|
(May 2003)
Paper's Table of Contents:
Overview - e-commerce is just business
The aim of this chapter is to provide practical guidelines for building effective business models and practices for new generation e-commerce. New generation here does not mean next generation innovations in either the electronic or commerce components of e-commerce. New generation refers to the plans made today to position for the markets and opportunities of the next 2-5 years. New “generation” implies an old generation. The simultaneous and interrelated crash of the NASDAQ, the collapse of the dot com boom and the recession that began in 2000 are the obvious marking of a shift that merits the term generational. The relevance of the old generation is for taking stock and drawing lessons but only for the purpose of moving forward.
The analysis and guidelines in this chapter are based on what may be seen as an unusual or even provocative claim. This is that the dot com collapse is largely irrelevant to any discussion of e-commerce. The more interesting and useful perspective is to view the old generation – Chapter 1 of a long story – as a large-scale set of experiments from which no general extrapolations about e-commerce can be made. The dot com crash tells us little about e-commerce as a whole, for two reasons. The first is that e-commerce far predates the Web and the dynamics of e-commerce revenue and cost structures, consumer response, and patterns of innovation are very much the same for the pre-Web as for the Web era. The dot coms have always been a tiny fraction of total e-commerce, under ten percent of the total U.S. Internet economy At their peak in 2000, according to University of Texas reliable and non-hype annual review and between one and two percent of total business revenues.
What was truly new about Chapter 1 of Internet e-commerce, the dot com surge, was that it created a volatile, varied, risky and often exotic laboratory. Just about every experiment in market targeting, pricing strategies, partnerships, new technology, auctions, uses of technology, promotions, alliances, customer segmentation, financing, products and services that the human mind could articulate as a “business model” and convince someone to finance got its chance.
A laboratory provides lessons and insights from individual experiments but does not throw much light on general patterns and trends. The discussion in Chapter 1 of Internet business was thus basically all about outliers. In an immature field explicitly driven by radical innovation and search for “new” business models, there are few patterns, little trustable evidence of long-term trends, and no normal distribution of samples – companies, customers, costs, industry sectors – from which to derive reliable lessons from experience and hence make reliable recommendations for new action. There are no “averages”, only individual cases. Striking instances then become the base for analysis and extrapolation. Because there is no normal distribution to compare against, it is completely unclear whether the companies will turn out to be just outliers, way off the scale, or represent a new average, a pacesetter or a soon-to-be mainstream.
In the euphoria of the dot com surge, commentators generated many assertions about the future of e-business from a small number of firms: Amazon, Ariba, Cisco, Priceline, Webvan, and the like. After the crash, the fashion shifted to extrapolation from disasters such as Boo, Exodus, Webvan and Ariba – often the very same companies but with opposite interpretations. Amazon remains a barometer for e-business as marking either sunny weather in store or storms on the horizon; assessments of its status and likely future are almost an e-commerce mood meter. The situation for the trade press has become analogous to the competition that the London Times newspaper held for its journalists in the 1930s: to create the world’s most boring headline. The winner was
“Small Earthquake in Chile: Not Many Hurt.” It became boring to write a headline
“Many Dot Coms Profitable.” It is as boring to write one that states
“eBay continues to do well” as it is to tell the world that
“Wal-Mart continues to do well.” As the fashion shifts, the headlines reverse the tone. Recently, the new headlines are largely about the successful embedding on the “e” in the “c” as more and more firms harmonize their channels, melding Internet, call center, physical locations and distribution services. There is still no statistical normal distribution and there remains a wide gap between e-commerce leaders and laggards, but the news is that e-commerce is not headline material for articles about dot bomb, dot con and Internet company shares selling for less than their equivalent as paper to decorate a wall.
Headlines are not needed for new generation e-commerce to move ahead and outliers should be treated as just that until there is a body of experience and stability to distinguish six-sigma special cases from one-sigma leaders and laggards clustered around a meaningful average. We have such data for supply chain and logistics, because this highly successful component of e-commerce reached momentum and payoff well before the Web. In the six-sigma headline mode of discussion, this precursor of what is now subsumed under the general label of B2B e-commerce was initially seen as sure to change the entire nature of commerce, with Cisco and Ariba two of the headline makers. Post-dot com collapse, it looks like a disaster; Cisco plunged and Ariba still struggles to stay in business. Focus on the history of supply chain management and the picture changes entirely: the University of Maryland’s in-depth studies and surveys show that over the past decade – pre- and post-Web – the percentage of U.S. GDP tied up in inventory and supply chain management costs as dropped by 40 percent, that in any industry the top ten leaders in the use of electronic SCM tools – including but not confined to the Internet – use half the working capital per unit of sales than their median competitors, half the overhead and half the inventories. Yet, B2B is in disfavor and, yes, Ariba and Cisco moved from exemplar to cautionary tale. So?
There are plenty of useful lessons to learn from these and many other B2B fables – stories with messages – but they are more about business and organizational issues for any high growth company in any volatile industry. Ariba did a great job as did Cisco. They both misjudged some business trends that had little to do with e-commerce and nothing to do with dot coms in general. Basically, Ariba did not recognize the sensitivity of its business to the high tech market; when that slumped, it was almost inevitable that any broker and intermediary whose revenues depended on their health would be caught up by the drop. Cisco misestimated the telecommunications industry slump and mistimed its inventory cutbacks.
