Cheaper than printing it out: buy the paperback book.

New Rules for the New Economy

Members Prosper as the Net Prospers

The distinguishing characteristic of networks is that they contain no clear center and no clear outside boundaries. Within a network everything is potentially equidistant from everything else.

Therefore the first thing the network economy reforms is our identity.

The vital distinction between the self (us) and the nonself (them)—once exemplified by the fierce loyalty of the organization man in the industrial era—becomes less meaningful in a network economy. The only "inside" now is whether you are on the network or off.

Individual allegiance moves away from firms and toward networks and network platforms.

Are you Windows or are you Mac?

This shift to network loyalty makes the potential of any network we might want to join a key issue. Is the network waxing or waning? Is the upside potential meager or tremendous? Is the network open or closed?

When given the choice between closed or open systems, consumers show a fierce enthusiasm for open architectures. They choose the open again and again because an open system has more potential upside than a closed one. There are more sources from which to recruit members and more nodes with which to intersect.

Identifying the preferred network to do business in is now a major chore for firms. Because more and more of a firm’s future lies in its networks, firms must evaluate a network’s relative open- and closedness, its circulation, its ability to adapt. Consultant John Hagel says, "A web limits risk. It allows companies to make irreversible investments in the face of technological uncertainty. Companies in a web enjoy expanding sourcing and distribution options, while their fixed investment and skill requirements fall."

As the destiny of firm and web intertwine, the health of the matrix becomes paramount.

Maximizing the value of the net itself soon becomes the number one strategy for a firm. For instance, game companies will devote as much energy to promoting the platform—the tangle of users, game developers, and hardware manufacturers—as they do to their games. For unless their web thrives, they die. This represents a momentous change—a complete shift in orientation. Formerly, employees of a firm focused their attention on two loci: the firm itself and the marketplace.


The prosperity of a firm is directly linked to the prosperity of its network. As the platform or standard it operates on flourishes, so does the firm.

Now there is a third horizon to consider: the network. The network consists of subcontractors, vendors and competitors, emerging standards for exchanges, the technical infrastructure of commerce, and the web of consumers and clients.

Commerce networks can be thought of as ecologies. Economist Brian Arthur states: "Players compete not by locking in a product on their own but by building webs—loose alliances of companies organized around a mini-ecology—that amplify positive feedbacks to the base technology."

During certain phases of growth, feeding the network is as important as feeding the firm. Some firms that already have large market shares (such as Intel, which owns 80% of the PC processor market) channel money, through minority investments, to younger firms whose success will strengthen the market for their products, directly or indirectly. They feed the web because it is good business.

In the network economy a firm’s primary focus shifts from maximizing the firm’s value to maximizing the network’s value.

Not every network demands the same investment. The music CD standard and web of suppliers is well entrenched by now. The new DVD video standard is not. A publishing company issuing music on a CD has to devote less energy to making sure the CD platform flourishes than does a movie company issuing their film on a DVD. The film company must devote substantial resources to ensuring the spread and survival of this emerging platform. They’ll work with the hardware manufacturers, maybe share costs of advertising by seeding the platform logo in their own ads, send reps to technical committees, and cooperate with other film studios in getting the new format accepted. The music company doesn’t need to make as heavy an investment with CDs. But they do need to make investments into new networks if they try to deliver music online—because online delivery is still in its embryonic phase.

Every network technology follows a natural life cycle, roughly broken into three stages:

  • Prestandard
  • Fluid
  • Embedded

A firm’s strategy will depend on what phase a network is in.

The prestandard phase is the most exciting. This period is marked by tremendous innovation, high hopes, and grand ambition. "Aha!" ideas flow readily. Since there are no experts, everyone can compete, and it seems as if everyone does. Easy entry into the field draws myriad players. For instance, when telephone networks began, there were few standards and many contenders. In 1899, there were 2,000 local telephone firms in the American telephone network, many of them running with their own standards of transmission. In a similar vein, in the 1890s, electricity came in a variety of voltages and frequencies. Each local power plant chose one of many competing standards for electrical power. Transportation networks, ditto. As late in the railroad era as 1880, thousands of railway companies did not share a universal gauge.

