Seven e-commerce lessons
By Prof. Dr. Hermann Simon
The internet hype has seen many herald the future business-to-business e-commerce as the cure-all to CEOs' prayers; massive cost savings, shorter sales channels and increased efficiencies to name a few. The reality is that these benefits are more likely to be achieved by improved processes within the organisation than anything else. The facts point to the true direction of the internet as a medium for commerce: business-to-consumer (B2C), with a focus on products that can be delivered digitally.
Guesswork, confusion and nonsense
Don't believe the numbers on e-business - they're an overhyped manipulation of the facts as they stand.
Misleading comparisons: Business-to-business (B2B) is presumed to be much larger than business-to-consumer (B2C). Forrester Research predicts that while the B2B sector is expected to generate € 1,318bn in the year 2004, B2C should bring in 'only' € 232bn. However, this comparison is meaningless. The volume of B2B transactions in each economy is much greater than B2C - my guess is by a factor of between five and ten. For each car a consumer buys representing a value of, say, € 15,000, there is a five or ten fold trading volume of parts, tools, chemicals etc. Nobody knows the exact factor for sure.
Confusion and guessing: The Strategis Group has predicted that the number of mobile internet users in the US will reach 17 million by 2005. IDC's estimates were higher at 40 million, surpassed only by Merrill Lynch's forecast of a whopping 161 million by that year. The estimates for mobile commerce advertising in 2005 were just as conflicting, with Forrester Research estimating $890m; Ovum: $4.2bn and the Yankee Group: $6.1bn. Are they just fooling around? The trouble is, journalists publish these numbers, people believe them, and they stick as if written in stone.
Cause and (exaggerated) effect: Time and again, we hear reports about huge savings, particularly in industrial purchasing. IBM Germany estimates that a 15 per cent saving is realistic for auto parts, 20 per cent for computer components, with purchasing cycles cut from 30 days to a mere three. But beware! How much of this is really due to the internet? In one particular case, an automotive company purchased steel at 15 per cent less than the previous year. They achieved this because they bundled their own and their suppliers' demand and forced the lower prices they paid on the steel producer. All this was enabled through the internet and fully attributed to it.
In reality, the internet can be attributed to 1 per cent, or 2 per cent at most - the rest was a result of pure purchasing power. Of course, there are true savings. But how big are they really? And for which products? Our guess is that the greatest savings over the internet are attained for the least important products, particularly those aimed at the consumer. The myth of the newly discovered supplier is also worth exploring as the Frankfurter Allgemeine Zeitung (FAZ) newspaper printed recently. "Everybody familiar with selling and purchasing processes knows that no purchaser has to search for new vendors. They all beat a path to his door. The purchasing specialist usually knows all relevant suppliers", explained the writer Rainer Stolz.
Preposterous savings: The Industry Standard magazine quotes Jack Welch as saying: "The internet could save the company $10bn over two years. Every day we uncover hundreds of millions of dollars from efficiencies". Where do these billions come from? Are they substantiated? Obviously there appear to be savings and productivity gains, but what we contest is the magnitude of these claims and the lack of substantiation.
Meaningless sales: According to FAZ, mail order companies in Germany reported internet sales of € 2.1bn for 2000. This amounts to 5.3 per cent of total sales - an impressive number. But what does this really mean? Not much! The internet is just another ordering channel for mail-order companies. The difference between telephone or mail sales and internet sales is not really relevant. Mail-order companies are not even sure whether they generate incremental sales from the internet but considering their limited overall growth of 1.7 per cent in 2000, it is unlikely. In a similar vein, Wuerth, the global leader in assembly products with sales of over € 5bn, has had electronic sales in the billions for years (through its automatic online replenishment system ORSYMAT). Today, these sales are made over the internet. At the 2000 e-Business conference in Bonn, it was suggested that many companies may boast about their internet turnover, but in reality all that has changed is the extent of information dissemination. In a similar strain, it was reported that Vauxhall, the British GM subsidiary, sold 1,500 cars (out of a total of 295,000) over the internet during 2000. But what exactly does 'sold over the internet' mean? Did the buyers test-drive the cars before they ordered or not? Did they visit a dealer, or did they just order
online? Nobody knows. One of the authors ordered his last car over the phone. But first, he visited the dealer and told him he would call him with his decision. Was this an electronic purchase? Hardly. The buying process consists of more than just placing the order. Figure 1 illustrates the so-called ”U” problem of the internet during the buying process.