The second reason for ignoring the dot com collapse as the base for a guide to new generation e-commerce strategies and opportunities is in the form of a decidedly non-boring headline:
“What Collapse? It Was Just Business As Usual.” The crash of so many dot coms was not a collapse of e-commerce; the most striking evidence for this is that the pattern of growth in e-retailing continued to be very much the same in the recessionary period of 2001 and 2002 as it had been in the two preceding boom years. Online business remained a small percent of the total economy – around 1% in 2000 and on track for 2% by 2005, a figure far below the exuberant forecasts of the boom period but still solid commercial growth and very close to the (short) historical rate of increase since the launch of Amazon in late 1994. Online retailing grew faster than did non-online business, even though so many big names like eToys had gone under and Toys”R”Us now relies on Amazon as its service arm to help it survive. Amazon is in far better shape than Kmart. Business Week in mid-2003 reported that far from being “bubble-era hype” was “stronger than ever.” Its special report includes the following claims: U.S. networked business-to-business transactions now amount to $2.4 trillion, consumer e-commerce was on track to reach close to $100 billion by the end of 2003, and that anyone who invested $1,000 in every single dot com e-retailer, would have earned a 35% return.
The dot com phenomenon, in all its phases, should not be viewed as at all unique. The lead up to the crash followed the same pattern of previous business expansion and innovation generated by deregulation and globalization (a form of quasi-deregulation of supply sourcing and market barriers), which in so many ways is what the Web has added up to. The pattern is apparent in the history of the U.S. airline industry, telecommunications worldwide, PCs, and electrical utilities. The first phase is that nothing happens. The status quo holds. In telecommunications, five years after deregulation in the U.S. and UK, AT&T and British Telecom still maintained around a ninety percent share of the long distance phone market. The Internet had been in operation for well over a decade before the Web browser began the transformation of Internet operations from a domain limited to academic and technical professionals to a browser on every PC.
Then, often for no apparent reason, some new player does something interesting. In e-commerce the something interesting included Amazon’s sudden success, Priceline’s new slant on pricing and eBay’s invention of the online yard sale. In the airlines, it was Southwest’s bypassing of the basic and very successful business model of American Airlines and British Airways: hubbing. It built its strategy on highly selective point-to-point routing and a new style of financial operations and customer care. In telecommunications, MCI’s launch of the first billion dollar telecommunications product, Friends and Family, began the breakdown of the telecommunications industry establishment’s control of the industry through the strength of branding. In the utilities industry, Enron was just one of the many companies to view power supply as just another futures market.
Such innovations capture interest and imitation and there is a resulting explosion of new entrants challenging the status quo of the industry and often claiming a new edge – business model – in their thinking, organization and operations. In the airlines, thousands of startups bought second hand planes and competed on the basis of a flexibility and small-is-beautiful model that overlooked issues of branding and “distribution” – marketing and reservation systems. The Internet saw the same pattern except multiplied by thousands across all dimensions of industries, small and large firms, new and established entrants. E-retailing took off, online auctions were everywhere and travel became one of the main consumer e-commerce markets.
At some stage and for whatever reason, realities of commerce start showing up and relatively suddenly there is a shakeout followed by a massive and often long consolidation. In the airlines, the realities of scaling and marketing began to bite into the hopes and forecasts of the new players. In e-commerce even before the dot com collapse, the dynamics of successful versus successful e-retailing were becoming very apparent – customer acquisition costs, process capabilities in distribution and order fulfillment, margins, etc. The shakeout was brutal and left many dot coms dead in the water, just as most airlines and telecommunications companies were left stranded well before the recession and 9/11 crashing of the airlines and the 2002 recognition of the overcapacity, misinvestment and accounting frauds and fiction of the leading telecommunications players. We do not label these as the dot jet collapse or the dot wire crash but as business failures. The dot com equivalents were business failures, often in business basics. After the shakeout, there is a long period of consolidation and a new establishment emerges. The airlines are back to a small number of megaplayers and consortia, with many well-known companies gone or likely to go. Obviously, Internet-based e-commerce is now in a post-shakeout consolidation phase. That consolidation often puts “innovation” and expansion on hold and looks for efficiencies, cost savings and rationalization: business basics.
Business basics are the issue. At some stage, the E of e-commerce will disappear, as it already has in banking and it will be integrated into everyday business and not seen as something distinct and often separate. The contrasts of clicks versus bricks and then clicks and bricks will become “what’s the difference?” In the mid-1980s, ATMs, electronic cash management systems and financial EDI (electronic data interchange) were all “electronic banking”; now, they are just banking, to the degree that a major problem for startup online banks was their lack of physical ATMs. Technically, an ATM is a general purpose computer with the distinguishing feature that a card activates it. That terminal can increasingly link via standard servers and switches to a range of financial communication networks, but also to government social service payments systems and license renewal processing systems, and even university administration systems. (There are many case examples of the exploitation of the ATM as in effect a portal.) Most ATM users do not think of it as a ”computer” or anything special. It is just a service point.
We are not yet at the same stage in online business, but the question now is not if but when this will happen and how to prepare for it. E-commerce today is a matter of emphasis – literally so – and where policy makers and executives place the emphasis strongly influences just about every element of their strategies. In its early stages, e-commerce was very much Electronic commerce – big E and little C. The focus was on the technology as the driver of business: the Internet as infrastructure, the Web browser as link to customers, intermediaries and suppliers, and a host of new software and data management tools as the foundation for product and services design.
Even before the dot com shakeout, the recognized challenge had become how to mesh technology and business and move on to Electronic Commerce, Big E and Big C. The new priorities became to use the power of E to first build revenues (instead of just “hits”), and then to turn revenues into profits. Many companies could not make the transition, but far more did than most reports suggest. Business Week stated in mid-February 2003, that over 40 percent of the 208 publicly traded Internet companies are now profitable, up by almost double from a year earlier. It comments that “In key areas such as e-tailing and online finance, profitability has become the rule rather than the exception. And those profits are measured by generally accepted accounting principles – no “pro forma” tallies need apply.” Around the same time,
USA Today reported that over half the companies on its Internet 50 list that mixes both large and small firms were profitable, up from half that number the year before. In other words, in one of the worst periods in recent business history and the worst for e-commerce, in the boring headline of the newspaper,
“Profit possible in post-bubble tech world.” (February 13, 2003) It is worth noting that both of these publications focus on averages not outliers; the data are now available and comparable. And it is the about EC, not E-commerce.
