9 INFORMATION PRODUCTS Chapter Introduction Information products - - PDF document

CHAPTER 9: Information Products -Peter O’Grady 1

9

INFORMATION PRODUCTS

Chapter Introduction

Information products are likely to have a significant market in the new few years. Such products include software, data and databases. They are characterized by having, at their core, information that can be digitized as data. Associated with these information products are the allied products that manipulate the data, such as fax machines and the computer appliances discussed in the previous chapter. Information products are diverse in nature, ranging from web brows- ers to data routers and from word processing programs to Zip drives. Since they usually have software as a main constituent, they are nec- essarily complex products (as discussed in Chapter 7). Modularity

2 The Age of Modularity PART III

will therefore be an essential ingredient of information products. An additional factor that compels the use of modularity is that informa- tion products will often be required to be part of an overall system and will be required to link with other portions of the system through standardized interfaces. For example, the Iomega Zip drive must link with the computer hardware (through a parallel port or through an in- ternal drive interface) and the Zip drive software must operate with the computer operating system. The Zip drive is therefore a complex product that forms a module with standardized interfaces to the rest of the system. Information products have four main marketing characteristics that give them a different behavior in the market from other products. The first characteristic is that they have low costs of reproduction, since they rely on digital data. It costs very little, for example, to copy a computer program or a database. A second characteristic is that the value of information products can often increase along with an i n- crease in the number of users, in what is termed ‘the network effect’. Users of e-mail, for example, experience increased benefits as more people use e-mail and there are more people with whom to exchange e-mail. The third characteristic is that users are prone to experience switching costs if they try to change technologies. A Mac user, for example, may have to experience the trials of converting files if they switch to a PC. This means that users can often be locked-in to a par- ticular technology although alternative technologies perform better. The fourth characteristic is that one standard will tend to rise to prominence in each product area, as a phenomenon known as tipping

• ccurs. The result is a ‘winner takes all’ environment where one stan-

dard dominates the others. Information products can also exhibit sudden, explosive growth as they are caught in a spiral of increased usage, leading to an enhanced

• verall value due to the network effect, which in turn leads to i

n- creased usage and so on. An example is the fax machine, which was

• nly used by a few companies as recently as 1982. As more compa-

nies started using faxes, the usefulness of the fax machine grew to the

CHAPTER 9: Information Products -Peter O’Grady 3

point where it became an essential piece of business equipment. By 1989, almost all companies of any significant size had a fax machine. Similarly, internet-based e-mail was used by few companies in the early 1990s but achieved explosive growth starting in 1995, and by 1998 almost all major companies had internet-based e-mail. Standards are essential for information products and there is the ten- dency to migrate to a single standard for each product as tipping oc-

• curs. Where there are competing standards for a product, customers

may be wary of purchasing out of fear that they might be left with a product that uses the non-dominant standard. For example, the 56K modem is undoubtedly useful since it allows communication at a higher speed than 28.8k modems. However, 56K modems were origi- nally developed with two different standards – the ‘K56flex’ standard from Rockwell and the ‘x2’ standard from U.S. Robotics. A fight between these two standards developed, with each company vying to make their standard the dominant one. Customers were reluctant to purchase either of these two products, fearing that they might be left with the loser in the standards fight. Eventually, both Rockwell and U.S. Robotics realized that neither standard would dominate the other and both decided to adhere to the open V.90 standard. Customers were then able to purchase 56K modems that adhered to the V.90 standard, knowing that this would be the dominant standard. The re- sult was that sales of 56K modems tripled.

Chapter Overview

Information products will be an important part of the economy in the future and since these products will tend to be complex, modularity will play a central role. This chapter describes how modularity can be used for these information products. The differences between infor- mation and conventional products are described. The manner in which module integrators and providers with information products will con- tinually form alliances with others in order to strengthen their position is described. Finally, the ceaseless quest for the dominant standard by module integrators and providers is discussed.

