Review: The Sexy But Dangerous Economics of Digital (Tech Strategy – Podcast 15)

In this class, we review the basics of digital economics, which is one of the key learning goals for Level 2.

You can listen here or at iTunes and Himalaya.

This is part of Learning Goals: Level 2, with a focus on:
  • #1: The basics of digital economics

Concepts for this class:

  • Digital and Information Economics

Companies for this class:

  • None


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, why are so many things free on the internet? Now, this is actually gonna be a review. I’ve laid out sort of 10 main learning goals for what we call level two intermediate. And one of those is about the economics of digital, of bits and bytes and ones and zeros and how they are quite strange. So this is gonna be sort of a short review of some of the important ideas there. And we’ll do it under the framework of, you know, why are things free? It’s weird. Anyways, that will be the class. But first, if you haven’t subscribed, I would really appreciate it if you can go over to and sign up for the class there. It really does support this effort and we’re sort of staffing up and getting bigger and moving forward. So that actually is a big deal and I would appreciate it. Okay, let’s get into the subject. So in a previous podcast on VIP Shop, I believe, we talked about the economics of digital. And for those of you who are subscribers, there is a learning goals structure to all of this, which you have received. And we break it into level one, level two, and level three, but right now we’re talking about level two, and sort of laid out 10 learning goals within that. And that’s kind of the process. And most of the… you know the talks and the articles are gonna start referring to those main ideas more and more so I thought it would be good to review the ten most important ones and today we’ll start with learning goal number one which is the economics of digital which is this idea of look why are things free on the internet I mean it’s weird it’s it’s so unusual and we take it for granted now but you know twenty years ago people didn’t give you free stuff in life. I mean, you walk through your day, you go into stores. Stuff wasn’t free. I mean, maybe you could watch the TV and that was kind of free and newspapers might be, but very few things. Now, like, if you think about all the stuff you use in a 24-hour period when you’re online, I mean, how much of that is free? It’s almost all of it. I mean, I’d say most of it. So that’s kind of really bizarre. Anyway, so the… Learning goal for this was the economics of digital, which I refer to as sort of the the sexy but dangerous economics of digital, which is one, the economics of digital are really powerful. They’re unique, they follow from software and things being made of bits and bytes and not physical atoms. So they are pretty amazing. However, that’s the sexy part. The problem is this They’re also open you up to a world of competition. It’s an incredibly difficult business to compete in, so they’re also really dangerous. It turns out most software is free, and most software is a commodity. And so if you can get the good economics, you know, you make all this money, but most people are just struggling to survive. So it’s a strange world in terms of economics, and it used to be sort of a creature of the internet. But more and more, those economics are invading other aspects of life, like services, which are now becoming about things like data and AI, physical goods like scooters and cars that are suddenly becoming and smart, which basically means they’re becoming infused with software. So a lot of the other businesses we’re comfortable with, retailers, automotive, things like that, are being altered by the economics of digital. So it’s kind of out there all over the place. In the previous talks, I basically said, look, there’s a lot going on here, and we’ll do this in multiple levels, but the simplest version of this is what we call zero marginal costs of reproduction, non-rival goods, cheap distribution, and usually non-consumed goods. Now there’s other aspects of digital which are important, like versioning and complements and bundling. I’m gonna do that in a separate talk, but for this sort of shorter review today, Just think about those zero marginal cost of reproduction, non-rival goods, non-consumed goods, cheap distribution. Okay, now the non, you know, the zero marginal cost of production is pretty simple. I mean, I spend six months writing a book, a lot of research, a lot of thinking, a lot of editing, and then, you know, we create it, refine it, and the end result is basically a PDF. I mean, that’s the end result of a book. and uh… you know that first pdf actually took quite a lot of time and effort and often a significant amount of money depending how you do it uh… okay so the first version of the pdf costs uh… quite a bit of money but the second version the marginal reproduction not the first the marginal one the next reproduction basically cost zero i hit control c and i hit control v and now i’ve got two copies you can copy and paste, copy and paste, copy and paste. So that second or that marginal cost tends to be free or very low cost. And that is kind of interesting. I mean, that’s why digital books are, you know, when I do books, I do a PDF version, which is an ebook, and then I also do print on demand, which is like an Amazon service, which is a physical book. Now, physical books don’t have, a zero marginal reproduction cost. You have to actually