The routinization of the e-commerce technology base
Today, we are very much moving on to the era of little E and big C: electronic Commerce. It is commerce that motivates this shift, not because the technology is a secondary factor but because the IT base for e-commerce is moving rapidly to becoming the equivalent of electricity: a set of standard interfaces taken for granted by the user, a massive infrastructure of technology utilities that link automatically and directly from producer to distributor, a variable cost pricing scheme, and clear standards that enable product and service designs. Outside the mainstream of e-commerce are many developments that do not as yet fit into this emerging e-commerce base. These include advanced multi-media interface tools (such as Flash MX), many aspects of wireless services and proven applications and tools that demand too much telecommunications bandwidth or add too much “latency” – delay – in processing to be fully practical in everyday usage. But all these and any future innovations are explicitly being designed to fit into this electricity-like infrastructure base; they have to for them to be successful. In a sense, this move to technology as routine is the e-commerce supply side’s basic business model. Even Microsoft is moving in this direction. Its packaging of its own Web services capabilities, .Net, is being marketed as the best way to achieve goals of modularity, integration and speed of development for Web-based initiatives, not as being a different direction and pathway. Similarly, IBM’s strategy is explicitly committed to the very opposite of its previous history: on-demand, pay as you go services via “grid” computing – a deliberate equating of Web-based services with the electrical grid.
These trends mark the increasing subordination of E to C in e-commerce planning, service design, delivery and operations. In terms of both the E and C of e-commerce, the hype is over. The initially underestimated difficulties of actually executing rather than announcing a business model are fully apparent. There will be no more grand multi-billion dollar experiments funded by naïve investors. While the technology continues to evolve and thus open up new e-commerce opportunities, it is now almost axiomatic that in and of itself successful innovations in technology do not directly lead to innovations in their use and impact. The most recent illustration of this point is mobile commerce. Here, the technology of 3G, WAP, and Web-enabled phones was predicted to create massive new markets. So far, they have not done so and much of the turmoil of the telecommunications industry comes from its repeating the big E, little C approach of most of the failed dot coms; the $200 billion spent in Europe alone on 3G licenses makes the dot com spending appear almost like petty cash. The industry has flailed in its search for effective business models; in the context of Web commerce, it is noteworthy that it typically calls these “revenue models.”
By contrast, the Web is now as routine in everyday life as the ATM. This is the first generation of business and society where the term “high tech” is a misnomer. The technology core of e-commerce is simply No Big Deal for most organizations and individuals, including in urban areas of countries that previously lacked telecommunications infrastructures and access to IT resources. There is no mystique left in “high tech” in an era of MP3, PDAs, Askjeeves, DSL, eBay, $300 full function desktops and $700 laptops, wi-fi and portable DVD players. Amazon.com is just another retailer. The IT hardware manufacturing industry is an extension of consumer electronics; PCs, laptops, mid-range servers and storage are globally sourced as components to be assembled and branded in the same way as televisions and camcorders. Dell now competes on its own terms with the branders of PDAs, storage, and servers. Companies like Sun Microsystems, EMC, and Compaq have lost much of their ability to set the terms of competition; their premium hardware has been commoditized and it has become harder and harder to find a new premium offer, through software, architecture and high- end product performance.
IT relatively suddenly became a mature industry at the very same time – mid-2000 onwards – that its stock market premiums collapsed. The PC industry is the leading example here: rapid market saturation, commoditization, overcapacity, and price erosion. In software, IT services and uses of IT in administrative functions, back office processes and call centers there is a rapidly growing emergence of global outsourcing and mega-utilities. 5-7 year deals in the $1-10 billion range are, if not commonplace, sufficiently frequent to merit little notice in the press. Many countries are joining the global outsourcing community. These are not casual deals. Apart from being big in financial terms, they are a major organizational shift. They are a recognition that there is no reason to retain in-house the many standard functions that can be both better and more cheaply handled via the transportability of business processes and either their technology base or access to it via telecommunications links. They are a more implicit recognition that there is no reason to retain in-house IT resources that can be both better and more cheaply handled by specialist firms abroad. The main message from just about every study of back-office and business process outsourcing is that the skilled labor needed for new generation technology application is available in many regions of the world – including those of traditional IT but also in many countries that are categorized as “lesser” developed. The Philippines, India, Mauritius and Slovakia are examples of new e-commerce skill pools and service providers.
IT is as a result of all the trends towards routinization moving from being a high upfront capital cost – with very high risk in terms of implementation, adoption and payoff – to a pay-as-you-go variable cost, including for many skills that in Chapter 1 of Internet e-commerce were very scarce but are now commodities. In 1998, anyone who could write HTML code for Web sites was a “Webmaster” and earned around $600 a day. Now, HTML is “Save As” in Microsoft Word. Today, XML skills are at a premium. Microsoft has announced that XML, too, will soon be a “Save As” option in Word. In 2000, no company could find enough Java programmers. Today, quite literally, it is hard to give them away; a number of software development and systems integration firms that tried to ride through the recession and retain this expensive talent found that clients would not pay for them even at cost. Mid-range skills are more and more being sourced globally; when you phone AOL for technical assistance, you get through to Makati, in the Philippines. That is where the Love Bug virus came from – his fellows describe the student responsible for it as pretty average compared to the rest of his class.
More and more “development” activities, are being handled not just outside IT departments but by teams of business people supported by just a small number of IT specialists, mainly in the areas of data quality assurance (including security), integration and architecture – these are IT skills which will not rapidly commoditize. These teams deliver results in 90-180 day “ventures” rather than the large-scale, long schedule standard IS projects. Examples are business process management (not at all the same thing as reengineering but with the same target, carried out more cautiously, in smaller increments, and with better process mapping and implementation tools), data mining, new generation e-commerce and Web-based services, and multi-media systems.
