, durable goods is pretty cool to think about. And last scenario, and that’ll be it for economies of scale. You’re selling a good, and one of the things that happens with digital, with software, is you’re no longer selling someone just one product. You’re selling them the product and its ability to connect with other things. You’re basically selling the cost that is being presented, which you would hope to have economies of scale in, is not just your cost structure, it’s your cost structure plus all of the compliments that go with your product together. That is actually the cost structure you want economies of scale in. Cause that is the price that is perceived by the customer. It’s not just your bit, it’s yours plus all the compliments. It’s your car, you’re selling the car, but you’re also selling a warranty and maintenance and service and software upgrades and maybe who knows what else. That is actually the cost structure that the customer perceives and when they look at the price. As things become more connected, because that’s what digital does very well, is it connects things. The economies of scale you’re trying to achieve is no longer just under your control, it’s you and your compliments. And so what all these players try to do, which is really funny, is they’re all trying to turn their compliments into cheap commodities. And everyone’s trying to do it to each other. You know, if I’m selling the car, and all the compliments are maybe the audio visual system from Panasonic, maybe it’s the apps you can download into the car, that would be a type of compliment. I don’t know what else could be, maybe it’ll be leather seat covers, all these things that would add value to the purchase. If I’m selling the car, my goal is to make everything that’s a compliment to me a commodity and dirt cheap, because if those things are dirt cheap, you’ll buy more. But I don’t wanna myself be dirt cheap. But that’s literally exact same game plan that everyone has. So everyone is trying to commoditize everybody else when you’re all compliments to each other. It’s a funny dynamic, but this whole idea of economies of scale is more and more becoming the composite of multiple goods that compliment each other because that’s ultimately what the customer sees is that effective price. Anyways, that’s another sort of fun bucket to think about. Okay. I think that’s enough on that subject. So economies of scale, very, very important. Digital is starting to really change this in a lot of businesses from what we’ve been comfortable with for a hundred years. And I’ve just sort of laid out five ways or at least five scenarios that I’m keeping my eye on. Okay, next topic, network effects. Now network effects again have been around forever. This is also called demand side economies of scale. The previous economies of scale was really about the supply side, the cost structure. This is more on the demand side. And the way that is usually explained is economies of scale on the cost side means when I’m bigger than my competitor, my cost structure is lower by virtue of being bigger. Now on the demand side economies of scale, network effects, as I am bigger than my competitor in terms of volume. my product or service is actually better. So the more volume I get, although it’s not always volume, it can be users, it can be a lot of things, but generally the larger I am on the demand side, the more people that use the product, the better the product is. So my standard thing is like you go to KFC, I go to KFC, because I went to KFC, your chicken didn’t taste better. But if you’re on WeChat and then I join WeChat, your WeChat actually got better because now you have more people you can talk to. The service actually improved when another person used it. So demand side economies of scale, network effects basically have a feedback loop. Economies of scale on the supply side don’t. That’s kind of important. Now, other things to think about. I have long sort of talked about platforms. versus network effects and I always qualify it. And I say, look, not every platform has a network effect. They often do, but they don’t have to. I mean, a platform business model could be a traditional newspaper. One user group would be the readers and the other user group could be advertisers. Now that’s two user groups. They’re in a platform business model. And that is why the reader often gets a lower price or a free price because you’re charging the advertiser. You’re using the two sides of the platform to subsidize price. But there’s no real network. So it has some of the advantages of a platform like chicken and an egg effect, price subsidy. Those are very important. But it doesn’t really have a network effect. If I’m reading the newspaper and you read the newspaper, there’s no one-sided direct network effect. And if you put a lot more advertisements on the newspaper, That doesn’t make it better for me as the reader. Now you could say, I guess if more readers are there, that adds value to the advertisers. So there’s a little something there. But generally speaking, most of what we’re talking about is a platform business model with some soft advantages as opposed to a powerful network effect like Taobao or a marketplace. So I’ve made that point many times. The point I also haven’t really touched on is sometimes you can have network effects without a platform. This actually exists. And standardization network effects, or also called interconnection network effects, this is a scenario where just by connecting things, it becomes, you get a feedback loop. You get demand side economies of scale, but there’s only one user group. There doesn’t have to be multiple user groups. So let’s say. railroad tracks. If we have one railroad company in Florida laying tracks and we have another railroad company in Georgia laying tracks, but the tracks are different sizes. Maybe they’re five feet wide in Florida and four feet wide in Georgia. Now if you were to standardize this railroad track size, let’s make them all five. Suddenly those two railroads could connect to each other. And suddenly people who make trains don’t have to make two sizes. And suddenly people who maintain trains have one, let’s say the railroad tracks, have one skill set. There’s a lot