Two examples of networks in the prestandard stage today are online video and e-money. You have the choice of many competing protocols with equal prospects. With both domains, the uncertainty level is high, but the consequences of being wrong are minimal. Little is locked in, so it’s easy to change.

Networks in the fluid phase have a different dynamic. The plethora of choices in the prestandard phase gradually reduces to two or three. Allegiances are mobile, and drift over time. During this period, networks demand the strongest commitment to their survival. Participants have to feed the web of their choice first, and the narrowing of choices allows substantial investment to spur rapid growth. The effects of plentitude and increasing returns kick in—more breeds more. Feeding the web on any of several standards still produces gains for all participants. Yet it is inevitable that only one standard will ultimately prevail while the other ones fail. The uncertainty level is nearly as high as during the prestandard phase, but the risks for being wrong are greater. Anyone who remembers the demise of 8-track audiotapes will appreciate the perils of this painful stage. Today such networks as digital photographs and desktop operating systems are in this fluid phase: Several well-established standards vie for ultimate dominance. Choose wisely!

The final stage in the life cycle of networks is the embedded phase, where one standard is so widely accepted that it becomes embedded in the fabric of the technology and is thereafter nearly impossible to dislodge—at least as long as the network exists. Regular 110-volt AC power is well embedded at this point (although, as the power grid becomes global, there could be some surprises). ASCII text is likewise deeply embedded—at least for phonetic languages. Some of the conventions of voice dial tone are so ubiquitous worldwide as to be permanent.

In any phase of innovation—prestandard, fluid, or embedded—standards are valuable because they hasten innovation. Agreements are constraints on uncertainty. The constraints of a standard solidify one pathway out of many, allowing further innovation and evolution to accelerate along that stable route. So central is the need to cultivate certainty that organizations must make the common standard their first allegiance. As standards are established, growth takes off.

For maximum prosperity, feed the web first.

Arriving at standards is often easier said than done. Standard-making is a torturous, bickering process every time. And the end result is universally condemned—since it is the child of compromise. But for a standard to be effective, its adoption must be voluntary. There must be room to dissent by pursuing alternative standards at any time.

Standards play an increasingly vital role in the new economy. In the industrial age, relatively few products demanded standards. You didn’t need a consensual network to make a chair and table. If you obeyed some basic ergonomic conventions—make table height 30 inches—you were on your way. Those industrial products that operated in networks—such as the electrical or transportation networks—demanded sophisticated standard-making. Anything plugged into the electrical grid had to be standard. Automobiles manufactured by separate factories shared standards on such things as axle width, fuel mixtures, placement of turn signals, not to mention the many standards of road construction and signage.

All information and communication products and services demand extensive consensus. Participants at both ends of any conversation have to understand each other’s language. Multiply one conversation by a billion, factor in a thousand different media choices, and then start to count three-way, four-way, n-way conversations, and the amount of consensus-setting skyrockets.

In the network economy, ever-less energy is needed to complete a single transaction, but ever-more effort is needed to agree on what pattern the transaction should follow.

Thus "feeding the web first" increases in necessity. Businesses can expect to devote great intellectual capital on formulating, negotiating, deciding, forecasting, and adhering to emerging standards. The question "Which platform do we back?" will not be confined to PCs. It will be asked in regard to calendars, cars, accounting principles, and even currencies.

As more of the economy migrates to intangibles, more of the economy will require standards.

But consumers will groan under the load of decisions. There is a yin-yang tradeoff in the new economy. The yin, or positive side, is that consumers keep most of the gains in productivity that are earned by technology. Competition is so severe, and transactions so "friction-free," that most of each cycle’s betterment goes not to corporate profits but to consumers in the form of cheaper prices and higher quality.

The yang, or downside, is that consumers have a never-ending onslaught of decisions to make about what to buy, what standard to join, when to upgrade or switch, and whether backward compatibility is more important than superior performance. The fatigue of sorting out options and allegiances, or recovering from them, is underappreciated at the moment, but will mount. The joy of the new economy is that the next version is almost free; the bane is that no one wants the hassle of upgrading to it, even if you pay them to do it.