These are the new mainstream of IT application and thus of new generation e-commerce. They can only increase, not decrease and, correspondingly, reliance on traditional IT systems development staff for e-commerce service development can only decrease. One of the implications of this routinization, even commoditization, of the IT skill base is that any region or country can be a player in e-commerce as niche supplier. Slovenia, Hungary, Jamaica, Bermuda, India, the Philippines and Mexico are examples. In its short history, the Web takeover of e-commerce from EDI and industry-specific services (online banking, airline reservations, retailing procurement, etc.) was very much driven from centers of “advanced” skill resources, most obviously Silicon Valley. As that situation changes along with the changes in the nature and supply of the e-commerce technology base, then obviously the options for supplying and obtaining the E also change rapidly. For chapter 1, technology and technical people were at a premium. They will both remain so in specific areas; end-to-end security and information assurance or RFID (radio frequency identification) applications in retailing are examples of today’s premiums. But Chapter 2 of Web-based e-commerce is already being based on sourcing, not building, standard technology.
It will also benefit from the shift in development of applications and services that has been building up since the beginning of the Web explosion and is gathering increasing and continued momentum. In the technology itself, the Web is being augmented by standards and tools loosely termed Web services that facilitate this trend by making it more and more practical to build and integrate new services in small modules. While there are many gaps in what may be termed Transitional Web services – those that are clearly sure to become part of the enterprise- and interorganizational platforms of new generation e-commerce but are not yet fully stable, secure, and proven – Foundational Web services are well in place (the key acronyms here are XML, the enabler of total sharing of data and electronic documents, SOAP, http (the core substructure of the Web), and C++, Java and their progeny). After forty years of sustained effort, the software field is close to its goal of building large-scale applications out of self-integrating, reusable “objects”, “components”, “applets” and the like. The browser wars – Netscape versus Explorer – are long over, as are the operating systems battles. Windows of course owns the desktop but Linux and Unix now co-exist with it. The operating system is a commodity. More and more organizations speak of becoming “agnostic” in the areas of browsers, operating systems and data management platforms that previously dominated IT industry competition and large organizations’ IT strategies.
This routinization of technology platforms has three major implications, all of them highly positive for new generation e-commerce. The first is that most of the infrastructures are in place and affordable across regions, markets and demographics. This obviously shifts the focus of e-commerce innovation from building technology capabilities and capacity to building business ones. The early e-commerce pacesetters uncovered many technology problems, such as scaling – being able to rapidly expand IT operations to handle often explosive increases in volumes of users and transactions and more recently decreases in them – security, privacy and protection against the almost open invitation to launch their viruses that hackers from many places and with many motives saw in the Internet. In general, the focus on IT-building got in the way of business management. In retrospect, it is obvious that many e-commerce players lacked basic skills and staffing in such areas as finance, human resources, and operations. In others, the primary of the technology imperative created cultures that neglected or even disdained general management skills and blocked the efforts of executives brought in to replace floundering founders.
The routinization of the e-commerce technology base thus frees up management attention, resources and skills. It makes absolutely no sense whatever for any company, except a few megaplayers in IT rather then just e-commerce, to build and maintain its own e-commerce technology operations. The second major implication of the routinization is that the shift from own and operate to rent and pay as you go is the guaranteed path for standard uses of IT. Had there been in place today’s technology industry, set of basic Web-based standards and availability of skilled and reliable outsourcing firms, e-commerce would have followed an entirely different finance path than it did. Fundamentally, the entry fee for e-commerce was upfront capital to be invested in two main areas: IT and customer acquisition. These fixed costs are analogous to R&D: invested yesterday and a drain on today’s cash flow and profits in the interests of profits tomorrow. As so often with IT, the full cost burden was underestimated by factors of 10-100. The Web site that could be built for, say, $100,000, cost millions to integrate and operate as an e-commerce platform that must link to legacy systems, other partners’ services and complex data base systems. In many ways, it has been the equally underestimated cost of customer acquisition that caused the crash of many online retailers and financial service firms.
The routinization of the technology platform transforms the upfront financial capital burden of e-commerce initiatives. It does not remove the burden of customer acquisition, however. This means that it increasingly routinizes technology costs and operations but not the commerce costs and operations. At the policy level of cities, regions and countries, the economic development issue here is whether these infrastructures will be publicly funded or left to the market. Whatever the answer, it is very clear that new generation e-commerce rents the E component rather than builds it.
Making sense of the e-commerce Chapter 1 laboratory
That means that the differentiator is commerce. At one level, that is commonsense. However, in the early surge of the dot coms the assumption was that this commonsense meant that new business models would be key to success – a different form of commerce. The evidence from the dot com laboratory is that it is more the execution of the model than the model itself that makes the difference. Each of the main innovators claimed some degree of uniqueness in its business model. But the patterns showed basically the same uniqueness just about everywhere. An insightful set of case studies edited by Stephen Elliot shows that the dynamics of e-commerce innovation, including explanations of success and failure, are very similar across the countries studied: Hong Kong, Greece, Australia, Denmark and others. The choice of initial target markets – B2B in the business sphere and retailing, travel and basic financial services in the consumer arena – are almost identical. The problems of management, scaling, and revenue-building are structurally the same, however much their situational differences.
Examples of situational differences that do have substantial impacts on e-commerce targeting and evolution are the relative costs of Internet telecommunications access and ISP services, and the availability of consumer credit. The latter is a surprisingly understudied element in the e-commerce field, even though it may well be the single most differentiating factor for new generation e-commerce as it spreads across the world and penetrates the SME (small and medium enterprise) markets, in terms of both e-commerce sellers and buyers. In the mobile phone market, consumer credit has led to the popularity – even necessity – of prepaid phones in the UK and their almost complete absence in the U.S., where very few users would have a clue what a SIM and “top up” are. However, the same structural patterns of mobile phone adoption and usage are apparent worldwide, reinforcing the view stated here that e-commerce in general follows fairly common paths of both supply and adoption.