4 The Age of Modularity PART III

The Chapter is divided into the following sections: How Are Information Products Different? This section gives an overview of the main differences between information and conventional products. Costly to Develop, Cheap to Reproduce The difference in the costs of reproduction between conventional and information products is described in this section. The issue of pricing information products is also discussed. The Network Effect This section describes the network effect, whereby the usefulness

• f an information product is often dependent on the number of
• ther users of that technology.

Switching Costs and Lock-In This section discusses the phenomenon whereby users can incur costs to change from one technology to another. These costs can act to lock-in a customer to a particular technology, even when alternative technologies are clearly superior. Positive Feedback and Tipping This section describes how the network effect can lead to a spiral

• f rapidly increasing sales for a product. The product can end up

‘tipping’ the market to become the dominant player. Modularity, Standards and Interfaces The issue of standards is crucial to the growth of information products and this section describes how standards are defined. The Future: The Quest for the Dominant Standard In a ‘winner takes all’ environment, such as that for information products, it becomes vital for module integrators and providers to be aligned with the dominant standard. In a changing world, new products will continually be developed with totally new stan-

• dards. Module integrators and providers will therefore be con-

CHAPTER 9: Information Products -Peter O’Grady 5

stantly striving to align with new dominant standards. This sec- tion describes the background to this ceaseless quest for the dominant standard. The Future: Alliances This section describes how module providers and integrators will be constantly forming new alliances to increase the sales of their products.

How Are Information Products Different?

As was discussed in the Introduction, information products differ from conventional products in four main ways:

• Information products are usually characterized by being costly to

develop, but cheap to reproduce.

• They demonstrate a network effect, whereby the worth of the

product increases with the number of users.

• Information products can have substantial switching costs which

lock-in the customers to a particular product by making it expen- sive to switch products.

• Information products exhibit the effects of positive feedback and

tipping, whereby a product can quickly take a dominant position in the market. Let us now look at each of these points in turn to gain a better under- standing of how information products behave in the market.

Costly to Develop, Cheap to Reproduce

The cost of products is made up of two main parts. The first part is the up-front cost of product development, which involves the costs of de- signing and producing the first version of the product. The second cost comes from reproducing the product and includes the costs of raw materials, purchased parts, manufacturing, assembly and distri- bution.

6 The Age of Modularity PART III

For conventional manufactured products, the reproduction cost can be a substantial part of the overall cost of the product. Although cost breakdowns can vary substantially between industry sectors, the final cost of a product is approximately made up of 50% for materials and purchased components, 14% direct labor and 15% indirect costs. The remainder of the costs being for product development. The reproduc- tion costs for conventional products can therefore often be approxi- mately 80% of the overall costs of a product, with the up-front devel-

• pment costs absorbing the remaining 20%. For example, the Ford

Taurus range of cars absorbed many millions of dollars in the devel-

• pment phase. Once the design of the Taurus was developed, the i

n- cremental cost of reproducing the car involved the costs of raw mate- rials and components, the manufacturing operations, testing and so

• n. The cost of producing a copy of the Ford Taurus can therefore be

a substantial portion of the final cost. Information products may also be costly to develop but, by contrast with conventional products, their reproduction costs may be very low. The Encarta Encyclopaedia CD, for instance, cost several millions to develop but copies of the CD can be made for a dollar or two. Simi- larly, the Netscape Communicator web browser software had high development costs, but can be reproduced inexpensively: since it is a product that naturally uses the web, it can be distributed over the Internet at costs even lower than those of reproducing CDs. An i n- formation product, therefore, is costly to develop but cheap to repro- duce, with a total cost structure as shown in Figure 1.

CHAPTER 9: Information Products -Peter O’Grady 7

Low High Quantity Produced Total Cost

Development Costs

Figure 1: Variation of Total Costs with Quantity Produced of an In- formation Product Such a flat cost structure raises a number of issues for information

• products. One of the issues is: What should be the sales price for the

information product? Conventional products are often priced by a simple markup of 10-50% of the manufacturing costs. Such a simple approach does not make much sense for information products, since this would do little to recoup the development costs – the sales price

• f a CD-ROM would then be around $2, since the reproduction cost
• f a CD is a little over $1. As alternatives to a simple mark-up of the

reproduction cost, companies can adopt a variety of strategies in pricing their information products. One strategy is to have prices dic- tated by calculating the value the information product represents to the customer. A real-time stock analysis product might have a high value to a customer and could therefore be valued at several thousand dollars, whereas software for a computer screen saver is unlikely ever to be worth as much to customers. Other strategies for pricing revolve around the network and tipping effects, described later in this chapter.