The fatigue will only worsen. The net is a possibility factory, churning out novel opportunities by the screenful. Unharnessed, this explosion can drown the unprepared. Standardizing choices helps tame the debilitating abundance of competing possibilities. This is why the most popular sites on the web today are meta-sites that sort the abundance and point you to the best.

Since the network economy is so new, we as a society have paid little attention to how standards are created and how they grow. But we should notice, because once implemented, a successful standard tends to remain forever. And standards themselves shape behavior.

I was associated with the genesis of the Well, one of the first public computer conferencing systems to be plugged into the internet. The Well was conceived and built by others, but as director of the poor nonprofit that owned it, and as one of the first participants to join when it opened, I was involved in creating its policies. It became clear almost from day one that the technical specifications of the software that the Well used directly shaped the kind of community growing within it. Other models of conferencing software used elsewhere produced different kinds of communities. The Well’s software—as implemented by the Well—encouraged linear conversations and community memory; it discouraged anonymity, but encouraged responsibility for words and topics; it permitted limited forms of dissent and retraction, and it allowed users to invent their own tools. It did all this primarily by means of Unix code—by the software standards set up within the Well—rather than by posted rules. The community it shaped was distinctive and long-lived. In fact the community, with all its quirks, is still going, even though the software that runs it has evolved into a web browser interface. The behavior-changing standards remain. The power to mold a community by code rather than regulation was eventually articulated by Well users into a serviceable maxim: Peace through tools, not rules.

The internet and the web also contain toolish standards that invisibly shape our behavior. We have ideas about ownership, about accessibility, about privacy, and about identity that are all shaped by the code of HTML and TCP/IP, among others. Currently only a small portion of our lives flow through these webs, but as cyberspace subsumes televisionspace and phonespace and much of retailspace, the influence of standards upon social behavior will grow.

Eventually technical standards will become as important as laws.

Laws are codified social standards; but in the future, codified technical standards will be just as important as laws. Harvard Law professor Lawrence Lessig says, "Law is becoming irrelevant. The real locus of regulation is going to be (computer) code." As networks mature, and make the transition from ad hoc prestandard free-for-alls to fluid hot spots of innovation, and then into full-fledged systems with deeply embedded standards, standards increasingly ossify into something like laws.

Standards also harden with age. They become resistant to change and they descend into hardware. Their code gets wired into the backs of chips, and as the chips spread, the standard infiltrates ever more deeply.

An elaborate process of legal overview monitors and analyzes our lawmaking. So far we have little of the sort for our standard-making, although these agencies, such as the ITU (International Telecom Union) will soon be as influential as courts. Standards are not just about technology. They are about soft and fuzzy things such as options and relationships and trust. They are social instruments. They create social territory.

A network is like a country in that it is a web of relationships regulated by standards. In a country citizens pay taxes and adhere to laws for the benefit of all. In a network, netizens feed the web first for the benefit of all. The network economy is a meta-country. Its web of relationships differ from those of a country in three ways:

  • No geographical or temporal boundaries exist—relations flow ceaselessly 24 by 7 by 365.
  • Relations in the network economy are more tightly coupled, more intense, more persistent, more diverse, and more intimate in many ways than most of those in a country.
  • Multiple overlapping networks exist, with multiple overlapping allegiances.

These hyperconnections can either strengthen or weaken traditional relationships. The extremely personal, highly trust-bound relations in a family stand to be strengthened, while the diffuse and nearly contractual relations in a nation-state are liable to weaken. Yet, as Peter Drucker points out, "The nation-state is not going to wither away. It may remain the most powerful political organ around for a long time to come, but it will no longer be the indispensable one." In its stead we’ll rely on nongovernmental agencies such as the Red Cross, ACLU, HMOs, insurance giants, the net and the web, and UN-like entities. These parapolitical organizations will supplement the embedded nation-state. They will be the indispensable networks we care about.