The increasing routinization of the e-commerce technology environment is encouraging for regions and economies that were situtationally left behind in the early e-commerce surge, for whatever reasons: the high cost of Internet phone access and ISP services (Japan, France), lack of technology and telecommunications infrastructures (Africa, rural regions in developed as well as underdeveloped countries), monopolistic telecommunications regimes (India, Japan) or political restrictions (China, Singapore). The routinization of the e-commerce technology infrastructure is removing most of these constraints but more importantly for national and regional development, the commonality of patterns of e-commerce innovation identified by Elliott and others suggests that the lessons of the dot com era in the U.S. can be used to guide selective prioritization and investment, at both the policy and individual firm level.
If that first era is viewed as a giant laboratory, there are so many instructive lessons to be gained. The disasters are then actually of value; they moved e-commerce forward even though, of course, they decimated their investors’ portfolios. The morality – even legality – of the financial dealings of the dot coms, VCs and Wall Street analysts with vested interests in the price of the stocks they analyzed obscures just how much of the dot com phenomenon was about e-commerce and how much was straight financial manipulation. In any case, outliers were the news for investors, analysts and business observers. In the early surge of dot com hype, hope and billion dollar gambles, wild extrapolations were made from individual cases to produce assertive claims about the death of bricks, then no – it’s clicks and bricks, no – the bricks have won, Amazon is dead, no – it’s on track, and the like. Forget the extrapolations. Even today, we lack the time series and sample size to make reliable statistical sense of e-commerce. There are few meaningful averages and standard deviations to guide future management and policy decisions. Yesterday, the extrapolations were very much made through rose-colored lenses. Every e-commerce “success” was to become the norm for entire industries. Today, the extrapolations are generally filtered through dark gray lenses that see individual disasters as the inevitability of the future.
Cases are just cases. Most of the first surge of e-commerce was the equivalent of a set of experiments in a high risk, high capital cost industry like pharmaceuticals. Most of them did not work out. That does not necessarily vitiate their messages. The explosion of B2B and subsequent crashing of the many B2B portals like Ariba, for example, highlights the complexity of process change in organizations; B2B relationships turned out to be highly dependent on existing processes, incentives, existing relationships, and internal systems. These eroded the value of software-based portal processes. The lesson should not be a dot com extrapolation but a recognition that business process management must be the driver of B2B. Similarly, the lesson from Priceline and others is that variable pricing does work, not that Priceline would or would not be a success.
Lessons from the Chapter 1 laboratory; the fundamental cost and profit structures of the firm
The start of Web e-commerce take-off can be appropriately marked by the late 1994 launch of the first version of Netscape’s Navigator browser and Amazon’s initiation of full operations. If there is just one lesson to take from the multiplicity of “experiments” between then and the April 10, 2000 crash of the NASDAQ that signaled the demise of the dot com era, it is that the fundamental issue for every single player in e-commerce was and remains the difference between the cost structures of the technology-dependent firm and the more typical organization. The diagram in Figure 1 below shows the difference in a schematic form.
Figure 1: The nature of cost structures in online
business
The traditional firm has relatively low fixed costs and high variable costs (labor, inventories, operations support, etc.) This means that its gross margins are low. Below the break-even point shown in Figure 1, it loses money of course but can generally work its way through short periods of tough times, mainly by reducing its variable costs, especially labor. Above break-even, it makes money but in most industries net operating margins are well below 10 percent. By contrast, gross margins for digital services are in the 85 percent range (eBay is an example). Fixed costs, however, are extremely large, as discussed below. Above the break-even point, the high margins lead to massive profits. Below break-even the losses are just as massive, with the fixed costs a burden that cannot be borne for long.
The stereotypical dot com was built on two high (often very high) largely fixed costs that demanded plenty of financial capital: IT platforms and customer acquisition. Customer acquisition here relates to the marketing, advertising, promotion, discounts, commissions and referral fees that it must spend in advance in order to build its customer base. While figures vary, most analyses indicate that it costs around $50 in retailing and closer to $200 in financial services to get a new customer to make a first purchase. This is not out of line with the costs in traditional insurance and start up of magazines. What was new was the scale of the market needed to recover the costs of customer acquisition and technology platform. Building a customer base of 1 million meant spending at least $50 million and more often $100 million. The dot coms largely assumed that the percentage of “hits” – the number of customers who contact the site – that would turn into sales would be high. This was very much an E-commerce assumption, that the technology itself would generate the business. One typical study in 2000 showed that the average large-scale e-commerce Web site generated 1.8 million hits a month. 15% of these hits turned into transactions. But the number of repeat customers was just 24,000 or 1.3 percent. The average purchase transaction was around $30.
The math just could never work out into a profit. $50 to generate a gross revenue of $30. Up to $300 million to build and brand a household name (the estimate of Charles Schwab’s head of advertising, who was a veteran of the comparable AT&T, MCI and Sprint long-distance phone service wars). In mid-2000, most dot com retailers were spending 65% of their revenues on marketing. The costs of building and operating a technology platform that could process 1.8 million accesses just to build a repeat customer base of 24,000 were in themselves bigger than the revenue base of most players. It is not surprising that so many dot coms could not sustain any growth.
Without going into details, this simple diagram explains so many of their failures. Even today, many online companies ignore the imperative of turning hits into transactions and transactions into repeat business. Many of them do not, for example, according to a 2000 survey respond to customer e-mails (40%) or even recognize that the customer is a repeat buyer (75%). Others do not track or respond to the customers who abandon their shopping cart – an indication of a strong interest in making a purchase but some glitch in completing it. Most companies’ business models assumed that online advertising would generate a whole new industry and revenue base. Today, banner ads are just a way of filling up the screen with displays most of us ignore. All in all, too many dot com players had business models that were not profit models in definition and execution. At best, they were “hit” generator models.