8 The Age of Modularity PART III

The Network Effect

The usefulness of information products is often dependent on the number of other users of that technology. For example, e-mail is quite useless if there are only a few others that use e-mail. Similarly, fax machines have little to recommend them if there are only a few others with whom to exchange faxes. The value of an information product rises along with the number of other users of the product. This is known as the network effect. As another example, the value of the Word word-processing software lies partly in the capability of the software itself but also in the num- ber of other users of the software. The large number of users of Word means that Word electronic documents can be exchanged between users without the difficulties of translating files. Similarly, Zip drives from Iomega become more useful when the number of other users be- comes larger since there are then more opportunities to exchange Zip disks with other users. In other words, the larger the Word or Zip drive network, the greater the overall value that Word or Zip drives have. The network effect has been formulated in what is sometimes known as Metcalfe’s Law. It was proposed by Bob Metcalfe, who was one of the main developers of the Ethernet. Metcalfe’s Law, as with Moore’s Law described in Chapter 1, is more of a guideline than a scientifi- cally proven law, but it is helpful in describing the usefulness of a

• technology. According to Metcalfe’s Law, if there are n users of a

technology, then the usefulness of that technology is proportional to the number of other users of that technology (n-1 in this case). The total value of the network of the technology is therefore proportional to the usefulness to all users, which is: n(n-1) = n2 – n. If n is large, as it will be for most information products, then n will be small relative to n2 and Metcalfe’s Law becomes:

CHAPTER 9: Information Products -Peter O’Grady 9

The total value of the network of a technology is proportional to n2. According to Metcalfe’s Law, for example, if the number of users of a technology doubles, the value of the technology network rises by four

• times. Similarly, increasing the number of users by a factor of 10, in-

creases the value of the technology network by a factor of 100. The value of a company can be partly based on the network effect. For instance, the value of America OnLine (AOL) rests on the sub- stantial empire that AOL has created by managing access to their

• network. AOL users can see if their friends are on-line and can inter-

act with them using Instant Messaging, where text messages appear in a window. The famous chat capabilities of AOL allow multiple users in a virtual chat room to send messages that all those in the chat room can see. The more subscribers that AOL has, the more useful that these messaging and chat capabilities are. In other words, AOL has a network effect.

Switching Costs and Lock-In

As a technology takes hold it can become harder to switch to another, rival technology. For instant, those that have been Mac users for sev- eral years will find that there are difficulties in moving to a Windows

• computer. Similarly, a company with the design data of their products

in the DXF format will experience costs in moving to another file

• format. In each case, switching costs are incurred when a user moves

from one technology to another. Besides the direct cost of the new technology, there are two main hidden switching costs: Costs of Retraining Users must have the time and training to become familiar with a new technology. For example, the QWERTY keyboard layout (given its name from the first six letters on the top letter row), is known to result in a lower typing speed than some other layouts. However, for a company to change the keyboard layout of their

10 The Age of Modularity PART III

computers would require considerable training and a significant, though perhaps temporary, drop in performance. Companies have therefore not changed from the QWERTY keyboard layout, in spite of its lower typing speed. Converting Data and Software If a new computer system is put in place, data and software from the old system must be converted to the new system. As the huge effort required to convert software for the year 2000 has shown, the costs of conversion can be substantial. In addition, there is the risk that there may be some problems with the translation and important data could be lost or incorrectly converted, with the possibility of considerable disruption. These switching costs lock-in a customer to a particular technology by making it difficult or expensive to move to another technology. Consequently, corporations may continue to use a technology even when it is substantially inferior to newer technologies because of the high cost of switching. Some companies try to lock-in their customers by deliberately making it difficult or expensive for users to change from their product. For example, if a customer moves from AOL, their e-mail is not for- warded to their new e-mail address but is instead bounced back to the

• sender. This inconvenience may be enough to deter some users from

changing to another internet service provider.