In both country and network, the surest route to raising one’s own prosperity is raising the system’s prosperity. The one clear effect of the industrial age is that the prosperity individuals achieve is more closely related to their nation’s prosperity than to their own efforts. Lester Thurow, an MIT economist, has pointed out that enabling the lowest paid to earn more is the best way to raise wages for the highest paid—the theory being that a rising tide lifts all boats. The network economy will only amplify this.

To raise your product, lift the networks it ties into. To raise your company, lift the standards it supports. To raise your country, increase the connections (in quality and quantity) that allow others to prosper.

To prosper, feed the web first.

The web is underfed right now. It is small compared to the rest of the world. In 1998 the internet boasted of an estimated 120 million people with access. But that means only 2% of human adults have a direct line to the online network.

But the net is growing exponentially fast. If current rates continue, by early in the new century, 1 billion people will have internet access, 75% of adults will access to some kind of phone, and, according to Nicholas Negroponte, there will be 10 billion electronic objects connected together online. Every year the net engulfs more of the world.

The net is moving irreversibly to include everything of the world.

As the net takes over, many observers have noted the gradual displacement in our economy of materials by information. Automobiles weigh less than they once did and yet perform better. Industrial materials have been replaced by nearly weightless high-tech know-how in the form of plastics and composite fiber materials. Stationary objects are gaining information and losing mass, too. Because of improved materials, high-tech construction methods, and smarter office equipment, new buildings today weigh less than comparable ones from the 1950s. So it isn’t only your radio that is shrinking, the entire economy is losing weight too.


Even industrial objects like atuomobiles follow new rules. An automobile's average weight is dropping and will continue to drop as information replaces its mass.

Even when mass is conserved, information increases. An average piece of steel manufactured in 1998 was vastly different from an average piece of steel made in 1950. Both pieces weighed approximately the same, but the one made recently is far superior in performance because of the amount of design, research, and knowledge that went into its creation. Its superior value is not due to extra atoms, but to extra information.

The wholesale migration from mass to bits began with the arrival of computer chips. This subtle disembodiment was first viewed as a unique dynamic of the high-tech corridors of Silicon Valley. Software was so strange—part body, part spirit—that nobody was surprised when the computer industry itself behaved strangely. The principles of the net, such as increasing returns, were seen as special cases, anomalies within the larger "real" economy of steel, oil, automobiles, and farms. What did such weirdness have to do with, say, making cars, or selling lettuce? At first, nothing. But by now every industry (shoe retail, glass manufacturing, hamburgers) has an information component, and that component is increasing. There is not a single company of consequence that does not use computers and communication technology. All U.S. companies (low as well as high-tech) together spent $212 billion on information technology in 1996. Often, the digital component of the firm, say the IT or MIS department, or the wizards running the technology, will be the first to feel the influence of the new rules and network dynamics. Consultants Larry Downes and Chunka Mui say, "Even though the primary technology of many industries may not be in transition ... every industry is going through a revolution in its information technology." As more of a company "goes online" nerd ideas begin to seep into the whole organization, reshaping the firm’s understanding of what it is doing. Over time, more and more employees will chase the opportunities that intensive information and communication networks bring.

New network technology and globalization accelerates the disembodiment of goods and services. The new dynamics of information will gradually supersede the old dynamics of industrialization until network behavior becomes the entire economy.

Bit by bit, the logic of the network will overtake every atom we deal with.

The logic of the network will spread from its base in silicon chips, to infiltrate steel, plywood, chemical dyes, and potato chips. All manufacturing, whether seeded with silicon wafers or not, will respond to network principles.

Consider oil—the quintessential atom-based resource. The classical theory of diminishing returns was practically invented to explain the oil industry. Easy oil is extracted cheaply at first; then at a certain point the expense of extraction doesn’t justify the cost unless the price goes up. But by now the oil industry is so invaded by chip technology that it is beginning to obey the laws of the new economy. Sophisticated 3D viewing software allows geologists to map oil-yielding layers to within a few meters; computer-guided flexible drills can burrow sideways with precision, reaching small pockets of oil. Superior pumps extract more oil with less energy and maintenance. Diminishing returns are halted. The oil flows steadily at steady prices, as the oil industry slides into the new economy.