Only a few firms – Amazon, AOL and Schwab are noted examples – recognized that e-commerce success in the consumer market rested on repeat business and on gradually reducing the high fixed cost burden of technology and acquisition outlays. Amazon’s business model was based from the start on this target. Since, as stated earlier, there are so many problems in assessing dot com accounting figures, it is not clear if Amazon is really profitable for the long-term but its pro forma profitability in 2002-2003 was significantly affected by its reducing the percentage of its overall revenues it spent on these two capital investments. (Technically, most of these expenditures were expensed just as R&D is, but they were clearly capital rather than expense in that they provided no short-term payoff and were a high risk investment in the firm’s future.)
But there is yet another catch hidden in the diagram above. The dot com, technology-dependent firm gains the benefit of very high operating margins that, if and only if it got above breakeven in recovering its programmed technology and customer acquisition costs. Then it made huge amounts of money. This was the case for AOL and Yahoo at their peak. AOL spent a decade and half a billion dollars in marketing before it broke even. It was able to generate $1.2 billions a year in cash flow on revenues of $6.5 billion (its “run rate” at its 2000 peak). Yahoo had gross operating margins of 80%. EBay’s were 74%.
The catch is that the cash flow machine that is operating above break even is a disaster when volumes slump. Schwab had dominated the online securities market through its customer service, customer retention and cross-selling of services from many other providers. It was able to charge $29.95 for a transaction that others provided for as little as $6. In a crowded competitive field, it achieved a 30% market share. Then the stock market crashed and the day traders fled. When high margin revenues disappear what is left in the dot com cost structure is all the high fixed costs, in Schwab’s case mostly technology and staff costs. For AOL and Yahoo, what became apparent was how much of their revenue – and Amazon’s, too – was based on the volumes of other e-commerce companies, either through fees they paid as commissions or for advertising or through often complex financial deals in which the companies took an equity position in other, smaller providers in return for payments they booked as revenues. The profit structures of the dot coms as a whole were highly sensitive to volumes. AOL derived much of its income from creative – and legal – accounting for business from its partners. It also failed and continues to fail to extend its income from its subscriber base. That base provides a monthly fixed flow of revenue but subscribers have shown little interest in paying for other services.
In many ways, the fundamental business issue for IT in general has always centered on the financial structures of the firm. IT has substituted fixed cost capital for variable cost labor, raising break even points, but raising variable margins. The diagram used to illustrate the discussion was introduced in the is author’s 2000 book From .Com to .Profit – or so he thought. Skimming through his 1986 book
Competing in Time: Using Telecommunications for Competitive
Advantage, he found the very same diagram summarizing “The Impact of Electronic Delivery Base on Costs of
Services.” (Page 50)
Beyond showing the impact of age on loss of memory, there are several points relevant to the 2003 discussion here. The main one is that the structural issues remain the same but the situational ones have shifted. The business logic underlying the diagram has not changed, but the nature of telecommunications costs has and continues to change. For new generation e-commerce, the opportunity is to shift from fixed cost investment to variable cost contracting. In 1986, the fixed costs of telecommunications were very high indeed, with only large firms having the scale to lease expensive data lines, build global networks or even afford 800 numbers. This meant that only a relative few players exploited what remains the core of business-to-business e-commerce: electronic data interchange. The technology base needed to enter the e-commerce market was prohibitive through the mid-1990s, which is why, though accurate, the analysis in Competing in Time was mainly of interest to large banks, airlines, utilities, etc.
What changed the situation was not the Internet as such. It was really in the context of EDI and standard industry e-commerce just an alternative delivery infrastructure and even today pioneers like Wal-Mart and American Airlines still rely on their core “proprietary” systems that are optimized to their own business patterns, security needs and services. These are, of course, evolving towards Web services and already incorporate links to the Internet and use of the Web. The Web introduced two new elements to e-commerce, both of which relate to the cost structures of the firm. First, it reduced the e-commerce entry cost for midsize and small companies in terms of both building their own capabilities and services and reaching customers. Secondly, it lowered the fixed cost investment and time needed to build a large-scale e-commerce technology base. While the costs of building these turned out to be far heavier than the low cost of the front-end Web site made it appear, the Web still lowered the investment base. There is no way that a startup firm like Amazon or eBay could have built a capability to serve millions of customers through traditional methods – including methods of financing. The earlier e-commerce innovators paid for their infrastructures out of their operations and established sources of equity and loans. It was the new financing levers associated with the Web that made the investment difference.
For new generation e-commerce, this pattern of cost structures – fixed versus variable, steepness of the margin line, and sensitivity of payoff to volumes changes in a key new way: the fixed costs become more and more variable costs. Scale can be added in smaller increments. Here is a simple guideline for new generation e-commerce basics, for both policy makers, the e-commerce technology and services supply industry and e-commerce players, whether dot coms or established companies incorporating electronic channels and services in the same way that banks incorporated ATMs and cash management and made the “electronic” term redundant: Think electricity. This is the real business model for e-commerce – and for information technology in general. IT must become a variable cost for its users instead of an upfront risk capital investment. For its suppliers and users, it will become a utility for basic everyday operations. The utility blueprint is Web services as an evolution path. For investors, the business model credibility is in the C of e-commerce with the E assumed as being provided through external services. This is the new commonsense of both technology supply and business use.
