CHAPTER 2

The Evolution of Online Retail

Starts from eBay and gang, then Amazon fulfillment followed by first responses from brick-and-mortar stores. Insightful data tells the physical and economic growth story of the fulfillment machine. Through four phases of evolution, the smart disruptors leveraged data, technology, and decision models into a customer delight. Identify the four traits of successful machine builders: prerequisites, timeliness, technology enablers, and a long view.

The Netscape web browser, launched in 1994, made it much easier for all of us to navigate the Internet. Almost immediately we were all looking for fun things to do while online. But few of us, including brick-and-mortar retailers, would have guessed that shopping would be on top of the list and better still a bunch of innovators would design, build, and operationalize solutions to satisfy that need. The history of online retail is recorded in many excellent books and articles, and you should read some of these to better understand how quickly online shoppers and retailers evolved. If you have the time for a leisurely review then read, The Everything Store: Jeff Bezos and the Age of Amazon13, if you just want the short 10-minute review then a good option is The Wired Guide to Online Shopping.14 But for a snapshot view, let’s start with some online shopping history trivia, courtesy Wikipedia:

NetMarket an online marketplace recorded the first online sale on August 11, 1994—someone from Philadelphia used a credit card to buy Ten Summoner’s Tales, a CD by Sting.

eBay, founded in the autumn of 1995, starts connecting buyers and sellers of unique, strange, and outright weird products. By 2000, it’s the largest online auction buying company with $431 million in sales.

Amazon, also founded in 1995, starts selling books online and the first sale occurs on July 16, 1995. Two years later, in 1997, the company processed its one-millionth order. By 2000, it was serving 20 million online customers with annual revenues of $2.75 billion.

The Chinese e-commerce company Alibaba was founded in 1999, launching an eBay-style business-to-business marketplace. In 2003, it launches Taobao, an Amazon-style marketplace for third-party sellers.

Starting from 2000, pure online retailers are launched in many different countries: Jingding or JD (China), Rakuten (Japan), Flipkart (India), Otto (Germany and France), GMarket (South Korea), Zalando (Germany), and MercadoLibre (Latin America).

Walmart acquires Kosmix in 2011 and renames it Walmart Labs, making online retail a top priority for the first time in its history. In 2016, it acquires Jet.com to remake its fulfillment operations.

Jumia was started in Nigeria in 2012 and soon became the largest online retailer in Africa. The company debuts its Initial Public Offering (IPO) in April 2019 with a $2 billion valuation.

Listed previously are only some of the successful innovator-led companies that have facilitated and validated the paradigm shifts listed in Chapter 1. Many lesser-known companies were also online retail builders; in many cases, these merged into larger rivals and some just ran out of capital or stopped growing. At least two legacy companies, UPS and FedEx, were also critical players in the build process and without their existing fulfillment infrastructure, most innovators would have struggled to get their ideas working. They are both supply chain inductees of the online retail hall of fame.

There were many players in the evolution of online retailing, but almost none came from within the ranks of established retailers. Even more remarkable, only a small percent of the individual innovators had any retailing experience. Note to self: Your machine-building team must include people with knowledge of the enabling technology and not everyone should be vested in the current machine. Success was not guaranteed to the early players and only a few survived the early years. Many of the players recognized some of the paradigm shifts and built their businesses to leverage the shift. But few, and possibly only one, saw all the shifts and they became online retail leaders. So why look at the evolution and not just study the machines they built. Because it is important to go through a discovery phase where you learn what worked and what didn’t, what are the enabling technologies, and what are the paradigm shifts you missed.

Data, Data, and Data

Everybody on your project team, your executive team, and your formal/informal consultant teams claims to have data that explain the past and project the future. Do not take any of this at face value. Be sure to always remember that one of the downsides of the online era is that information is easily posted and readily available to everyone. The reality, though, is that a lot of data are not validated, likely tainted by a confirmatory bias, or even downright wrong. It is OK, rather a must, to question and investigate any publicly accessed data you plan to use in decision making or project planning. Fact check, authenticate, validate, call it what you want but don’t forget to do it. So, let’s review selected validated data that provide instructive insights on the evolution of online retail and the associated supply chains.