Positive Feedback and Tipping

The network effect means that the overall value of a technology i n- creases with the number of users. Often, an increase in the number of users increases the overall value of a product, which makes it more likely that customers will buy that product. One result can be a spiral

• f increased sales as one product rapidly takes a dominant share of the
• market. For example, as mentioned in the Introduction, the fax m

a- chine was something of a rarity in 1982 and so was of limited value to

CHAPTER 9: Information Products -Peter O’Grady 11

• wners, since there were relatively few others that could send or re-

ceive a fax. The fax machine was still of sufficient value that the numbers of users gradually grew, which meant that the overall value

• f the fax ‘network’ grew even faster. With such a large growth in the
• verall value of the ‘fax’ network, customers rushed to buy their own

fax machines, so that by 1989 most companies had at least one fax machine. This is an example of positive feedback, similar to the way in which a piercing screech can be produced by placing a microphone close to a

• speaker. This screech is started by a small sound being picked up by

the microphone, amplified, and then passed through the speaker. This louder sound is then picked up by the microphone and the cycle re- peats itself, with a rapidly increasing volume of sound from the

• speakers. In the case of information products, it is the network effect

that amplifies the value of the product, potentially resulting in a rap- idly growing level of sales. The presence of positive feedback means that a technology can rap- idly become dominant in a market as customers rush to a dominant

• product. This is called a tipping effect, as the market tips across to the

dominant product as shown in Figure 2. The winning technology takes over the market, displacing the losing technologies. An example is that of the Word word-processing software which started in 1991 to displace the then dominant word-processing package, WordPerfect. This displacement was due to the slow response of WordPerfect to develop a Windows version after the success of Windows 3.0.

12 The Age of Modularity PART III

Time Market Share

Tipping Winner Loser 100 0%

Figure 2: The Tipping Effect The classic example of tipping is the battle of videotape standards that took place in the 1980s between Betamax and VHS. Although Be- tamax was arguably better technically, VHS had the larger market share through flexible licensing arrangements. The providers of the videotapes therefore produced many more VHS than Betamax video-

• tapes. Customers then moved quickly over to VHS and the market

tipped away from Betamax. VHS became the winner and Betamax eventually disappeared. The tipping effect creates a ‘winner takes all’ environment where the company with the winning technology is in the happy position of dominating the market. The winner can then expect substantial sales, while switching costs can prevent any rivals from taking over their dominant position. Conversely, the loser is in a poor position with sales and revenues diminishing to low levels. Given this environment, it is not surprising that companies will desperately try to have their technologies become the winners.

CHAPTER 9: Information Products -Peter O’Grady 13

How can corporations help their technology tip over into the domi- nant technology? There are three main ways: First to Market a New Technology. The first to market with a new technology may be able to grab hold of the market before other technologies can be established. The company can then use the effect of switching costs to make it difficult for users to change to any rival technologies that are de-

• veloped. For example, RealNetworks was one of the first compa-

nies to market a technology for audio and video streaming over the Internet using their RealAudio and RealVideo products, later combined into their RealMedia product. They managed to attract many users by distributing the RealMedia player client free over the Internet, while still charging for the RealMedia server prod-

• uct. The network effect generated by the rapid acceptance and

use of RealMedia meant that RealMedia server sites popped up around the world. By the time competitors arrived, RealNetworks had a large number of users, each of which would have to endure switching costs to change to the rival technology. Microsoft tried to break this cycle of positive feedback for RealNetworks by giving away both the client AND server versions of their own product, called NetShow, but has been unable to make significant inroads in the market thus far. Similarly, Intuit has a considerable network effect for their personal finance product Quicken and has been able to use both the network effect and switching costs to resist the efforts of Microsoft with their rival product, Money. Have a Technology that is Substantially Better Establishing a new technology can be difficult but is made easier if the technology is substantially better than existing technolo-