Another point to note: beware of revenues as such. It's easy to generate revenue when you sell at extremely low, subsidised prices and have negative margins. Amazon sells on negative margins. Priceline sold at prices that were heavily subsidised by venture capital. Is it astute pricing to validate high losses by boasting a high turnover? We doubt it! One lesson holds true: It is easy to sell but very difficult to make money with low prices.
Lesson 1: In order to avoid wrong decisions and investments, don't believe the numbers on the New Economy and the internet. Instead, try to understand what's really going on. E-Business has to start all over again.
What 's really unique about the internet?
The internet possesses two outstanding traits:
1. The ability to distribute digital products at close to zero costs to a large number of customers. This applies to both push ('the supplier pushes') and pull ('the customer pulls') information dissemination.
2. The ability to network, i.e. to connect large numbers of people.
In order to understand the implications of these traits, we have to ask ourselves a few simple questions:
· What kind of product do we want to sell (digital or non-digital)?
· How many transactions do we have?
· Who are our customers (in particular, how many do we have)?
· What role does networking play in our business?
Many products and services are either totally or largely non-digital: cars, steel, chemicals, food, hardware, haircare, hospital services, even tourism. On the other hand, some are almost completely digital: music, data, stock prices, schedules, software, banking, insurance.
As well as digital / non-digital products there is a composite alternative that is physical but can be made digital. These products include newspapers, books, entertainment, films, financial services, knowledge management and recruitment tools. This third category holds the greatest potential for the internet.
By combining the product type ('non-digital', 'digital/digitisable') with the number of customers, we get a better understanding of productivity gains through the internet. The highest (perhaps the only high) gains will be achieved for digital/digitisable products that are sold to many customers. The smallest gains have to be expected for non-digital products that are sold to few customers. Interestingly, the much talked about automotive suppliers fall into this latter category.
Amazon.com falls into the upper right quadrant (see Figure 1) - not a very promising position. Why? Most probably as a result of an insubstantial digital segment within their internal value chain. On the other hand, their logistical capability is much larger. Experts estimate that logistics costs are around € 6 per package shipped. Is it realistic to believe that these costs plus a margin can be recovered when you sell a book or CD at € 10? Commenting on amazon.com's $1.4bn loss on revenues of $2.8bn in 2000, an editor from the Wall Street Journal commented: 'In retrospect it seems clear that loading up 10,000 consumer items in a van and trucking them to a central location (known as a 'store') is cheaper than sending 10,000 vans to deliver the same goods individually to 10,000 doorsteps.'
To reiterate, most products and services are not purely digital or non-digital, but a composite of the two. The critical question to ask is: 'How large is the digitisable segment of the product's value chain?" More often than not, this percentage is amazingly low. Take the tourism industry for example, where planes, trains, automobiles, hotels, food and personnel cannot be made digital. Banking, music, films or even a medical report can be experienced virtually, but when a customer rents a car, there is no escaping its physical characteristics.
Both the volume of transactions that take place and the number of customers that transact in the marketplace have a cumulative effect on the savings that can be made from the internet. If there are many transactions for a product, possibly each with a small value and accordingly high transaction costs relative to the total value, there can be substantial savings. Again, the impact depends on the segment of the value chain that can be transferred into a digital format.
Cisco, one of the pioneers in e-business, has shown what is possible on the internet. Figure 3 provides a summary.
Nonetheless Cisco had an inventory backlog thanks to poor forecasting for 2000/2001 (subsequently writing off $2.2bn of inventory in the third fiscal quarter of the year ending July 2001). Sales declined 30 per cent in the same period, the company posted its first loss, and the share price slumped to $13.63 on April 6th compared to a high of $82 thirteen months previously. Based on this evidence the internet is not necessarily contributing towards the firm's strategic problems.