The first of the online paradigms is online shopping, anecdotally we all know this is happening, but studying the data tells us when online shopping first became significant and what the future trend is likely to be. In 2000, online retail accounted for only 0.9 percent of all U.S. online retail sales, and experts would opine that online was a niche retail channel and posed no threat to a traditional brick-and-mortar retailer. Over the next several years, all the paradigm shifts had occurred, and by the end of 2018, fully 10 percent of U.S. retail sales (Figure 2.1) were online.15 What are the future trends? Projecting the current growth rate of about 1 percent per year, we expect that 25 percent or fully one-fourth of all U.S. retail sales will be online by 2033. What does this mean for the fulfillment machine? It must more than double and possibly triple in capacity in the next 15 years. Manufacturers must strategically plan and execute a shift of their supply chains from a retailer to a direct online focus.

image

Figure 2.1 Online retail sales as a percent of annual U.S. retail sales

Next, let’s investigate the revenue data for Amazon and investigate how it relates to the paradigms. Currently, Amazon reports sales in five categories: Online retail store; Physical stores (e.g., Whole Foods); Third-party seller services; Amazon web services; Subscription fees; and Advertising. Of these, the first two are product sales of Amazon-owned inventory, while the third represents sales of non-Amazon-owned products. It was only after 2010 that these categories were reported separately by Amazon, meaning total retail product sales or gross merchandise value (GMV) before 2010 will have to be estimated from the available data.16 The third-party seller ratio, which is the share of physical gross merchandise sales sold on Amazon by independent third-party sellers, is reported since 2000. Combining this ratio and the likely historic growth of non-product revenues,17 we can create a relatively reliable estimate of Amazon’s annual GMV product sales since 2000 (Figure 2.2). Why are we focused on this sales number? Because these sales are directly associated with the value throughput of the fulfillment machine.

image

Figure 2.2 Amazon annual GMV product sales (billions)17

The GMV product sales growth rate is phenomenal with an 80-multiple increase from 2000 to 2018. Splitting the data into five-year snapshots, the annual growth rates were: 30 percent from 2003 to 2008, 36 percent from 2008 to 2013, and 25 percent from 2013 to 2018. Was there a threshold year? I would have to say it is 2005 when online product sales crossed $10 billion. At that point, the existing U.S. parcel fulfillment infrastructure was close to capacity, and future growth would have to be processed through a new fulfillment machine. Not surprisingly, the five-year period 2008–2013 has the highest growth rate, it’s the period immediately following the launch of the iPhone, a product that accelerated the digitization of retail.

An interesting and significant trend in the sales data is the rapid growth in third-party sales, rising from 3 percent in 2000 to 58 percent in 2018. The data validate the online supply chain subscription paradigm. Many small- and medium-sized manufacturers were able to use Amazon’s fulfillment machine to quickly and efficiently serve a world of online shoppers. Jeff Bezos opened his 2018 letter to shareholder18 with a direct reference to this new fulfillment channel.

Third-party sellers are kicking our first party butt. Badly. The annual growth rate for our first-party business (1999 to 2018) is 25%. But in that same time, third-party sales have an annual growth rate of 52%. We helped independent sellers compete against our first-party business by investing in and offering them the very best selling tools we could imagine and build. There are many such tools, including tools that help sellers manage inventory, process payments, track shipments, create reports, and sell across borders.

—Jeff Bezos, 2018 Amazon Shareholder Letter18

The year 2014 was a pivotal year for the Amazon fulfillment machine, for the first time more than half the GMV product flow was from third-party sellers. Paradigm #8, supply chain subscription, was now rapidly changing the retailing world. Validation is a key step in data analysis, but an equally important step is to question whether any data segments are missing, else one could end up with conclusions from only a partial picture. If one had looked only at Amazon’s company-owned product sales, the size and growth of the fulfillment machine would have been greatly underestimated.