And what could be more industrial-age than automobiles? Yet, chips and networks can take the industrial age out of cars, too. Most of the energy a car consumes is used to move the car itself, not the passenger. So if the car’s body and engine can be diminished in size, less power is needed to move the car, meaning the engine can be made yet smaller. A smaller engine requires a yet smaller engine, and so on down the slide of compounded value that microprocessors followed. The car’s body can be reduced substantially using smart materials—stuff that requires increasing knowledge to invent and make—which in turn means a smaller, more efficient engine can power it.

Detroit and Japan have designed cars that weigh only 500 kilograms. Built out of ultra-lightweight composite fiber material, these prototypes are powered by high-tech hybrid engine motors. They reduce the mass of radiator, axle, and driveshaft by substituting networked chips. They insert chips to let the car self-diagnose its performance, in real time. They put chips in brakes, making them less likely to skid. They put microprocessors in the dashboard to ease navigation and optimize fuel use. They use hydrogen fuel cells that do not pollute, and electric motors with low noise pollution. And just as embedding chips in brakes made them better, these lightweight cars will be wired with network intelligence to make them safer: A crash will inflate intelligent multiple air bags—think "smart bubblepak."

The accumulated effect of this substitution of knowledge for material in automobiles is what energy visionary Amory Lovins, director of the Rocky Mountain Institute, calls a hypercar: an automobile that will be safer than today’s car, yet can cross the continental United States on one tank of hydrogen fuel.

Already, the typical car boasts more computing power than your typical desktop PC. Already the electronics in a car cost more ($728) than the steel in the car ($675). But what the hypercar promises, says Lovins, is a car remade by silicon. A hypercar can be viewed as step toward a vehicle that is (and behaves like) a solid state module. A car becomes not wheels with chips, but a chip with wheels. And this chip with wheels will drive on a road system increasingly wired as a decentralized electronic network obeying the network economy’s laws as well.

Once we visualize cars as chips with wheels, it’s easier to imagine airplanes as chips with wings, farms as chips with soil, houses as chips with inhabitants. Yes, they will have mass, but that mass will be subjugated by the overwhelming amount of knowledge and information flowing through it. In economic terms, these objects will behave as if they had no mass at all. In that way, they migrate to the network economy.

Because information trumps mass, all commerce migrates to the network economy.

MIT Media Lab director Nicholas Negroponte guesstimates that the online economy will have reached $1 trillion by 2000. Most tenured economists think that figure is terribly optimistic. But actually that optimistic figure is terribly underestimated. It doesn’t anticipate the scale on which the economic world will move on to the internet as the network economy infiltrates cars and traffic and steel and corn. Even if all cars aren’t sold online immediately, the way cars are designed, manufactured, built, and operated will depend on network logic and chip power.

The current concern about the size of the online market will have diminishing relevance, because all commerce is jumping on to the internet. The distinctions between the network economy and the industrial economy will likewise blur and fade, as all economic activity is touched in some way by network rules. The key distinction remaining will be between the animated versus the inert.

The realm of the inert encompasses any object that is divorced from its economic information. A head of lettuce today for instance does not contain any financial information beyond a price sticker. Once applied, that price is fixed, too. It doesn’t change unless a human changes it. The economic consequences of lettuce sales elsewhere, or a change in the general global economy do not affect the head of lettuce itself. Instead, lettuce-related information flows through wholly separate channels—news programs or business newsletters—that are divorced from the lettuce itself. The lettuce is economically inert.