Lesson 2: In order to assess the impact of the internet on a specific business, one must profoundly understand the underlying economics. The internet will not create a revolution in non-digital products. But the potential is high for products that are currently physical but can be digitalised.
Networking
According to most experts, the general rule of thumb is that what you cannot do offline, you cannot do online. Generally, that's true. But there is an exception: networking. Some transactional networks can only be created through the internet.
Figure 4 shows a so-called one-way-hub and two-way-hub. The one-way-hub allows the seller to distribute to many customers. If the product is digital, this distribution structure can be extremely economical over the internet - but keep in mind, the one-way-hub is not a network.
On the other hand, a two-way-hub is a network. ebay, the C2C auction site, serves to explain the economics of a two-way-hub. Assuming there are one million sellers and one million buyers in the system, 106 x 106=1012 contacts are possible, an incredibly high number. If only one in 1000 of these potential contacts results in a transaction, we get one billion transactions. The internet is currently the only tool that can realise a network of this magnitude. In one year, ebay initiated 120 million parcels in the US alone. It's no wonder that ebay is one of the few profitable e-commerce companies and one of the most visited shopping sites, with 24.4 million hits. In contrast, amazon.com was second with 22.6 million.
Networks aren't limited to C2C. Twin-Gears is a company based in Germany that works on establishing a four-dimensional network for bearing and drive products. Figure 5 shows the structure.
The limiting factor within this industry arises from the enormous amount of different items and mechanical parts, which, though small and inexpensive, are critical to a firm's very existence. If a spare is not immediately available, a machine might stand idle for hours or even days. During that time, it's likely that the part is actually floating around somewhere in the system, but nobody knows where. With the Twin-Gears system, the part can be traced immediately and delivered as quickly as possible. With complete coverage, the system could achieve an incredible 10^17 contacts. Of course, Napster, Gnutella, Freenet, Aimster, OpenNap and similar mechanisms take advantage of the same network opportunities. The technology of music sharing, so-called P2P systems (for peer-to-peer), no longer even requires a centralised server, which means there is no one to sue. Services like OpenNap allow every computer worldwide to take part in music and file sharing and are already widely used. It's the simple fact that because digital products can be copied and distributed at no cost, traditional distribution structures have become inefficient.
The potential for both internal and external networks is unlimited. The authors of The Cluetrain Manifesto, a book which discusses the death of business as we know it in the wake of the internet revolution, actually consider networks between customers and employees, and their overlap as the most important aspect of the internet. As they put it, 'there are two conversations going on. One inside the company. One with the market. These two conversations want to talk to each other'. They also see the real attraction of the internet as being 'an altruistic throwback to the prehistoric human fascination with telling tales. In many ways the internet more resembles an ancient bazaar than it fits the business models companies try to impose on it. To its participants, it is primarily a place in which all participants are audience to each other'. Indeed, from mobile phones we know that the most important content is the content produced by the users. According to the GSM Association, the number of SMS's (mobile text messages) sent worldwide in 2001 was about 300 billion! If one assumes an average price of 12.5 cents per SMS, this means an annual revenue of roughly € 36bn for the service providers. In Japan, the wireless i-mode has found huge success, with 30 million subscribers as of January 2002. According to The Economist, 'the fancy features on the new phones that people are most likely to want are those that enhance the experience of communicating with each other. To flourish, telephone companies need to turn their customers into their content.'
Lesson 3: Each company must understand what the networking potential of the internet means for its business, both outside and inside the company.
To B or to C?
What are the implications of these diagnoses? The reality points to a reversal in commonly held opinions - B2C holds more promise than B2B. The future of e-business lies in the distribution of digital products and services to many customers and/or to fewer customers with frequent transactions. This also suggests that the distinction between B and C is misleading. It doesn't matter whether a customer is C or B. What matters is the economies of transactions that take place on the internet. Wuerth, a B2B supplier, has 1.5 million customers, 400,000 in Germany alone, all with frequent transactions of small value. Wuerth's distribution problems are more similar to
those of a home delivery service than to those of an automotive supplier who has only a handful of customers and, often, few transactions.