image

Figure 2.3 Amazon annual fulfillment costs

Building and operating the fulfillment machine are expensive, and the capital investments require a long-term view. The point-of-use delivery and free shipping paradigms require a cost-efficient fulfillment system that extends all the way to the customer address. Without cost efficiency, free shipping would be just a short term and ultimately suicidal business strategy. Figure 2.3 reviews Amazon’s annual fulfillment costs over the past 18 years19 and confirms the scale of the fulfillment machine and its progressive growth over the years. From 2000 to 2005, Amazon was building the first version of the fulfillment machine, and the fulfillment cost to GMV product sales ratio decreases from 14 percent to about 6 percent. The machine that changed retailing had a coming-out party, and Prime membership with free shipping was launched in 2005. For the next five years, the ratio remains relatively flat, but since 2010 it has been rising and in 2018 it hit 10 percent. Why this increase? Very likely the online shopping paradigm is growing faster than the fulfillment machine is evolving. Expect to see even more capital investments from Amazon, UPS, and the Walmarts of the world as they expand their fulfillment infrastructure to meet the continued growth of paradigm #1, online shopping. One view of the changing retail landscape classified retailers on an information and fulfillment matrix, four quadrants were proposed3: traditional, pure online, shopping and delivery hybrid, and online retail plus showrooms. Progressively, we would see a blending of these quadrants as the stronger retailers expanded into other quadrants.

The last data we want to review are the current state of online retail, and a snapshot view is provided by the top 10 online retailers in the United States for 2020 (Figure 2.4). The data provide several takeaways. (i) The fulfillment machine is dominant; in a remarkable statistic, we find that after two decades of online shopping one company still accounts for close to 40 percent of all online sales. (ii) A few brick-and-mortar retailers are putting up a good fight and have successfully pivoted their growth strategy to online retail, and are making significant capital and talent investments to build a fulfillment machine. (iii) The new era pure online retailers are growing fast and starting to enter the top 10. In 2018, we saw only Wayfair, the online furniture innovator, break the top 10 but others such as Etsy and Chewy are growing fast. (iv) Manufacturers or brand owners are building out their Omni channel supply and see fast online growth, both LBrands and Nike are in the top 25.

image

Figure 2.4 Top 10 U.S. companies’ online product sales

Source: eMarketer, February 2020

Four Phases of Online Retail Fulfillment

Starting from 1995 and continuing through today, four phases in the evolution of online retail are identifiable (Figure 2.5). Here, we are less concerned with the sell- or consumer side of online retail, and more on the delivery or fulfillment side. In each phase, the innovations and transitions in the fulfillment machine are distinguishable.

image

Figure 2.5 Four phases of online retail fulfillment

Peer-to-Peer (1995 to 1998): The first significant online retailer is eBay, but it did not offer a curated catalog from which buyers could select merchandise. Legend has it that founder Pierre Omidyar created eBay to sell a collection of Pez candy dispensers. In the process, he created a peer-topeer retail platform or a direct buyer-to-seller network. This innovation has been critical to the growth of e-commerce and is the principle on which many online retailers operate even today. Some key features of the peer-to-peer phase are as follows: (i) The retailer maintained no inventory, rather inventory was distributed throughout the buyer–seller network, it was an inventory blockchain long before Bitcoin. (ii) Product pricing was elastic, and a key function of the retailer was to match buyers and sellers and negotiate a price consensus, it was an online NYSE trading floor. (iii) Buyers and sellers were unknown to each other, and consequently, the transactional certainty was low. The retailer served as a guarantor or arbitrator and would resolve any product delivery or payment issue. This was a critical catalyst for the early growth of online retail.

There were a few fulfillment innovations during this phase. Transaction volumes were relatively small and offered no economies of scale. Buyers were in no hurry and speed was not essential. Most sellers were shipping products from their homes, and the existing parcel delivery infrastructure was sufficient. The one exception was Dell, which recognized early on the online shopping shift and effectively transitioned its mail-order fulfillment process for online orders.

Novelty (1999 to 2002): Profitability in-store retail is a function of scale, and this has been a key driver in the successful growth of big-box retail. A key store metric was revenues per square foot, and products with low or scattered demand were usually taken off the shelves. This constraint disappeared in online retailing, the world was your market, and products with scattered demand could be aggregated to a single fulfillment center. The original motivation for Amazon was to sell those low-volume books that were not available at Borders or Barnes & Noble. Innovation was on steroids and several start-up companies were created, many of these focused on some novelty product category. Overstock, Razorfish, Home-Grocer, Webvan, Zappos, and Pets are just some of the names.