print it out and bind it. And you know, that costs a couple dollars usually. Like I can sell an ebook for like, I could sell an ebook online for basically $1. For a physical book, print on demand is, you know, it’s like I sell for like $8 or something, pretty small. But yeah, okay, so that’s kind of. One way to think about this, but we’re not talking about the total production, the cost, the total production cost, you have to think about the cost of the first version as well. Now it turns out one of the reasons books are such a terrible business, I just told you these are cool economics because you can produce very cheaply. Okay, here’s the dangerous side of that. Books are a terrible business. I mean, they’re awful. If you want to make money in life, do anything but publish a book. Like anything, sell sodas on the sidewalk. anything books are the worst why because anyone can do it anyone with a laptop can sit down because it turns out the initial production cost of the first version is mostly time and effort and people can kind of do it and are willing to do it without getting paid so you know the total production cost cash wise can basically be zero and then people tend to write off the sunk cost of time and effort They sort of view that as a sunk, can’t get it back anyways. So that’s why books are so terrible because you basically can’t charge for them. And there’s millions of them coming out every year. Everyone with a laptop can do it. They’re all up online. And you’re competing with all of these books and not only every book that’s written this year, but every book that’s ever been written. So it’s terrible. Physical books are actually better if you do them print on demand because the physical… Aspect printing it shipping it creates some degree of costs That’s a little bit better, but that’s all definitely better than the traditional way of shipping them in in trucks Putting them in store inventory renting retail space hiring staff You know physical retail of books is just awful and I stay away from that but yeah, when you think about these sort of production costs think about the difference between the fixed and upfront costs and the ongoing variable costs. And usually in digital, the ongoing variable cost is close to zero and the fixed and upfront costs, you know, if you’re making the Avengers movie, you’re talking about a couple hundred million dollars, which is why very few people can make a movie that big. If you’re talking about a book, pretty low. If you’re talking about a podcast, which is one of these in podcasts are not a good business. It’s because anyone can do with a microphone. Making videos, which I do from time to time, is actually a little bit better because it takes a little more technical skill to throw a drone up and all of that. But generally, digital goods are created by labor and they tend to have these type of economics. And the really cool, powerful businesses, like Facebook, I don’t wanna say Twitter, I don’t really like Twitter, TikTok, they tend to capture just the second half of that equation. where Facebook doesn’t create content, that first version, that upfront and fixed cost, they don’t do that. They let journalists and everyone do that. They just capture the reproduction costs because that’s where you share it. So they’ve got a really clever business model where everyone comes to their platform, spends the time and effort to create the first version, which doesn’t cost them any money. And then the sharing aspect is what they capture, which by the way, that part’s free. So it’s a pretty good business model if you can avoid the cost of content creation, which Instagram, TikTok, a lot of these groups, companies, platforms, try to do this where they talk about how their model is based on user generated content. When everyone says our business model is about user generated content, that usually means they’re trying to avoid that fixed and upfront cost of content creation. Or other types of software could be application development, it could be a lot of… Could be a lot of types of intellectual property. Okay, so that’s number one. Number two, these are non-rival goods. That’s an economics term. I don’t know why they call it that. It’s kind of hard to remember. Needs a better name. Non-rival, which means, you know, multiple people can use the good at the same time. You go down to the corner store. If I buy an apple, you can’t buy that apple. If I’m reading a physical book, you can’t read the same physical book. That limits your usage, it limits your revenue, but if you have a non-rival good, I can read the PDF at the same time you’re reading the PDF, at the same time a thousand people are reading the PDF. So that tends to just sort of expand your usage when your good can be consumed by multiple users at the same time. And an airplane seat would be a rival good. Only one person can sit in the seat. A pair of headphones would be a rival good. Only one person can wear them. But the music you’re playing on the headphones, non-rival. A lot of people can be streaming at the same time. Driving in a car, rival. Only one person’s driving the car. The freeway itself is non-rival. That lots of people can be using the freeway at the same time. And… The reason I bring that one up is it’s generally worth pointing out that non-rivalry is only within a range, typically. If enough cars go down the freeway, it becomes rival because more cars can’t enter the freeway. So non-rivalry is usually within a spectrum. Anyways, interesting. Okay, point number three, consumed versus