CHAPTER 5

The New Economics of Disaggregation and Scale

Traditional economics is based on accumulation, aggregation, and concentration to create scale and market power. The Ford Model-T business model is about demand aggregation by offering a single, mass produced car model at a very low cost. Today, the manufacturing industry is moving in the opposite direction—away from uniform standardization for low cost mass production and toward mass customization at a scale achieved through deep analytics for product personalization. Adidas has pioneered a new analytical craftsmanship that allows for the mass production of ergonomically personalized sneakers that we will discuss in Part II of this book.

The Hilton hotel chain is an example of growth through the accumulation of assets. This model was turned upside down by the completely asset-less and massively successful short-term rental model of Airbnb. In the 1950s, the shopping mall model became very successful because it concentrated a large number of diverse shops into a single location. This made the shopping trip easier for consumers and simultaneously increased the share-of-wallet due to the larger variety of goods. Amazon successfully took the physical mall out of the business model to save consumers the entire trip and to gain even a larger share of the consumers’ wallet.

All of these business models have produced large-scale operations that have captured significant market share in their industries. However, today the successful models from the physical economy are in tight competition with the new business models of the digital economy, and the economics of growth is radically different.

Disaggregation, Concentration, and Accumulation

Complete aggregation, that is, the ability to aggregate and capture the entire demand for particular goods and achieve a monopoly, is impossible in the physical world. Yet, it is quite possible in the digital world through the process of disaggregation. It is the fundamental reason for the fast growth of the digital giants.

No long ago I was discussing with Gerald Cohen, founder of Information Builders Inc., whether we can judge correctly the likelihood of success or failure of start-ups. We agreed that we place a great value on the business model, but even more so on the impression that the founder makes on us. And then Gerry said: “But even that can be quite misleading and quite difficult to judge.”

He then told me how he met Jeff Bezos when he was just starting and pitching his early stage Amazon venture. Gartner Inc., the global research firm that provides insights and studies on emerging technology trends and technology management practices, organized regular lunches between CEOs of successful, established companies and newly founded start-ups. At one such event Gerry Cohen and Bill Gates had to share a table with two aspiring entrepreneurs. Both founders were selling goods online. The difference was that one was selling wine and the other books.

When the two founders left the table, Gerry and Bill started to discuss which one of the two had more chances to succeed. They both thought that the wine business had a bigger chance to succeed and were almost certain that the book business would fail.1 They thought that the business models were the same and, hence, put more faith in the charming personality and storytelling of the wine seller. As we know today Gerry and Bill were wrong.

It is interesting why, at the early stages of online retailing, selling books turned out to be more successful than selling any other type of product. Bezos gives the following two reasons why he chose books.2 According to him books have a distinct retail advantage because they are sortable, packageable, and shippable. Second, there is an overwhelming number of already existing books that cannot be displayed and sold even in the largest brick-and-mortar bookstore. In other words, the demand for books cannot be completely aggregated as no retailer can sell all the books in the world. Furthermore, no retailer could even own copies of all the books existing in the world. But Amazon did not have these limitations. It did not need to own copies of the books either as it simply allowed people and organizations to sell second-hand books on its platform. In this way Amazon was able to expand its catalogue of books beyond anything that its brick-and-mortar competitors could do.

The supply and demand argument that Bezos emphasized could be made for any type of product. There are more bottles of wine in the world than any shop or liquor retailer can own, store, and sell. The markets for clothes, shoes, electronics, and so on all share the same limitations. Supply cannot meet all the demand in the world because of physical constraints.

On the other hand, Bezos’ first reason about why he chose books cannot be made for many types of products. Wine is not easily sortable, packageable, and shippable. It is also hard to store. For many years the fear of shopping online was caused mainly by concerns about the delivery of the purchased goods. What would happen if the bottle of wine was broken during shipment, what if the shoes were not the right size, and what if the material of the dress was not as it appeared on the picture and in the description? Many successful online retailers had to build in their business models the means to ensure the trust of the consumer that the delivered goods would arrive and be of the quality that they expected at the point of purchase. Zappos’ success of selling shoes online is largely attributed to the invention of the free returns which alleviated shoe shoppers of the fear of regret.