Online sales were increasing rapidly during this phase, but the lack of a fulfillment machine meant costs were rising quickly. Unfortunately, the innovation did not cross over to the delivery side and the fulfillment strategy was simply to use traditional mail-order style warehouses. Investor capital was used to subsidize fulfillment costs, and for many of the online retailers, this solution was not sustainable. Financial challenges would force many of the start-ups to either close or be acquired. One of the prophecies of online retail was prices would be lower, and this phase proved it to be incorrect. The savings from not having physical stores were quickly absorbed by fulfillment costs. Convenience rather than price emerged as the key motivator of online shopping. Online retailers needed a low-cost fulfillment machine to grow and sustain the business. The process had already started at Amazon and its first two fulfillment centers were built in 1997.9

Machine Building (2003 to 2012): By 2003, online retail was on a secular growth trend and Amazon had over $5 billion in annual sales. The customer pivot function was sensitive to four factors: (i) Availability, (ii) selection, (iii) delivery speed, and (iv) reliability, all four were improving progressively and buyers were pivoting from physical stores to online stores. This was the golden period of fast fulfillment, the need was clearly defined, performance targets were specified, and the technology was available. I don’t know whether the big box store retailers were aware of what was evolving, or they simply chose to ignore what was happening, but during this decade Amazon was earnestly and quietly building the fulfillment machine. Teams of brilliant innovators were designing and building order processing systems, fulfillment warehouses, transportation modes, and vendor partnerships—all of which were not modifications of existing infrastructure, but new designs focused exclusively on the online customer. There was a very large-scale expansion in the number of available items, and paradigm #6—The warehouse is the store—was becoming a reality. Online customers were excited and demanding more choices, and paradigm #5—Variety multiplication—the fulfillment machine had to build speed, size, and intelligence simultaneously.

By the end of this decade, Amazon would operate over 80 fast-moving high-volume fulfillment centers in the United States, with more than 75 million square feet of warehouse space. An additional 120 centers would operate in the rest of the world. The machine was constantly evolving, and the fulfillment centers would be labeled as generation 1, 2, 3, and so on. The innovation never stops and the eighth generation was introduced in 2014 and the ninth generation is currently being built.21 During the same period, Walmart also expanded its online business, but orders were fulfilled from either store inventory or its general merchandise distribution centers. Walmart would open its first 100 percent online order fulfillment center in 2013, but it has been investing heavily in building out its fulfilment machine and in 2020 there are 25 such facilities.22

Consumer Selection (2013 to 2022): In 2013, online sales were close to 6 percent of U.S. retail sales and growing at a rate of 1 percent. In this current phase, not all the sales growth is from pivoting customers, a good portion of it is coming from an expanding selection of items. Consumer buying experts will tell you that that ease of search, ordering convenience, and price efficiency are changing the why and how we buy things. The fulfillment machine had made this transition seem easy and seamless. The expanding selection was driven by many small- to medium-sized product manufacturers and distributors setting up online stores. Paradigm #8—Supply Chain Subscription—let them focus on product innovation and let the machine builders take care of the fulfillment side of the business. Consider Jumia the largest online retailer in Africa, which is close to a pure fulfillment machine with 90 percent of its sales coming from third-party sellers. The fulfilled by paradigm is rapidly enabled by the machine. Both Amazon and Alibaba let small vendors sell and fulfill their online orders through their warehouses. Additionally, large manufacturers could bypass brick-and-mortar retailers. Price democracy was prevalent, and consumers gravitated to the lowest price. Margin redistribution was occurring rapidly.

During this phase, the fulfillment machine would sort of hit a speed limit, two-hour delivery. By 2016, it was offering two-hour delivery service for selected products in several U.S. cities. By 2018, Amazon would offer free same-day delivery for over three million items on qualifying orders over $35. Several crowdsourced delivery companies would also offer a similar service in partnership with large retailers; in this case, orders were fulfilled from store inventory. Fulfillment time equivalency, between online and physical store purchases, would have been achieved. Equivalency implied the inherent advantage of a physical store would be greatly diminished, and the fulfillment machine war between Amazon and Walmart would accelerate. In May 2019, Amazon announced plans to invest another $1 billion to expand the number of products with free one-day shipping. A few days later, Walmart announced next-day delivery on a wide range of general merchandise. Interestingly, we see the physical store retailers developing an alternate fulfillment machine, the strategy would be to use store-located or forward-positioned inventory. Buy online and pick up from stores becomes the retail mantra of physical retail chains. In the end, it will be speed and costs that will determine the winner.