• gies. Andy Grove from Intel has suggested that the new technol-
• gy needs to perform at least ten times better than existing tech-

nologies in order to sweep the market. While the figure of ‘ten times better’ may be approximate, it does give an indication of how much greater the performance should be if the costs of switching are to be overcome. For example, Iomega developed

14 The Age of Modularity PART III

their Zip drive technology, which can store 100 MB of data on a small removable disk. This is substantially more than the 1.44 MB of data that can be stored on a conventional floppy disk, in fact it is almost 70 times greater. With this substantial improve- ment over conventional floppy drives, Iomega was able to estab- lish themselves in the market for data storage. Have High Credibility Much of the network and tipping effects are produced by custom- ers flocking to what they perceive as the dominant technology. Customers perceptions are therefore important in determining which technology will be dominant. Their reception to a new product will be more positive if the technology is produced by:

• A dominant player in an allied technology. For example, a

new networking technology would have a better reception if it were produced by Cisco Systems than by an unknown company.

• A very strong group of allied corporations who support the
• technology. For instance, Sun was able to propel the Java

programming language into prominence by assembling a large number of influential corporations as supporters of the technology. In each case, a dominant corporation can also use expectations management to announce new technologies, even if somewhat prematurely, to dissuade others from entering the market. This approach can only go so far before the announced technologies become viewed as ‘vaporware’.

Modularity, Standards and Interfaces

Because of the inherent complexity of information products, modu- larity will be vital to their success. Since information products will

• ften be required to operate as part of an overall system, as the exam-

ple of the Zip drive in the Introduction illustrated, these modular i n- formation products will link to the rest of the overall system through

CHAPTER 9: Information Products -Peter O’Grady 15

standardized interfaces. The specification of the interfaces is therefore

• f prime important to the development of information products.

Standard interfaces are a necessary part of most technological devel-

• pments and have been developed since the early days of the Indus-

trial Revolution. The use of interchangeable parts by Eli Whitney is an example of the early use of standard interfaces, although these were necessarily ad-hoc in nature. As the process of industrialization has gathered pace, so too has the formulation of standards, ranging from standards of measurements (the metric standard is an example), to that of computer networks (for example, the specification of TCP/IP). Complete and adequate standards allow for interaction be- tween individuals, groups and corporations, since each party can base their operations on the same standard interfaces and avoid the need- less confusion that may otherwise result. As products become more complex and as there is a move to modu- larity, the need for standard interfaces will grow much stronger. For- tunately, customers exhibit a strong desire for standard interfaces, since standards allow them to avoid many of the switching costs de- scribed earlier. The result is, as we have seen, a tipping effect, as there is a mass migration to a particular standard. As in the example of VHS and Betamax, the presence of multiple competing standards is inherently unstable since customers will tend to move toward a single standard, due to the combined phenomena of the network effect and tipping. However, where there is uncertainty about which standard will emerge as the winner, customers may not want to purchase either product, for fear of being stranded with the losing technology. Where there is no clear winner in sight, there is therefore a considerable incentive for module providers to work out a single standard themselves. For example, as mentioned in the Intro- duction, there have been two competing standards for 56K modems,

• ne called ‘K56flex’, developed by Rockwell, and the other called

‘x2’, developed by U.S. Robotics. Each attempted to corner the mar- ket by aggressive sales and marketing but neither was able to ‘tip’ the

16 The Age of Modularity PART III

• market. Customers were unsure which would end up as the dominant

standard and were therefore reluctant to buy either of these modem

• standards. Eventually, in December 1997, both Rockwell and U.S.

Robotics came to an agreement with the International Telecommuni- cations Union (ITU) on a new standard for 56k modems, called V.90. Both Rockwell and U.S. Robotics benefited from this standard since sales of 56k modems increased three-fold. There would appear to be two main mechanisms whereby standards are formulated:

• 1. Standards Organizations. The first mechanism is where a body,

usually international in nature, develops a standard based on a consideration of the multiple factors that are of concern. An ex- ample is the Initial Graphics Exchange Specification (IGES), which was adopted as a standard in 1981 to allow for the e x- change of Computer Aided Design drawings. There are a large number of standards setting bodies around the world. The ITU sets telecommunications standards for products such as modems and faxes. Other formal bodies include the International Standards Organization (ISO) and the National Institute of Science and Technology (NIST). This method of developing standards tends to be extremely slow with frequent delays caused by the delibera- tions of the, usually, many bodies involved. The main advantage

• f this method of developing standards is that a wide variety of

considerations can be brought to bear in the standard and methods are usually developed for maintaining the standard.