durable goods. Basically the idea that some goods you consume, some goods you can use forever, you buy a table, you can use it forever, you buy a PDF of an e-book, you can read it forever over and over. Physical books are pretty durable as well, although they do wear out. But you buy an apple or a banana, a Coca-Cola, Snickers bar, you know, you use it once and then it’s gone. If you ever go into apples, like I spent some time a couple years ago going through the financials of like apple and like literally physical apples, not the electronics company and wine companies and the economics are just the worst like They’re really just terrible like it’s so hard to grow and sell an apple You got to buy the land. That’s a lot of real estate. You got to grow stuff that takes time It’s very dependent on weather and bugs and trees. You got to pick the apples You gotta take them, you gotta do distribution. The apples only last a very short period of time. They’re perishable goods. You take them down to the supermarket. The retailers are pretty powerful. They squeeze you, and finally your apple’s in a bin at the back of the supermarket, and customers don’t care about brand at all, you know? And it only lasts a couple days. And only one person can eat it. It’s, you know, it’s a rival good, and it’s got significant marginal production costs. and it’s consumed by one, I mean it’s just, the economics are just the worst. Selling potato chips is dramatically better. They come out of a factory, you don’t need any land, you don’t have insecticide problems, they last kind of a long time, they’re kind of addictive, people care about the brand name, they want Lay’s or whatever. Anyways, but the idea of consumed versus durable, that’s not really specific to anything digital. People talk about that all the time. The reason I bring that one up is because consumed versus digital becomes really interesting when you start thinking about software and data. Because software in theory is a durable good, it lasts, but as the technology changes and the complexity changes, sometimes it’s better to have new software versus old software because you have to make the old software still work. There are legacy problems. that get created when software gets older and older and you’re on version 0.8. And your new version of Microsoft Office version nine has to work with all the previous seven versions. So that’s almost a limitation compared to a new company with no legacy issues. And data is also interesting to think about consumption versus durable because some data is very, very durable. A dictionary. is very durable. You build a business, we’re going to translate English into Spanish. One, that’s a commodity. That’s not awesome. And two, once it’s done, it’s pretty much set in stone, which makes, you know, it’s very hard to protect it as a business. But a lot of businesses like Waymo that use data, you know, nobody wants the traffic data from yesterday. In fact, nobody wants the traffic data from this morning. So data, it’s not, I wouldn’t call it consumed, but it does sort of degrade. are almost amortized very, very quickly and become unusable. So software companies and data companies where the data is either consumed or sort of non-durable tend to be very interesting. Mapping is very powerful in that respect. And there’s a lot going on with how to think about data these days. All right, number four, we’re only gonna do five here, four to five. So we’re in the home stretch. I told you today would be short. Digital goods tend to have zero or very small distribution costs. I think this is one that people get wrong because it’s like, oh, okay, I wrote a book. It’s a PDF. I can email it to you and the distribution costs are zero. I can put it up on my web page. Someone could go put in a credit card and download it immediately, no distribution costs. That is in fact true. software and a lot of these companies don’t have distribution intellectual property a Lot of things I’m talking about You know, I’ve been talking about software and digital but a lot of the same ideas are true for intellectual property rights patents ownership of content, you know who owns the Avengers or the spider-man rights and things like that. A lot of this is the same because it’s kind of all intangible assets in practice distribution costs are actually significant. If you’re gonna sell a book, you have to go through Amazon. Pretty much. They dominate the market in the West, and they’re gonna charge you 30%. Whatever you price your book at, they’re gonna charge you 30%. That’s their distribution cost. If you’re gonna go through Facebook to distribute various things, Facebook is basically gonna act like a gatekeeper, and you’re gonna pay them, it’s gonna be called marketing money, but it’s really a distribution or a gatekeeper type. A lot of people think e-commerce sites like the Alibaba’s and Amazon’s that’s basically what they’re doing. You have to spend on marketing money to keep your distribution channels open. So ironically here, you know, distribution of digital goods was supposed to be free. Turns out it’s not free because these platforms have such a power in the marketplace. And now people are looking back at retail stores as… a place to distribute that actually lets you go direct to consumers. And the argument is maybe retail stores were never distribution. Maybe retail stores are really about a marketing channel where that’s how you can tell your story to direct directly to consumers with no interference by a platform. So maybe distribution is free, but