Successful Machine Building

We all want to build an innovative machine that then creates an efficient business and commercially successful business. Later in Chapter 5, we will discuss different types of innovation and how they differentiate businesses. In an Internet-driven world, the business will design-build one or more disruptive innovations and these will drive the business to success. I summarize four traits a machine building team must acquire to ensure success. For each trait, I identify questions the team should ask itself before starting their machine building project. The answers will determine whether the team is ready.

Prerequisite: College courses frequently have prerequisites; these ensure that a student is well prepared for the course and reduces the risk of failure. Most machines are a complex manifestation of one or more ideas, and integrate several hardware and software components. This inherent complexity can quickly bog down an innovative idea. Why? Lack of contextual knowledge and experience. Successful innovation teams must meet the prerequisites: process knowledge, performance pivots, customer excitement, and technology trends, what is the competition, and likely capital and resources needs. A key role of venture capitalists is to check whether the team has the prerequisites, and if needed add people to close the gap.

The prefrontal cortex develops the ability to better communicate with other parts of the brain so that all areas of the brain can be included in complex processes such as planning and problem solving. The term that psychologists use for this sort of neurological maturity is an executive function. Executive function is simply the ability to see ahead and plan effectively, to connect actions to possible consequences, to see the probabilities of risk and reward.

—Rick Kalgaard, in Late Bloomers23

Executive function is a critical prerequisite for machine building; it separates the winners from the losers. Business machines involve many dynamic trends that can quickly deflate a great idea. The inability to see and operationalize the risk–reward relationship in the idea execution plan is a frequent cause of failure.

Questions to ask—What are the critical to success prerequisites for idea execution? Are these prerequisites sufficiently met by one or more members of the team? Do we have enough executive function ability in the team?

Timeliness: An innovative idea can be too early, too late, or just in time. Many of the early players in online retail did not survive because they were too early. Why? There was insufficient sales volume to sustain the business or the supporting technology and infrastructure were still being developed. Likewise, Walmart and Macy’s are innovating too late, building new fulfillment machines to catch-up with Amazon.

Why did independent sellers do so much better selling on Amazon than they did on eBay? Of great importance are Fulfillment by Amazon and the Prime membership program. In combination, these two programs meaningfully improved the customer experience of buying from independent sellers. With the success of these two programs now so well established, it’s difficult for most people to fully appreciate today just how radical those two offerings were at the time we launched them. We invested in both programs at significant financial risk and after much internal debate.

—Jeff Bezos, 2018 Amazon Shareholder Letter18

Timeliness is often not binary, rather an innovative idea or business machine is more likely to progressively meet customer needs. As Jeff Bezos describes in his letter to shareholders, the fulfillment machine, the third-party seller community, and the world of online customers all evolved in parallel. Classical economic analysis seeks that instant of timeliness, making it impossible to justify a radical idea with a delayed breakeven point. We all wonder why the plug-in electric vehicle industry was led by Tesla and not one of the large automobile companies with big budgets and deep talent.

Questions to ask—Yes, customers will pivot to the new machine, but are they ready to do it when you are ready? What are the machine-building milestones that relate capability with customer needs? What is the likelihood, and what can we do to reduce the uncertainty, that customer needs and wants will materialize?