Auctions will be largely irrelevant in the B2B sector. There, breadth of commodities is much narrower than in the B2C market. Is cement a commodity? Yes - but not the delivery of cement. It comes as no surprise that a recently held large cement auction with a transaction value of € 50m turned out to be a flop. The auction covered about a dozen construction sites with totally different requirements: just-in-time, continuous casting, varying lot sizes, difficult truck access etc. There are few commodities that fit the B2B auction model when one considers getting the product to the right place, at the right time and in the right volume. In an auction for B2B parcel services, the three main contenders simply refused to participate. Only one vendor, who had a dubious reliability image, was willing to bid. As a result, the initiator called the auction off.
Strategic procurement calls for the ever increasing cross-integration of value chains between vendor and customer. This not only precludes auctions, but any kind of easy switching between vendors. One cannot have both integration and the lowest price at all times. One must make a choice.
In summary, the internet will play a limited role in industrial procurement and supply chain management. Market players in areas such as the automotive industry, consumer goods, construction industry, or chemicals will therefore only be able to radically transform their respective industries if they integrate horizontally - something that has little to do with the internet and is also very hard to achieve.
Lesson 4: The greatest potential of the internet lies in being able to do business with large numbers of customers. That means business to consumers and business to small businesses, but not classical B2B. The frequency and size of transactions also plays a role. The impact of the internet on procurement and the supply chain will remain limited. Auctions will be exceptions rather than the rule.
The radical changes
Does this article run contrary to accepted wisdom? Is it defeatist? The former yes, the latter no! Indeed, many radical changes will occur. As mentioned previously, the most radical change will be seen in products that exist in physical form today, yet hold the capability to be transformed into digital format, e.g. newspapers, books, music, film.
Information-intensive, however, is not synonymous with digital. Consulting, investment banking, legal advice are all information-intensive, but only a small part can be digitised (essentially the storage and retrieval of information).
One critical question is whether content can be sold profitably on the internet. There is only one industry today that makes money on the internet and telephone: sex. Will other industries be able to achieve this? The jury is still out, and developments in the wake of Napster's challenge to the music industry will tell. The fact is that consumers have been educated to obtain all kinds of free information on the internet. Re-educating them to pay for this information will be difficult. In addition, once a product is in digital form on the internet, it is highly likely that consumers will be able to multiply and distribute it for free. Peer-to-peer (P2P) is likely to induce a rapid diffusion of attractive content. As a result, some experts predict the end of intellectual property and the collapse of the music and the publishing industries. Nobody can control the distribution of freeware such as OpenNap or other P2P-systems. If OpenNap were used by a large community of users, the music industry would no longer face one centralised opponent, Napster, but would have to sue millions of its customers - an impossible task. The same danger appears imminent for books, movies and any kind of product in digital form. This remains a major challenge for all content providers.
Revenue and pricing models will play a central role for digital businesses. Threatened industries don't seem to be sufficiently aware of and prepared for this challenge. Truly sophisticated pricing is called for: subscription models, time-based pricing, volume-related pricing structures (number of bits), and, most importantly, value pricing. However, in the first phase of Paid Content, pay-per-use pricing models are likely to be more successful than subscriptions, since the users aren’t ready to decide yet which offers provide so much value to them that they are willing to pay a subscription fee. Figure 6 illustrates some success factors of Paid Content.
We agree with Don Tapscott of the Wall Street Journal Europe and Oliver Samwer, founder of Alando and Jamba!, that micropayments will play a decisive role in this new world. Companies should adhere to the advice 'If you can't bill it, kill it'.
Lesson 5: The internet will affect different industries in totally different ways. None of the truly radical changes has yet occurred. Businesses which currently sell physical products that can be digitised will be turned upside down. Revenue and pricing models will become essential for survival.
Time, time, time
The radical changes will take much longer than anticipated. This is due to three factors:
· Time needed to develop devices
· Time needed to develop processes
· Tme needed to change habits.