retail is becoming a marketing expense. So these two, these terms are getting kind of mixed up, but they’re fun to think about. Okay. But generally speaking, you could say digital goods by and large have low distribution costs, especially if you start to think about distributing globally. If you want to sell a book, it used to be if you sold a physical book, it was a total pain. Like I did one publishing deal with a traditional publisher, it was not good, I’ll never do it again. And to distribute your book all over the world, the publisher has to do secondary deals with every country around the world that you want to go into. This for me was English speaking markets. Then you’ve got a distribution in all these countries, then you get on retail. It was really hard to get your book into bookstores on all the streets around the world. Digital goods can be globally distributed very, very easily. So in that sense, yeah, it is remarkably cheap. Netflix can go into country after country just by opening up a webpage and changing the language. That’s pretty amazing. Okay. And one of the interesting things about these low distribution costs is it allows you to start to ignore the mass market where, you know, we’re not doing a mass market good like ketchup Heinz that goes through, you know, mass market production. We do it all in one big factory because we get better unit economies, then to mass market distribution, then to the mass retailers like the Walmarts and whatever, you know, and then we try and reach basically everybody. you know, we do a general good for as many people as possible. That was really how the world was focused. Everything was mass market, not niche. And then marketing was the same way you advertise on the Super Bowl, which is, you know, designed to reach a lot of people. As the distribution costs and the cost of reaching people through sort of programmatic marketing has dropped, little niche players can do very, very well by ignoring the mass market and going for just… very small groups of people that are scattered all over the place, but you can now reach them and sell to them. So suddenly you don’t want the mass market, you want 10,000 people somewhere in the world who really care greatly about this type of sun tan lotion or allergy free lotion or something like that. I mean, it enables you to switch from mass marketing to sort of niche focus. And a big part of that is dropping the distribution costs, not just in your geography, but globally, where you can start to aggregate a sort of niche market on your own. Okay, those are the four big ideas. The other one I brought up, which I’m not gonna go into, is basically platform business models, which to a large degree are about connecting digital goods and services between people. That’s what they do best. Now they can also do this in the real world. You can use Uber to get a ride by connecting with a driver and then the good is actually the services delivered in the real world, which is getting in the car. But a lot of times these Palette forms, particularly the powerful ones are about connecting various users online in the exchange of physical goods and I’m sorry, digital goods and services. So platforms are a big part of this. Okay, and you know. Final comment, you know, digital goods and services, generally speaking, up until now, are created by people, and they’ve mostly been intangible assets. For example, intellectual property, patents, copyrights. R&D is really a type of intellectual property that could be put in this. User-generated content, which we just talked about, organizational capital, like business processes, techniques for production. business models, the software that runs your business, human capital. Now within this, what’s kind of getting interesting is this idea of what if digital goods and services are not created by people? What if they are created by software? What if AI and various types of machine learning are now creating the content without people being involved in a significant degree? So then we have intangible assets. creating other intangible assets, which is pretty cool. And it could be physical goods as well. I mean, they could just go to the factory and get printed there, but. So that’s kind of getting interesting of where also these economics gonna start playing out in more advanced types of software like machine learning. And yes, we’ll get into that into later classes, but that’s basically what I wanted to cover for today. So I think I kept it pretty short, 20 minutes. Not terrible for me. My typical lecture at a school is three hours, so this is actually quite good for me. But yeah, I think that’s a good review for learning goal number one, which is the sexy but dangerous economics of digital. And the question of, okay, why are things free on the internet? Yeah, I mean, basically, most of those reasons are part of it. And then there’s some other aspects here like bundling, versioning, complements, which I’ll go into another one. But I think that’s enough for today. Okay. I think that’ll be it. Thank you so much for listening. One last reminder, if you haven’t signed up, please go over to and sign up there. I would greatly appreciate it. And if you have any feedback on this, I’ve been talking to quite a few people about what’s working and what’s not, and it’s been super helpful. And if you have any issues or comments like that, feel free to reach out to me, or just on my webpage, and I would appreciate that. But that’s it. Have a great week and I will talk to you next week.

Leave a Reply