Technology Enablers: The evolution of online retail was driven, maybe even pulled, by several technologies (Figure 2.6) that powered how we, and where we, connected with the Internet. None of these were sudden developments and their introduction and availability were widely advertised. Successful machine builders had the innate ability to utilize not only proven technologies but nascent technologies in their designs and solutions. They also had the prerequisites and futuristic insight to project how the technology enablers could be exploited to disrupt a business process, value chain, or entire industries.

image

Figure 2.6 Online retail technology enablers

When you are coming up with product ideas such as the iPod, do you try to solve a problem? There are different approaches—sometimes things can irritate you, so you become aware of a problem, which is a very pragmatic approach and the least challenging. What is more difficult is when you are intrigued by an opportunity. That, I think, really exercises the skills of a designer. It’s not a problem you’re aware of, nobody has articulated a need. But you start asking questions, what if we do this, combine it with that, would that be useful? This creates opportunities that could replace entire categories of device, rather than tactically responding to an individual problem. That’s the real challenge, and that’s what is exciting. We don’t do focus groups—that is the job of the designer. It’s unfair to ask people who don’t have a sense of the opportunities of tomorrow from the context of today to design

—Sir Jonathon Ives, Chief Design Officer at Apple24

While it may seem many online retailing innovations were accidental, the reality is that many innovators saw, planned, and executed their strategies in anticipation of these technologies. The value of these technologies is often not obvious to the final users. Forward-looking innovators, such as the designers at Apple, imagined how existing and upcoming technologies would enable new business machines that added exciting value for customers and employees.

Questions to ask—What technologies and infrastructure do you need to support the market release of the machine and are they readily available? What value opportunities will the machine create from the current and future technology enablers?

Long View: It is often mistakenly assumed that a Eureka moment occurs when you have an instant revelation, idea, or success. Rather, it usually denotes the culmination of long ideation, development, and analytical process. The machine-building highway is littered with ideas that ran out of capital, talent, and customers. There was no eureka moment for these innovators. When management, investors, or the builders themselves have a short view then the build process is handicapped from the get-go. The basic premise that increased levels of resources will result in faster project completion and success is not universally true. Sure, if you are constructing a residential tower or building a bridge, this assumption holds. But many business machine ideas are dependent on external evolutions that are outside the builder’s control.

I think what Mr. Buffett realized in 1969 is that being a long-term investor with short-dated capital is just ultimately going to lead to a bad outcome at some point in time.

—Bill Ackman, CEO Pershing Square Capital,
commenting on Warren Buffett’s transition from managing
a hedge fund to running a publicly-traded company—April
2019, 13-D Active-Passive Investor Summit25

Ackman’s comments hold true in building the fulfillment machine. Paradigm #1—online shopping, provided a natural impetus for quick revenue growth, allowing many early online retailing innovators to succeed easily. In many cases, their backers (venture capitalists) were holding short-dated capital, and investments were directed to immediate and visible success. Little-to-no innovation talent or investment capital was directed to fulfillment; the assumption was that it’s a commodity service provided by the brown truck guys. In stark contrast, Amazon starts building the fulfillment machine within a few years of its start. Later, many naysayers accused Amazon of subsiding delivery costs with investment capital! The instructive take way from a review of the online retailing evolution: machine building requires a long-term view supported by long-term capital and acceptance of short-term deficits.

Questions to ask—What is the time series value profile of the idea, that is, revenue and profitability projections? How long is long, when will investors, management, and/or the team lose patience or when will we run out of capital? How short is long, when will an innovative idea just become another idea?

I am always intrigued by the profiles of serial entrepreneurs or serial innovators. How is it that the same person is involved in multiple business successes, while so many other smart, and possibly smarter people, have failed to design-build their single idea? I suspect these serial geniuses were gifted with the foresight to plan for the four traits. They would put together a good team, build the four traits, and then gestate the idea successfully through an innovation process.

Chapter Summary

Reviews how data and technology were used by several successful innovator-led companies to facilitate and validate the paradigm shifts listed in Chapter 1.

Investigates the growth of Amazon with an in-depth analysis of GMV product sales and the third-party seller ratio. The importance of fulfillment as a subscription is highlighted and the enormous value to small sellers becomes apparent.

Building and operating the fulfillment machine are expensive, and the capital investments require a long-term view. Analysis of Amazon fulfillment costs reveals that the fulfillment cost to GMV product sales ratio decreased from a high of 12 percent (2000) to 6 percent (2009) but has been increasing since and was back to 11 percent (2018). The point of use delivery and free shipping paradigms cost money, and businesses need to be ready for big investments as the start machine building.

The innovations and transitions in the four phases of fulfillment machine evolution are identified and discussed.

Four team traits for successful machine building are proposed. For each trait, prerequisite self-check questions are presented.

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
54.165.248.212