Lessons from the Chapter 1 e-commerce laboratory about business targets
So far, the arguments of this chapter hide a major potential risk for new generation e-commerce that has to be recognized up front: competitive differentiation via electronic differentiation too often is not sustainable. Again, it is useful to look at e-commerce as pre-dating the dot coms. The classic e-commerce innovation was the ATM. Today, it is so embedded in the structures of everyday life that it is easy to overlook how risky and radical it was and the many parallels to today’s e-commerce that bring forth the old adage that those who ignore history are doomed to repeat it. When Citibank launched its ATM initiatives, Bank of America was publicly alarmed not because it feared Citi’s success but that its inevitable failure could put the entire U.S. banking industry at risk. Citi in turn saw the opportunity to create and maintain a proprietary competitive advantage. It built its own ATM network so that other banks could not piggyback on its innovation. It made several entrepreneurs very rich as Citibank grew to depend on their technology. Other banks came into the then electronic banking game with their own networks and technology standards. Increasingly, though, smaller players drew on the emerging ATM utilities like Cirrus and Most that offered shared infrastructures. Customers began to expect that they could use their bank’s ATM card at any machine. The end of this more than a decade-long sequence was that the premium item that Citibank created had become a commodity and shared networks the norm and no bank had any competitive advantage. They had added to their own cost base and greatly benefited their customers but even Citibank had to surrender to the inevitable: the ATM card as an interface to any bank worldwide. Many commentators argue that the industry would have been best advised to accept this long-term reality and collaborate from the start to build shared networks and common interfaces.
In the Internet e-commerce era, the laboratory shows many examples of a similar loss of competitive differentiation, added cost base and increased price competition at the expense of the provider but very much to the benefit of the customer. The obvious examples are airline ticketing and car purchases. Historically, airlines, travel agents and car dealers had an information edge over buyers in their knowledge of prices, inventory and range of customer choices. They used this edge to optimize their margins and retain a strong degree of control over their marketplace. Internet e-commerce essentially gives away that edge. The customer can use the airlines’ own Web sites, search engines and online intermediaries to locate the best deals. They often use this information to bypass the very providers who provide them with access to it. Here, the customer gains but the industry in many ways loses, analogous to the evolution of ATMs. Travel agents have been the main victims here; online e-commerce has contributed to a 20% decrease a year in the number of agencies in business as their services are commoditized, as airlines pressure the agents by reducing their commissions and providing financial incentives for passengers to book online and use their own reservation systems, and as intermediaries like Travelocity, Priceline and Expedia erode their prices. In the business market, travel departments are using the new openness of pricing and inventory information to negotiate deals with airlines that cut prices by as much as 80%. In the auto industry, while dealers still maintain control of sales outlets via distinctly non-competitive regulatory restrictions and while such Chapter 1 e-commerce stars as AutoByTel have not been able to move from dot com to dot profit, car buyers routinely use the Web to gather information about prices, options and promotions that they use to provide them with new bargaining power, bypassing the information provider who created that benefit for them.
Economists talk about asymmetry of information to capture the advantage that airlines, car dealers, brokers and distributors have historically used to leverage their own role and profits in industry value chains. Web e-commerce outliers looked like breaking that asymmetry and building customer power that would be rewarded by those customers gravitating to them. Many of these portals went under in the dot com crash. The laboratory experiment lesson is not that they failed but that they failed to build a W3 capability. Here, W3 does not stand for World Wide Web but Win-Win-Win relationships – a win for the provider of an e-commerce service, for its customers and for relevant business partners. Two out of three is not enough to ensure stability. This W3 principle underlies many of the successes of Chapter 1 to the degree that it must be an imperative in the design of new generation e-commerce services. Here are a few examples from the laboratory, again with the implicit statement that it is irrelevant whether or not they refer to a specific dot that thrived or died:
- Customer self-management underlies many of the successes in both consumer and business to business e-commerce. Basically, the art form is to make the provider’s back office administrative expense the customer’s valued front-office and to provide incentives for other online providers to help extend the provider’s range of services and brand. Yahoo, Dell, Fedex, Cisco, Schwab and many airlines and banks are examples here. The economic driver is the cost difference between a customer handling a query or making a transaction at a cost to the provider of a few cents versus the typical $10 for call center response. Self-management shifts out many costs but it works only if customers gain, too, in terms of personalization, special deals and a level of convenience and responsiveness that goes beyond what they can get by picking up the phone and dialing 1-800 something. The scale of the self-management opportunity is indicated by Fedex’s saving of $25 million a month from the 2.4 million daily tracks of packages that previously were handled via its 1-800 phone number. Its CIO comments
“And best of all, customers prefer doing business with us on the
Internet.”
- Many B2B portals did not create a W3 situation. They and the customer gained but dealers, brokers, and other intermediaries lost out or worried that they would lose something somewhere. The auto industry’s ambitious Covisint portal was potentially a big win for the major manufacturers and their largest “Tier 1” suppliers but the smaller players saw in advance that they would be sure losers and opted out.
- The complex issue of consumer trust, privacy and security is really a W3 agenda that was poorly handled by many Chapter 1 e-commerce players. They demanded and used – even misused – personal information mainly to maximize their selling opportunities. Consumers saw this as an intrusion and abandoned the site quickly. Others abandoned personalized Web site offers when they heard about the misuse of cookies, selling of email addresses and uses of software that captures information about customers to profile their behavior across many sites.
- In mobile commerce, the basic W3 challenge is that customers will not pay for services that are just conveniences. In standard Internet e-commerce, the same has long been the case. The old adage about the Internet before it became a business services platform was that “Information wants to be free.” The (mainly strongly anti-business) Internet community used it to share information and ignored issues of security and in many instances of copyright. Information is a universal and common good. That viewpoint underlies what many adults see as immoral – Napster’s enablement of a large-scale industry of downloading music for free (until it was legally forced to end the service) – but that a whole generation sees as perfectly reasonable. Perhaps information does want to be free; certainly, no company has as yet succeeded in building large-scale services where customers are willing to pay for information: online newspapers, search engines, or the many city and restaurant guides that can be accessed from a Web-enabled mobile phone.
- One of the world leaders in Internet and mobile banking, Nordea (which operates in the world’s leading Internet and mobile communications societies in the world, Scandinavia) insists that giving away services has distorted the entire economics of e-commerce. Doing this puts the provider into a lose-lose situation; if the customer values the service they will use it but insist it remain free. It will be very hard to charge for it later. If it begins by charging for the service, it comes up against the growing expectation of free information or the competition from providers who will give it away in order to attract customers. Mobile commerce has been badly blocked by its inability to create a revenue model for data services, beyond SMS.