• 2. Market-Driven Standards. The second mechanism of develop-

ing a standard may be thought of as a direct result of widespread

• use. As we have seen, the network effect leads to tipping and a

dominant market share for the chosen technology, which then be- comes a market-driven standard. Windows, VHS and the CD standard are examples of this. This mechanism also applies to the interface used by the dominant technology. For instance, if a computer file format is in widespread use, then this can become a

CHAPTER 9: Information Products -Peter O’Grady 17

market-driven standard. An example is the DXF file format for Computer Aided Design files that was used initially by the Auto- CAD computer package. As this package had the main market share of CAD packages on personal computers, the DXF file for- mat became a market-driven standard. The main advantage of de- veloping standards in this manner is that of speed, since such standards can be very quick to emerge. The main disadvantages are that these standards can change very quickly, they can be pro- prietary in nature and there may be no international body that maintains the standard so that it can fragment. In the real world, the setting of standards can involve much jockeying for position and influence by the organizations involved. As an exam- ple, a new product category that is just emerging is that of smart cards which look similar to credit cards but which contain a small computer

• chip. Smart cards have the advantage of being able to hold much more

information, with much greater security, than the magnetic strip of a conventional credit card. Smart cards can therefore be used for a much greater range of tasks than credit cards and they have the po- tential to attract a very large market. For instance, U.S. benefit pay- ments must now be made electronically and smart cards would be an ideal method of distributing payment. Growth in the use of smart cards has been lower than many had expected. One problem had been the absence of a single standard. Without a single standard, consum- ers would need to carry a number of smart cards and this would be a deterrent to the acceptance of smart cards. Visa and MasterCard worked on a standard to allow cards issued by either of the organiza- tions to be able to be read by a single reader. The result was that MasterCard developed the Mondex system and Visa the Visa Cash system, but both of these can be read by a single machine. As other

• rganizations develop their own systems there will be substantial

pressure to make sure that their systems can be read by the same reader as the Visa and MasterCard systems.

18 The Age of Modularity PART III

The Future: The Quest for the Dominant Standard

As we have seen, the highly complex products of the information economy will likely be composed of a number of modules from dif- ferent providers. Providers will then have to make sure that their modules are complementary to the overall system:

“In the information economy, companies selling complementary components … are equally important. When you are selling one component of a system, you can’t compete if you’re not compati- ble with the rest of the system” 1

The future will therefore be marked by a continual search for stan- dards – open or proprietary – to allow different modules to inter-

• perate. Stability may not last long before a new technology is devel-
• ped and another search for standards is started. Module providers

will therefore be on a never-ending quest to align themselves with the dominant standard interfaces. Module integrators will try to make sure that the architecture in their product is readily accepted; the aim being to persuade module provid- ers that this architecture is indeed the dominant standard. As men- tioned earlier, module integrators can help to do this by being first to market with a new technology, having a technology that is substan- tially better than the existing technologies, or by having high credibil-

• ity. Module integrators will continually pursue the aim of making

their product architecture the dominant standard, expanding their ar- chitectures as technologies change. Microsoft, for example, has been very successful with this strategy. Having established the Windows

• perating system as the dominant standard, they have pursued every
• pportunity to expand the standard architecture when it is endangered

by a new development. For example, the rise of 3D capability, pri- marily as a result of computer games, threatened to allow a 3D stan- dard to emerge that could pose a threat to Microsoft’s dominance. For a time, Silicon Graphics, with their OpenGL graphics standard for high-end graphics, and 3Dfx, with their ‘Voodoo’ chip for computer games, seemed to be the major contenders for the dominant standard.