The realm of the animated is different. It’s vastly interconnected. In this coming world a head of lettuce carries its own identity and price, displayed perhaps on an LED slab nearby, or on a disposable chip attached to its stem. The price changes as the lettuce ages, as lettuce down the street is discounted, as the weather in California changes, as the dollar surges in relation to the Mexican peso. Traders back in supermarket headquarters manage the "yield" of lettuce prices using the same algorithms that airlines use to maximize their profits from airline seats. (An unsold seat on a 747 is as perishable as an unsold head of lettuce.) In relation to the net, the lettuce is animated. It is dynamic, adaptive, and interacting with events. A river of money and information flows through it. And if money and information flow through something, then it’s part of the network economy.
The progression by which the old economy migrates toward the new follows a relentless logic:

  • Increasing numbers of inert objects are animated by information networks.
  • Once the inert is touched by a network, it obeys the rules of information.
  • Networks don’t retreat; they tend to multiply into new territories.
  • Eventually all objects and transactions will run by network logic.

One is tempted to add "resistance is futile." The overwhelming long-term trend toward universal connection may seem Borg-like, as if all things will lose their identity and become part of one large mindless swarm. Two things should be made clear: 1) constant, ubiquitous connections do not per se eliminate individuality; and 2) by "all" I mean an ongoing trend that approaches an asymptote, not a finality.

One might say that industrialization eradicated hand-crafted production to the point where all objects are machine-made. That is true by and large, and it accurately describes the destination of a trend. But the trend has a few notable exceptions. In an era of objects made completely by machines, hand-made items are a scarcity and thus command very high prices. A few—but only a few—shrewd artisans and entrepreneurs can make a living crafting items by hand, items such as bicycles, furniture, guitars, that would ordinarily be stamped out in a factory. Resistance is marginal, but profitable.

The same will be true in the networking of the economy. Resistance will not be futile. In a world of ubiquitous connection, where everything is connected to everything else, scarce will be the person not connected at all, or the company not pushing ideas and intangibles. If these mavericks are able to interface with the networked economy without losing their distinctivness or value, then they will be sought out, and their products priced high. One can imagine a successful idea-artist in the year 2005 who does no email, no phone, no videoconferences, no VR, no books, and who does not travel. The only way to get her fabulous ideas is in person, face-to-face at her hideout, live. The fact that she is booked 8 months in advance only adds to her reputation.

MIT economist Paul Krugman has an alternative vision of how information technology will invert the expected order. He writes: "The time may come when most tax lawyers are replaced by expert systems software, but human beings are still needed—and well paid—for such truly difficult occupations as gardening, house cleaning, and the thousands of other services that will receive an ever-growing share of our expenditure as mere consumer goods become steadily cheaper." Actually we don’t need to wait for the future. Recently I had to hire two different freelancers. One sat in her office moving symbols around. She transcribes tape-recorded interviews and charges $25 per hour. The other is a guy who works out of his home repairing greasy kitchen appliances. He charges $50 per hour, and as far as I could tell had more business of the two. Krugman’s argument is that these "manual crafts" (as they are bound to be labeled when so high-priced) will level the salary discrepancies that now exist between high tech and low tech occupations.

My argument is that great gardeners will be high-priced not only because they are scarce and exceptions, but also because they, like everyone else, will be using technology to eliminate as much of the tedious repetitive work as possible, leaving them time to do what humans are so good at: working with the irregular and unexpected.

At the dawn of the industrial age it would have been difficult to imagine how such quintessential agrarian jobs as farming, husbandry, and forestry could become so industrialized. But that is what happened. Not just agrarian work, but just about every imaginable occupation of that period—especially menial labor—was intensely affected by industrialization. The trend was steady: The entire economy eventually became subjected to the machine.

The full-scale trend toward the network economy is equally hard to imagine, but its progression is steady. It follows a predictable pattern. The first jobs to be absorbed by the network economy are new jobs that could only exist in the new world: code hackers, cool hunters, webmasters, and Wall Street quants. Next to succumb are occupations with old goals that can be accomplished faster or better with new tools: real estate brokers, scientists, insurance actuaries, wholesalers, and anyone else who sits at a desk. Finally, the network economy engulfs all the unlikely rest—the butchers, bakers, and candlestick makers—until the entire economy is suffused by networked knowledge.