It is either the stroke of a genius or pure luck that such fears did not exist for shipping books. The postal service has been delivering letters, papers, magazines, and books packaged in envelopes for many years. People are accustomed to receiving printed materials in the mail and know that there is minimal risk for damage during shipment. Because of this, consumers had no psychological barriers to shop for books online. On the contrary, they just could benefit from the large variety and lower prices. The pre-existing consumer trust in the new business model provided Amazon with easy access to early adopters and a staying power through the early stages of online retailing that ultimately allowed Amazon to expand into other product categories. Trusting Amazon for the delivery of books gradually led to trusting Amazon for the delivery of anything.

Amazon is an example of the ability to disaggregate demand to an extreme. As the merchandise catalog expanded, Amazon could satisfy more and more niche preferences. This has become known as long tail retailing. “Long tail” is a statistical concept about the distribution of events. Popular items have a higher frequency of occurrence than less popular items. In retail, more frequently purchased items tend to be easier to find. They are placed in more visible locations in the stores; there is more advertising of them, and more reviews. This was the fundamental idea behind the Ford Model-T. The lower the cost, the higher the frequency of purchase. The higher the frequency of purchasing the more awareness in the market as everyone sees and hears about the Model T. This leads to an always increasing frequency of purchase.

But there are a lot more products and untapped revenue opportunities in the long tail if a retailer can make special items with low frequencies of purchase as easily accessible as the high-frequency products. Some economists have estimated that the sales potential in the long tail is at least equal if not greater than that of high-frequency items. The economic theory of the long tail states that:

Our culture and economy is increasingly shifting away from a focus on a relatively small number of “hits” (mainstream products and markets) at the head of the demand curve and toward a huge number of niches in the tail. As the costs of production and distribution fall, especially online, there is now less need to lump products and consumers into one-size-fits-all containers. In an era without the constraints of physical shelf space and other bottlenecks of distribution, narrowly-targeted goods and services can be as economically attractive as mainstream fare.3

The more an online retailer can expose items from the long tail to consumers, the more customers with special interests and tastes it will attract. The technology that made this possible is search because it eliminates the traditional placement and advertising advantages of high-frequency products. It is search that allowed Amazon to disaggregate demand completely and ultimately to become the store for everything.

At a press conference in 2014, Bezos describes what they are trying to achieve and how it affects the consumer:

We are not trying to build a great service for tens of millions of consumers. We are trying to build a great service for one consumer. If you think about it that way, the consequences of building a great service for one consumer is you get millions.4

With search even the most particular preferences and needs could be matched to a niche product and a niche supplier. And when this happens, consumers really feel that the store is just built to satisfy their unique needs. As Bezos points out in the above statement, the corollary of disaggregation at scale is that it leads to extreme concentration of suppliers and consumers. Amazon attracted a large number of vendors and suppliers who sell items from the long tail and, thus, it grew beyond what any superstore or supermall could do.

But there is one more important factor in such growth stories. Disaggregation and concentration are fueled by the accumulation of data. The data about products, consumer preferences and behaviors, supplier prudence, and much more is what keeps the engine running and growing. The more data the company accumulates, the easier it is to answer any consumer question which, in turn, makes self-service retailing easier and frictionless.

The Socioeconomic Effects
of Disaggregation at Scale

The process of disaggregation impacts not only businesses and market competition but also the operations of the job market and the market institutions. The underlying reason behind this far-reaching impact is that the technology that facilitates disaggregation is also rapidly lowering transaction costs, that is, the cost of executing a trade in the market. There is a branch of economics called transaction cost economics that studies just the effects of lowering or raising the expenses associated with the exchange of goods and services and their impact on the structure of business organizations and market institutions.

How Do Transaction Costs Affect the Job Market?

An important question in business and economic theory is why firms exist? Adam Smith argued that firms exist to facilitate the coordination of resources. The production process can be organized far more efficiently when all employees work in the same factory. The assembly line that produced the Ford Model-T makes this self-evident. Hence, the modern factory-based manufacturing system emerged and delivered big economies of scale by the use of machines and division of labor.

Ronald Coase,5 who is considered to be one of the founders of transaction costs economics and is a recipient of the 1991 Nobel Prize in Economics, argued that firms hire employees because it is cheaper and more efficient to enter an employment contract instead of negotiating on a daily basis the terms of the work to be done.6 Imagine several hundred workers coming to the factory whenever they want, at whatever time they want, and deciding on the spot whether they will take on the available tasks or not. The cost of organizing work-for-hire in such a way is too high. The company would need many hiring managers to interview and explain job tasks to applicants on the spot and to coordinate the various shifts on the factory floor.