In each of these and the many other Chapter 1 e-commerce examples, the business model of providers in one way succeeded. They offered something of interest and value to customers. They assumed that the customer would create value for themselves, via sales or commissions from other companies for referrals. Many expected online ads to compensate for their free services. As mentioned earlier, even for the small numbers of branded portals such as AOL and Yahoo that did build a strong, high margin revenue base through ads, that base was highly sensitive to scale. The only major and successful brands in e-commerce information services have from the start accepted that information is free (Google, AskJeeves, Yahoo) or at the very least a free part of a paid subscription (AOL). Only the Wall Street Journal has built a critical mass of paid subscribers for its online newspaper and that is both small in volume and a special case. News is available everywhere via Yahoo, AOL, individual publications and search engines – for free.
W3 may well turn out to be the single key element in any effective business model for new generation e-commerce. The evidence favoring this view is apparent again across the Chapter 1 laboratory: the emergence over time of the concept of value networks. These may be often loose online collaborative arrangements in supply chain management where the technology is used to create flows of information for collaborative forecasting; reports show that this has reduced inventory levels along the supply chain by around 25% for all parties. The complexity and interdependencies of modern business increasingly make collaboration and alliances essential and obviously these will arrive at and maintain a stable equilibrium if and only if they are W3 in nature. It can be argued that many of the dot coms that crashed were a big success – for the customer. Others were a success for the provider and its customer base but a loss for many established intermediaries.
The design and maintenance of W3 value networks is the agenda for new generation e-commerce. Two out of three is just not enough.
Conclusion
The line of reasoning outlined in this chapter provides a way of thinking about new generation e-commerce. It does not point to specific actions or business targets but offers some simple guidelines for creating a clearing in the forest of muddle, claims, headlines, assertions and often intellectual undergrowth and vendor debris that so marks e-commerce discussion. The line of reasoning here is:
- The underlying structural issue for e-commerce is the cost dynamics of the firm, with e-commerce pushing towards heavy investment in basically fixed or at least programmed costs in two areas: technology and customer acquisition.
- Customer acquisition costs can be recovered only through repeat business, the cornerstone of Amazon and Schwab), cross-selling of other services (Yahoo, which “runs” well over 15,000 online stores – for commissions and advertising fees) and strong ongoing relationships (Cisco’s online self-management services for established customers and Dell’s customized Premium Page Web sites for its largest ones.
- Technology fixed costs were a blockage to e-commerce diffusion in the pre-Internet e-commerce era and a massive burden for the fast startup, rapid expansion dot coms. The routinization of the e-commerce Web-based technology infrastructure opens up growing opportunities to shift to technology as a variable cost and source services and skills globally and selectively. This opportunity cannot be missed and is surely the middle-term mainstream for IT in general – the equivalent of the electrical utility industry.
- The commercial challenge for any e-commerce initiative is how to ensure it is W3. This applies whether the business model is focused on the consumer or business market, e-commerce, mobile commerce, clicks, bricks, clicks as part of bricks, portals, retailing, banking, healthcare, manufacturing or distribution.
E-commerce has too often been thought of in revolutionary terms. It is an evolutionary force, in terms of both business and technology. The word evolution is just one letter different from revolution, of course. The firms that focus their e-commerce initiatives to take advantage of the lessons from Chapter 1 on the online economy, especially concern cost structures, will be part of what historians will almost certainly label as the Something Revolution, with many companies out of business because they failed to treat e-commerce as commerce. Evolve or erode. What most stands out in 2003 about e-commerce is how robust in aggregate it has been, however volatile the fortunes of individual companies and most of the dot coms. E-commerce is everyday business, which makes it an everyday business management responsibility.
Boyson, Sandor, Thomas M. Corsi, Martin E. Dresner and Lisa H. Harrington,
Logistics And The Extended Enterprise: Benchmarks And Best Practices For The Manufacturing Professional (New York: Wiley, 1999)
Dennis, A.R., J.F. George, L.M. Jessup, J.F. Nunamaker, and D.R. Vogel,
“Information Technology to Support Electronic Meetings,” MIS Quarterly, Vol 12, Issue 4, December 1988, pp. 591-624
Hammer, Michael and James Champy, Reengineering the Corporation: A Manifesto for Business Revolution (New York: Harper Business, 1993)
Keen, Peter G. W., Shaping the Future: Business Design Through Information Technology (Cambridge: Harvard Business School Press, 1997)
Keen, Peter G. W., The Process Edge: Creating Value Where it Counts (Cambridge: Harvard Business School Press, 1997)
Keen, Peter and Mark McDonald, The eProcess Edge (New York: McGraw-Hill, 2000)
Keen, Peter G. W. and Henk G. Sol, Rehearsing the Future (in manuscript, 2003)
Winograd, Terry and Fernando Flores, Understanding Computers And
Cognition, (Norwood, New Jersey: Ablex, 1986)
“A new economy nightmare: value America rode the Internet up like a rocket and came down like
one”, Washington Post March 5, 2000, p. H1
David Puttruck ad Terry Pearce, Clicks and mortar: passion-driven growth in an Internet
world, San Francisco: Jossey-Bass. 2000
Competing in Chapter 2 of Internet business: Navigating in a new
world, The Eburon Netherlands: TU Delft, 2000
stf pre-Internet costs
cit
Sandor Boyson, Thomas M. Corsi, Martin E. Dresner, And Lisa H. Harrington,
Logistics And The Extended Enterprise: Benchmarks And Best Practices For The Manufacturing
Professional, New York: Wiley, 1999, pp 135-138
Robert D. Hof, Overview: "The “click here”
economy”, business week, june 22, 1998
|
|