CHAPTER 9: Information Products -Peter O’Grady 19

Microsoft recognized that they would be left behind and quickly de- veloped a standard interface called Direct3D, later combined with sound and other interfaces into a combined set of interfaces called

• DirectX. The size and credibility of Microsoft has been sufficient to

persuade many module providers of 3D graphics cards and game software to shift to the DirectX standard. However, the high-end graphics market has been more hesitant to change from the OpenGL standard and this has forced Microsoft to enter into discussions with Silicon Graphics to determine a single, compromise standard for this market. From the module providers viewpoint, their aim is to align with a dominant standard so as to avoid developing products for the losing

• standard. For example, those companies that developed software for

the CP/M operating system standard were left high and dry when DOS became the dominant standard. Similarly, the rise to dominance

• f Windows meant that companies that had invested in developing

products for the rival OS/2 operating system found that their invest- ment had been wasted. Module providers will therefore often try to keep their options open before committing to a particular standard and must always be ready to make a switch.

The Future: Alliances

The ease with which information products can be reproduced, com- bined with the network effect exhibit by these products, means that companies will try to form alliances with other companies in order to further their own products. Given the fast pace of change in informa- tion products, these alliances may change rapidly. In the future, such shifting alliances will be the usual mode of operation for organiza- tions providing information products. For example, a computer m

• dem usually comes with a CD-ROM, containing the various drivers

necessary for the modem together with some documentation. The large capacity of CD-ROMs means that other software can be added at almost no cost to the modem manufacturer. A pruned version of the WinFax software from Symantec, or a similar product, is often added

20 The Age of Modularity PART III

to the CD-ROM. This helps the modem manufacturer, since the extra software may increase sales of the modem. It also helps Symantec (or the other software provider) in that users will be exposed to their product and may buy the full version if they find the pruned version useful. Other alliances may be longer lasting. For example, Intel and Micro- soft have formed a very successful alliance to enhance and improve what is termed the ‘wintel’ platform. Though it has been shaky at times, Intel has concentrated on the processors, while Microsoft has concentrated on the complementary operating system and applications software, such as the Office suite. Each has an interest in the success

• f the modules provided by the other, since such success will help

their own modules. Improvements in the Intel processors, for exam- ple, will improve the performance of Microsoft software and tempt those using other operating systems to move to Windows, increasing the sales of Microsoft products. Similarly, if Microsoft improve the

• perating system, customers will move to Intel-based computers,

much to the benefit of Intel. The ‘wintel’ alliance is an example of the

• ften uneasy alliances that naturally form in the information econ-
• my.

Key Points

Information products are characterized by having digitized data at their core. These products include software, databases, audio, video and the associated products that manipulate this data. Information products will be an important part of the economy in the future and since these products will tend to be complex, modularity will play an important role. The key points raised include:

• Information products are usually characterized by being costly to

develop, but cheap to reproduce. This produces a flat cost curve where costs do not rise appreciably with the number of informa- tion products produced.

CHAPTER 9: Information Products -Peter O’Grady 21

• Information products demonstrate a network effect, whereby the

worth of the product increases with the number of users.

• Metcalfe’s Law states that the usefulness of a technology is pro-

portional to the number of other users of that technology.

• Information products can have substantial switching costs which

lock-in the customers to a particular product by making it expen- sive to switch products.

• The switching costs are mainly made up of the costs of retraining

and the costs of converting data and software.

• Information products exhibit the effects of positive feedback and

tipping, whereby a product can quickly take a dominant position in the market.

• Companies can help their technology tip over into the dominant

technology by being the first to market a new technology, by having a technology that is substantial better than other technolo- gies, or by having a high credibility.

• Modularity will be a vital part of the complex information prod-

ucts of the future.

• Standards are developed by standards organizations, which tend to

be slow to respond to new developments, or by the market (as market-driven standards), which tend to respond faster but may be proprietary.

• In the future, module integrators and providers will be on a con-

tinual quest for the dominant standard.

• Also in the future, shifting alliances will be the usual mode of op-

eration of organizations providing information products.

References

22 The Age of Modularity PART III

[1] Shapiro C. and Varian H.R., 1999, “Information Rules: A Strategic Guide to the Network Economy”, Harvard Business School Press.