of, one, there’s externalities that are important. But generally speaking, there’s just a lot of savings advantages that happen. And for all the users, it’s much better because you can now travel multiple places. there’s a lot of benefits to, hey, let’s all have five foot railroad tracks and everything connects. Now there actually are some externalities there, but generally speaking, I think you get the point. So in the sort of 70s and 80s, when a lot of this tech stuff was really taking off and people were talking about, how do we get one PC and connect it to many PCs? And there’s a lot of discussion about protocols. and how can one computer talk to another computer and you do that enough times and you get the World Wide Web. And so there’s sort of TCP and the other protocols and that’s when this idea of sort of standardization network effects really play out. Or actually it happens in a lot of places, but a lot of talk about that. And this is when you got Metcalfe’s law, which is not something that’s really a law, but everyone says it’s a law, it’s just an idea. And the argument was the value of a network increases at the square of the number of nodes connected. Basically, the more places connected with the same railroad tracks, the more valuable the whole rail system is to everybody. But in this case, he was talking about connecting computers and other things and ether nets together. So it was a lot about connecting the pieces of what later became the World Wide Web. So connection protocols. That would be a type of standard. that would get you a standardization network effect. Look, the more people that connect their computers with this type of protocol, the better it is because they all connect. Fine. Another type, which you always heard Bill Gates used to talk about this all the time, he would always talk about file types and file formats that we want everything to be used in Microsoft documents and we want everything to be PowerPoint and we want everything to be Excel and PDFs. And the more people that use Adobe and have a.pdf at the end of their files, then there’s more people you can send your document to who can read it in their reader. Or if I’m using Word doc and you’re using Word doc, I can send it to you and you can edit it. So the more people that just use these formats, the more valuable it is for everybody. But it’s not a platform, it’s just when we’re all on the same standard for a file format or a connection protocol or something like that. you get a network effect. And the network effect is pretty much proportional to the number of people using it. So it really is sort of Metcalfe’s law that the value increases with the number of nodes, the number of people, the number of connections, not necessarily their usage, which happens in other places. So there’s a lot of fight to become the standard. China is actually talking about this a lot right now, that they want to, concerns, let’s say, of when China looks at their digital strategy and their technology strategy going forward. As a country, one of the things they talk about a lot is a lot of the standards that are used internationally were created by the West over the past 20 or 30 years in semiconductors and other things. And they are trying to be more involved in creating the next generation of standards. And they have a lot of institutions working on this. So that’s actually something you hear about a lot in China is the fight for standards. What are the benefits? I mean, why is it better if everyone uses it? I mean, you can make a long list. Training costs. We don’t all have to learn 10 different types of ways to create documents, because we’re all using Word. You only have to learn one. A lot of cost advantages, it’s more efficient, it requires less investment. Sending things back and forth, there’s lower transaction costs. Risks, I guess you could say there’s a risk reduction effect. If you know everything is connected on the same standard, if you’re gonna make an investment in a new product, you know it’s gonna work because you can see the protocols and you can build against them. You can make bigger decisions. I mean, there’s just a long list of sort of advantages if we’re all on the same standard, especially when it’s about an interconnection. Okay, I think I’ve made the point there redundantly at this point. I was listening to, well, I was reading, but I was listening to an audio version of one of Bill Gates’ annual letters from the late 1990s. And the whole thing he was just talking about standardization and file formats within the ecosystem. It was like half the talk he kept talking about this. It was really interesting. So anyways, you can get network effects without a platform. It’s rare, but it does exist. There’s a lot of fighting about how you get standards to emerge. and lawyers are all over this. I think this is a pretty big corporate law practice because if you’re a big network, let’s say you’re a big railroad, maybe you don’t want other people to connect with your railroad, which is really what the railroads didn’t want this. I mean, they were kind of like, if I’m dominant in the Northeast and I’m the big player and you’re the dominant railroad in the Southeast, it’s better if we’re not on the same. Protocols same standards because then I don’t have to compete with you. Maybe I don’t want it to connect that would wreck my world a little bit Companies like Ford in the early days and Tesla today, they don’t necessarily want everything they do to be easily interchangeable in terms of parts, connections with other car companies. Because they’re kind of in the lead. Instant messaging, you know, it used to be, you know, you couldn’t send from various messenger systems like AOL to others, they didn’t connect. There’s always this argument of like, I don’t wanna give up my little fiefdom and… The counter argument is always, yeah, but if we all do this together, the whole pie will get bigger. You know, it will become a bigger business, a bigger market, more robust. So yes, your slice of the pie, you’ll go from being the big slice of a small, the big slice, the big portion of a small market to a smaller portion of a big market. And you kind of got to get everyone to agree to that, which is problematic