But this is exactly what Uber and other companies in the gig economy do. Uber drivers can start driving whenever they want and drive for as long as they want. They can also reject rides without having to provide any reason either to Uber or its customers. A full-time cab driver does not have the liberty to do the same.

Uber has changed the employment relationship because its device and software application make the transaction cost of contracting and paying drivers for their services very low. And while people may argue that driving does not require specialization of labor, we can see how even high-end, highly qualified services such as design, video editing, professional editing, proof reading, and many more can be organized and offered as gigs as the very successful Fiverr (fiverr.com) has done. The employment contract can be disaggregated and transformed into a freelance gig when the market demand is successfully disaggregated at scale.

But what about government and nongovernment market institutions and intermediaries?

Another Nobel Prize laurate, Douglass North,7 argued that institutions set rules and norms that are key in determining the transaction costs in the market. Well-designed institutions lower the transaction costs and thus lead to economic growth and vice versa. Financial intermediaries exist because of their ability to lower the transaction costs. One way of doing it is by bundling together individual investor funds into a single large fund and thus save transaction costs on the individual investors through economies of scale. As the size of the bundled fund increases, the cost of doing a larger transaction with the fund does not increase proportionally. Another example of a government intermediary is the Bureau of Motor Vehicles. One of its functions is to administer vehicle titles to ensure trust when cars are bought and sold.

But these intermediaries are expensive. What if the transaction cost can be lowered and even reduced to zero? Doesn’t it imply that the intermediaries will become obsolete? If individuals can save the costs that they pay to private intermediaries to do business, they will be better off. If society does not need some of the services of government funded intermediaries, various taxes and fees can be lowered. Thirty years after the invention of the economic theory of transaction costs, a technology called blockchain promises to reduce or altogether eliminate in some cases the need for some intermediaries.

Blockchain is a universal ledger for transactional records. In the future, potentially all business transactions can be stored in this universal ledger. This is very different from the current business practice where each company has its own ledger and accounting system. In the past three decades the emergence and evolution of ERP (Enterprise Resource Planning) systems offered by Oracle, SAP, and other companies have allowed businesses to improve management and automate many processes, thus leading to significant cost savings. However, during mergers and acquisitions the integration of ERP systems is one of the costliest and most time-consuming processes. If all ERPs were built on a common universal ledger many of these due diligence and integration costs would be reduced or eliminated.8

Like the demand for Uber services critically depends on the reputation of the company, so the adoption and functioning of blockchain critically depends on the verification of the recorded transactions. Institutions exist in the traditional economy because verification takes time, money, and effort. How can blockchain make this verification costless?

The invention of costless verification, which is built into the backbone of the blockchain technology, is the key system component that makes it possible to slash transaction costs. Blockchain stores the most important transaction attributes in individual blocks that are like virtual Lego pieces. Each subsequent modification of the transaction is recorded in a new block that is chained like a new Lego piece to the entire chain. Since the entire blockchain is replicated and distributed across many networks, it is practically impossible to corrupt the individual blocks and commit fraud.

As it can be seen, the same principle of disaggregation and scale has been applied to traditional accounting practices to make fraud quite costly and verification costless. Multiple copies of each record are stored on many random nodes of the system. If the transaction is tampered with the copies will not match. Imagine the buyer of a second-hand car who can instantly verify not only the vehicle title but also the entire prior ownership history of the car. If this can be done instantly by the buyer, the sale will occur right away, and the title transfer fee will be saved.

Blockchain Drives Creation of New Data-Driven Business Models

Since blockchain facilitates direct peer-to-peer low cost transactions, many new data-driven business models will be built directly on it. When transaction costs are low, the provision of goods and services can be disaggregated and scaled too. A new breed of decentralized software applications is emerging called dApps that allow micro peer-to-peer transactions. Using dApps people can rent the free processing power of their computers to crypto miners or other companies, can make money on their social posts as it is being now enabled by the Steemit platform, get paid for visiting locations as the Geon App allows its users to do, and many more. Like the gig economy has changed employment, dApps are changing how people can deliver and earn money on microservices because of disaggregation at scale and costless transactions. With disaggregation at scale we are seeing an economic transformation where the importance and management practices are shifting from macro- to microtrends, which in turn will change how analytics is done and now decisions are made and managed as we will discuss later in this book.

---