Do we need special devices to use the internet? Isn't the PC the all purpose-machine? No, it is not! At the breakfast table we are not going to change from our beloved print newspaper to a PC. That said, we don't care about the incredibly huge printing and logistics operations that run every night in order to guarantee that the paper is on our doorstep in the morning. We also don't care about the tons of paper we have to get rid of over the course of the year. A typical Saturday issue of the Frankfurter Allgemeine Zeitung, Germany's leading daily, weighs 1.1 kilograms. With a circulation of roughly 620,000, this adds up to 682 tons of paper in just one day. Sunday newspapers in the US are even heavier. We would definitely prefer a domestic newspaper device in our homes. Every day, we would recharge our e-paper, and out comes the newspaper, most likely on one page - it's convenient, and our desire for a real 'hands-on' experience is satisfied. After all, we are used to having a newspaper in our hands, not on a screen, at the breakfast table. Obviously, it takes time to develop such a device, but research is well advanced in this field. The two main contenders are a joint venture between Xerox and 3M along with the E-Ink Corporation, Cambridge and Mass. Please note that we’re not talking about an online version, which most newspapers have already. The Wall Street Journal started testing electronic delivery on February 7, 2001, at 9 a.m. Each day, they send interested customers a small news update per e-mail. Printed out, it's a handy, short newspaper. This is a first step towards the domestic newspaper machine.
Some other devices are further in their development cycles. At the technological exhibition CeBIT 2001, the first wireless, portable internet pads were presented. Easy to use, they can be carried throughout the house, and at a low usage fee it's financially possible to keep them always switched on. We need similar devices for cars. Volkswagen introduced the first internet-equipped car in the summer of 2001. Obviously, internet access in a car should be voice-controlled since our hands are busy steering the vehicle. We need user-friendly, always-on devices for mobile use. Japan's i-mode is a precursor. It was launched as a consumer product right from the start. We need devices for the remote control of activities in our homes. Extended development times have their parallels in history. It took many decades to develop the numerous electronic appliances that exist today. Only a few decades ago there were no microwave ovens, no CD players and no sophisticated heating systems. Today, we have hundreds of electric motors, controls, and machines in our homes, offices and cars.
New processes are equally critical. Internet security is still an issue for many consumers. Micropayments (payments over the internet) are complicated and still perceived as unsafe, credit card fraud is on the rise and the electronic signature is still in a nascent stage. Standards have to be agreed upon, and legal problems have to be solved. These two bottlenecks have so far stunted innovation in areas such as telemedicine. It will take years, if not decades, to cope with all these challenges. Only then will the full potential of e-business be realised.
Probably the most time-consuming bottleneck derives from consumer habits. Because of this, we speculate that we won't see the full impact of the internet for another 20 to 30 years. A generation has to grow up with a new technology to exploit its potential fully. The development of television provides a fitting analogy. TV today is quite different from what it was 30 years ago. In 1970, TV was a one-way, one-medium broadcast mechanism. Today, it is a multimedia-enhancing, interactive 'master-medium'. In Germany, there are four daily soap operas, which in recent years reached an audience of 12 million viewers every workday - the most popular being Gute Zeiten Schlechte Zeiten (GZSZ). Its ancillary monthly magazine, GZSZ, has a circulation of 500,000. The most popular TV network, RTL, also has the most successful content page, with around 50 million visitors per month. Each week, 130,000 people call in wanting to participate in Who Wants to be a Millionaire? - paying per minute for the privilege. Millions of viewers call reality shows, such as Big Brother, on special telephone numbers that charge them a per-minute fee. This interaction has become a big revenue generator. Some TV channels (e.g. TM3) build their business models on this interactivity. To achieve similar active participation levels on the internet requires the development of new habits. Young people who grow up with the medium will use it differently than older consumers with more passive attitudes. How will the internet be used in the future? We can't even begin to imagine today how it will develop. The time it takes to form new habits in businesses and companies is hardly shorter than in households.
Revolutions are effective in destroying old structures, not in creating new ones. Destroy.com is working. For create.com, evolution is needed. And evolution requires time. These facts affect the speed with which established companies should move. The eminent 'first mover advantage' is likely to be a chimera - it's more important to be right than to be fast. Think first, then act!
Lesson 6: It will be decades before the full impact of the internet is known. The first mover advantage is a chimera. This is not a call to become passive, but to assess the situation realistically before acting.