The three great currents of the network economy: vast globalization, steady dematerialization into knowledge, and deep, ubiquitous networking—these three tides are washing over all shores. Their encroachment is steady, and self-reinforcing. Their combined effect can be rendered simply: The net wins.


Maximize the value of the network. Feed the web first. Networks are nurtured by making it as easy as possible to participate. The more diverse the players in your network—competitors, customers, associations, and critics—the better. Becoming a member should be a breeze. You want to know who your customers are, but you don’t want to make it hard for them to get to you (IDs, yes; passwords, no). You want to make it easy for your competitors to join too (all their customers could potentially be yours as well). Be open to the power of network effects: Relationships are more powerful than technical quality. Especially beware of the "not-invented-here" syndrome. The surest sign of a great network player is its willingness to let go of its own standard (especially if it is "superior") and adopt someone’s else’s to leverage the network’s effect.

Seek the highest common denominator. Because of the laws of plentitude and increasing returns, the most valuable innovations are not the ones with the highest performance, but the ones with the highest performance on the widest basis—the "highest per widest." Feeding the web first means ignoring state-of-the-art advances, and choosing instead the highest common denominator—the highest quality that is widely accepted. One practical reason to pick the highest-per-widest techniques and technologies is because complex technologies require passionate and informed users who can share experience and context, and you want the maximum dispersion of usage that doesn’t sacrifice quality.

Don’t invest in Esperanto. No matter how superior another way of doing something is, it can’t displace an embedded standard—like English. Avoid any scheme that requires the purchase of brand new protocols when usable ones are widely adopted.

Apply an embedded standard in a new territory. Is there a way to accomplish what you want using existing standards and existing webs in a different context? Inventing a novel standard for an existing network is quixotic. But some of the greatest success stories in current times are about firms that master one network and then use its embedded standards to exploit an established network in need of improvement. This process is called "interfection." The present revolution in telephony is all about zealous internet firms that are interfecting the old Bell-head world of moving voices with newly established protocols for moving data on the internet (known as internet protocols, or IP). The huge increasing returns that spin off the internet give them a great advantage. Indeed, one telephony standard after another is falling before the relentless march of IP. Likewise, aggressive companies are leveraging the established desktop standard of Windows NT—with all its plentitude effects—to interfect new domains such as telephone switching gear. Even the huge cable TV networks have something to offer. The emerging standards for video transmission, such as MPEG, are trying to migrate onto the internet. In choosing which standard to back, consider dominant standards outside your current network that could interfect your own turf.

Animate it. As the network economy unfolds, more firms will begin to ask themselves this question: How do we put what we do into the logic of networks? How do we prepare a product to behave with network effects? How do we "netize" our product or service? (The answer is not "put it on a web site.") Architects, for instance, generate huge volumes of data. How can they be standardized? How can the data about a physical object (say a door) flow through or with that object? What are the fewest functions we can add to glass windows to incorporate them into networks? What steps can a contractor take to allow the networked flow of information from any architect to any contractor to any builder to any client? How do we increase the number of networks our service embraces?

Side with the net. Imagine that in 1960 an elf let you in on a secret: For the next 50 years computers would shrink drastically and cheapen yearly on a predictable basis. Subsequently, whenever you needed to make a technological decision, if you had counted on the smaller and cheaper, you would have always been right. Indeed you could have performed financial miracles knowing little more than this rule. Here is today’s secret: In the coming 50 years, the net will expand and thicken yearly on a predictable basis—its value growing exponentially as it embraces more members, and its costs of transactions drop toward zero. Whenever you need to make a technological decision, if you err on the side of choosing the more connected, the more open system, the more widely linked standard, you will always be right.

Employ Evangelists. Economic webs are not alliances. There are often few financial ties among members of a web. An effective way of establishing standards and coordinating development is through evangelists. These are not salespeople, nor executives. Their job is simply to extend the web, to identify others with common interests and then assist in bringing them together. In the early days when Apple was a cocreator of the emerging PC web, it successfully employed evangelists to find third-party vendors to make plug-in boards, or to develop software for their machines. Go and do likewise.