in a lot of cases. Anyway, so there’s a lot of stuff of standards negotiation and who’s the standards leaders, and there’s a standards war between DVD and Blu-ray and whatever, and anyways. It’s this whole world of thinking of how standards emerge. It’s pretty funny, it’s not really my area. So that’s sort of the first point in terms of how does digital impact a competitive advantage like network effects? Well, network effects have been around forever. I mean, an old downtown marketplace in ancient Rome. was a platform business model with a network effect. The more shoppers that showed up, the more benefit to the consumers and vice versa. Okay, as we start to move digital in network effects, I mean, just the list of things that are happening is really important, because I mean, what is digital doing? I mean, what digital is really doing is it’s connecting things that were never connect. First it digitizes things and then it connects them. Like railroads. connected locations. So we’re just connecting more and more of the world and of business and society. So network effects and standardization, obviously, are becoming a bigger and bigger subject. And I think that’s going to continue. So that’s sort of point one of the intersection of digital and network effects. Point two, which will be the last one for today, is this idea of pricing and network effects and customer tranches and how this kind of evolves over time. It really is kind of complicated when you think about if you’re going to get a network effect as it grows over time. It sounds really easy. Well, if more people use it, it gets more valuable. Well, I mean, it’s a little more complicated than that. Like, often the first people that use something with a network effect, like let’s say going on Taobao or, I don’t know, getting a ride on Didi or Uber, although they don’t have their network effects aren’t huge. Usually it’s that first wave of people that get the most value, that had the most compelling reason to join. But I’ve said, but as more people join, the value actually goes up. That’s what the network effect is. That’s true. So you do have, okay, these people join and then the value of the service goes up. But at the same time, often what’s happening then is you’re also moving from your first tranche of customers to a second tranche who maybe didn’t have as compelling a reason to join. maybe they didn’t need the ride home as much. You know, you’re moving from your best customers to your second best customers, which is a lot of what happens when you grow a business. So yeah, the value of the network is increasing, but at the same time, you’re also moving to people for whom this service is less valuable than your first group. So there’s sort of opposing forces that play out as that evolves. That’s worth thinking about. Then you have to think about, okay, How do I price for that then? Because in theory, I should be able to raise my prices as the network effect gets going because it’s more and more valuable of a service. So the prices show go up, they should be the willingness to pay should go up. It’s a heck of a lot more to go on a marketplace like Taobao when there’s five million sellers as opposed to 10,000 sellers. That should be more valuable to me. Yeah, I mean, that’s true. But then there’s also the idea, well yeah, but maybe this group doesn’t care as much, that’s a problem. And then the other thing is, but maybe my goal is, maybe I should lower my price for the first group because my goal is not necessarily to get money from them, but I should be thinking in terms of penetration pricing, that I should lower my price to get as many people in that first wave as possible because that will create more value. quickly for everybody else. So is this, I mean, the pricing question becomes very complicated. So there’s actually kind of a lot going on within that when you start to take apart the pricing. And this is something like venture capitalists are very good at this. This is a little bit outside of my world, but they’re very good at those early stages of building a network effect and get it going. I tend to focus more on the later stages of, I like when a network effect starts to decline and suddenly users are going away and the management starts to panic because they’re worried the network effect is gonna flip into reverse and go down very quickly. You know, my argument is like a network effect is like flying an airplane, that it’s very hard to get that thing off the ground, which is what venture capitalists focus on. But then if you get it into the air, it actually goes much faster than any, just about any other vehicle you can think about, cars, bicycles, whatever. So it goes really, really fast. But if you start to lose users, the value of the platform starts to decay very quickly, and that’s like losing altitude. And you realize once you’ve got your network effect going, I called this like the sustained innovation trap of network effects. that once you start to lose altitude, you start to lose activity or users in a business built on network effects, it’s literally like your plane is going down and management gets panicked and there’s not a lot they can do. There’s no way to land the thing. If you start losing a lot of consumers, well, then your merchants go away. If you lose your merchants, well, then the consumers find it less valuable and it goes in reverse. And so you see a lot of frantic behavior by. companies in this situation where they start spending a huge amount of money on marketing just to keep people coming to the platform to stop this. Didi actually had this scenario when there was a couple violent incidents in their cars in China and that really hit at the core of the business where basically some passengers got killed. And that really hit at the core of the business which is trust. And you can see what could happen to their network effect could flip into reverse very quickly. If people don’t trust that it’s safe, they will stop going and then drivers will stop being interested. And then with fewer drivers, fewer consumers, I mean, trust is the centerpiece of ride sharing. It’s the same for Airbnb as well. If you don’t trust that you’re staying in someone’s house, that thing goes down very fast. That