The internal effects
In many established companies, it may well turn out that the internal effects of the internet are more important than the external ones. As The Cluetrain Manifesto points out, the employee network is one of the two webs of communication essential for a company, the other being the customer network. Our own guess is that in our company with 170 employees and eight offices worldwide, 80 per cent of the e-mails sent are internal. At a recent meeting, Bertelsmann CEO Thomas Middelhoff suggested that his employees could now communicate directly with him through e- mail. This, however, is not a new concept, as they have always had the option of writing him a memo. So what's new? The difference is that he can now write a letter to all employees at zero costs. Actually, that is exactly what the Simon-Kucher Chairman does every Friday at 9 a.m. sharp. One mail to many works; many mails to one doesn't work. Nor does it make sense.
As a communication tool for R&D the internet is invaluable. Knowledge management is obviously another area where the internet can offer terrific advantages. The same applies to human resource management, with its very information-intensive processes. Most of this information is digitisable. Recruiting processes now use internet applications and are rather advanced - decisive tools in the so-called talent war. A lot of the work can be delegated to applicants through sophisticated questionnaires, interactive games, or scoring models.
We are more sceptical about e-learning. The learning process as such still requires human interaction. We don't see hordes of people effectively learning in front of a computer rather than a teacher. Programmable learning, educational videos, learning cassettes, distance learning have had limited success, and the same may be true for e-learning. According to the Harvard Business Review (January 2001), IBM does 30 per cent of its employee training online. Thirty per cent of what exactly? And what does this number really mean? Maybe the complementary nature of personal teaching and e-learning holds some promise.
To make our position clear: Of course the interaction will be an extremely effective tool for the distribution and multiplication of learning material. But this is not what we mean by e-learning. The internet also gives access to events which cannot be attended personally. In a seminar of Handelsblatt held in Frankfurt in June 2001, 131 participants were present, 437 followed the live internet transmission, and about 1,000 more looked into the stored internet file during the following ten days. Thus, the number of viewers in the internet was about ten times higher than the number of people in the lecture room. This is a large increase in reach and efficiency.
Most internal processes have to be radically rethought and reorganised before the hidden internal treasures of the internet can be revealed. This requires a lot of organisational 'de-learning', and, as we all know, de-learning is often more difficult than 'new learning', so ultimately these changes will also take much longer than originally expected. The paperless office is only the superficial metaphor for these changes. Whether paper continues to be used or not is quite irrelevant. What counts is how effective and efficient the processes are.
Lesson 7: In many established firms, the internal impact of the internet could become more important than the external impact. Companies are huge information exchange systems, and that's all the internet is about. To realise the potential, all internal processes have to be reinvented.
There are many other issues which need to be addressed. One is the information stockpile available to customers. With the help of the internet some patients are better informed than their doctors, and some investors know more than their financial advisors. This aspect has been discussed. But information overkill is also a problem.
Issues such as 'stickiness' or loyalty will remain critical. Customer loyalty is not a variable which lends itself easily to direct control, so firms must look back to its origins - customer satisfaction. The book Pricing Opportunities in the Digital Age (2001) demonstrates that the level of customer satisfaction is fundamentally linked to perceived value. As value is more difficult to communicate over the internet than price, a dangerous instance of information asymmetry could be introduced. Understanding consumer behaviour on the internet will be just as critical as it has been in the 'Old Economy'.
How does the internet interact with the sales force? In a study for an insurance company of 2,400 sales agents, we found that only 31 per cent used the internet regularly. The role of the internet in a multi-channel system is an unresolved issue with conflicts amongst channel partners a major problem amongst established firms.
Some authors proclaim the increasing importance of partnerships. Whether they are right or not remains an open issue. We believe that strong companies have a penchant for going it alone. The internet will force competitive strategies to evolve. In some cases, the competitive posturing has to be radically redefined. The eminent strategist Michael Porter suggests applying his traditional model of competitive forces and value chain concept to the emerging internet enabled markets. These concepts are indeed helpful to interpret and understand better the competitive implications of the internet when married with the concept of business and its communication networks.
The Overall Lesson: You don't need an E-strategy. You need a strategy, and you need to know how to employ E-tools for this strategy. The internet is a tool - nothing more, nothing less.
 |