plane really goes down. And so what DD did, because they’re very good. in terms of a management team, is they focused their number one priority for their company was safety. And I went to their campus in Beijing and I mean there were banners everywhere about safety and they had like daily sessions. I had to leave because they had a daily session that was about safety and the whole the whole company came down to the auditorium to talk about safety. And that was you know they focused on that. And then I went to Brazil and I met with Didi there which was 99 in Sao Paulo. And I asked them the same thing, what’s your priorities? Is it growth, is it market share? And they said, our number one priority as a company is safety. I mean, they were laser focused on restoring the trust of everybody and in making it safe, not just talking, but making it totally safe. Anyways, so that’s something to think about, this sort of price discrimination, various tranches of customers, and how these things decay and go down, which is really my… interest is, it’s not how you build the competitive advantage, which is the venture capital question. It’s more how you defend it and when it starts to decay or can you attack it. I kind of like that second phase a little better because I do later stage stuff. Okay, I think that is enough for today. So I’ve covered three ideas today. Economies of scale, how digital can really change something we’ve been very comfortable with for a long, long time. Network and then also standardization network effects, which are pretty important. I think that’s enough for today in terms of ideas. And this all goes under learning goal number 29, which is digital economics two. So the last couple classes, I’ve been basically doing theory. I was basically filling gaps that there were sort of things that I had. not really talked about within the scope of the content. So that’s what’s been going on. I do want to get back to companies and cases, which we’ll do next week. So this was the last sort of filling in the gaps in terms of content. But I mean, I think it’s interesting, but I think cases and company discussion tends to be more useful. And the theory does get to be a bit much after a while. Anyways, I’ve had a pretty good week. I’ve just been running around doing administrative stuff. Visa processing and all that and getting a driver’s license in Thailand which has turned out to be a real adventure. I’m going back in the next day which will be my one, two, three, fourth, fifth trip down, fourth or fifth trip down to get this done because there’s registration, you got to make an appointment, you got to go to the embassy to get a letter, then you got to go back, then you got to sit there and listen to… hours of a video on safety laws in Thailand, which was of course in Thai, although they did have subtitles. That went on for hours and then you have to take tests which are surprisingly, let’s say not difficult, but obscure. I didn’t pass those, I had to go back and take it. I did learn something by the questions I missed. I didn’t, you know, I didn’t quite know the helmet laws for novice monks. which they don’t have to wear a helmet. Didn’t know that one, got that wrong. I didn’t know the proper procedure for cleaning acid stains off a battery of a car. The answer to that one is you don’t use vinegar. I got that wrong. You don’t use lemon juice. You use groundwater and then you apply grease. Got that wrong. Yeah, there was a whole bunch of questions. They’re like, I was just, I have no idea. I’ve never thought about this my whole life. If your car goes underwater, what should you do first? I actually got that one right, which is, you know, go for the door. So this was what I learned in the Thai exam. My favorite part was, because there’s only a handful of foreigners, they’re not set up to do foreigners, right? It’s for Thai folks. And So they hand you these photocopied pages. They hand everyone else these formal books. Here are the laws, please study these, sit here for three hours and then take the test. But they don’t have that for foreigners, so they hand you some photocopied pages, about 30 or 40 of them, that are not high quality photocopies. They’ve been photocopied like 100 times. These pages with questions and laws, but the grammar is not awesome and the English was not awesome. So people would correct them. expats probably so they’d hand write on top of it like it would be like this you know you cannot drive a car more than 2.5 meters in bubble and it would be crossed out and say not more than but less than Because they had mistranslated the english which is kind of an important distinction So this was all throughout there and then about every other page In the handwritten notes, which they were photocopying and handing out these booklets with the handwritten notes in the photocopying There would be obscene comments like like really obscene, like some of them are just jokes and then some of them are really obscene. So you’d be reading a question about what is the proper way to signal slowing down in a vehicle and you put your arm out and you wave it up and down and then it would be like, I won’t even repeat it but it will be expletive, expletive, you drive too slow. Why expletive, expletive, you drive so slow, expletive. Like there’d be these cursing words all throughout a lot of the pages. And some of them were really out there. And I think they have no idea. I think they’ve been handing out these photocopies for years and I’m like, someone actually pointed them out. The guy next to me is like, you know, there’s a bunch of cursing in these books cause they had no idea. It was pretty funny. That was my highlight of the day was the expletives in the driver’s tutorial at the Bangkok Lane Transport Office. Anyways, that was an interesting part of my week. Okay, that’s enough for today. Thank you so much for listening. I hope this is helpful. We’re still doing a lot of phone calling out to subscribers, you know, getting feedback. That’s super helpful. Thank you to those of you who I’ve spoken with. If you’d like to provide feedback, please just send me a note and definitely want to talk to you. But that’s it. Have a great week and I will talk to you next week.

Leave a Reply