Computing Applications Viewpoint

Innovation from the Edges

How innovation originates from market participants with multiple perspectives about commercial value.
  1. Introduction
  2. At the Outset
  3. Support for a Surprise
  4. Market-based Experiments
  5. Contrasting Experiences
  6. Conclusion
  7. Author
  8. Footnotes
Innovation from the Edges, illustration

Seen from the perspective of a knowledgeable observer circa 1990, the Internet would have appeared to be ill-suited for commercial life. There was no orientation toward market needs. The National Science Foundation (NSF) allowed users to develop applications that were not for sale, and most applications stressed technically advanced functionality, not user-friendliness. The network behind the scenes also did not charge prices, and seemed not to provide experience relevant to organizing a mass market for data services from many providers.

Compare those impressions with what did happen. The NSF privatized the Internet backbone, and not long thereafter the growth and deployment of the commercial Internet changed the way everyone lives, works, and plays. Most astonishing of all, the deployment of the commercial Internet brought about these changes at a fast pace. These characteristics are typically associated with only the most transformative technologies, such as the deployment of electricity or the automobile.

What lessons can be learned from this experience? Something is missing from common understanding—a view about how commercial markets shape which frontier information technology comes to users, and at what price, and in what organizational form. I developed that theme in a recent booka and discuss that further in this Viewpoint. As with any technology, many archetypical economic forces shaped the diffusion of the Internet. For example, the surprising start of the commercial Internet reflected well-known tensions in the transfer of technology out of universities into private hands; the rush to start businesses resembled a gold rush, and it was followed by investment typical for deploying a network good; and the bulk of investment dollars from commercial actors was oriented to adapting technology for market needs, not with invention per se.

What was unique then? It was innovation from the edges, a process that arises when innovation originates from market participants with multiple perspectives about commercial value. Many perspectives brought a diversity of viewpoints to new market opportunities, reflecting a variety of opinions about how best to build businesses to address those opportunities. Many participants came from a wide breadth of organizations, some commercial and some not. Implicit in this framework are multiple origins, that is, no single entity or commercial decision maker coordinated the Internet for an extended time with a single economic interest as motive.

This framework explains how there could be a commercial Internet that relied so heavily on markets, yet lacked an advanced plan for orchestrating the design, or a commercial firm that built and operated the entire system. It can also explain how government policies—both antitrust actions and regulatory decisions—could have encouraged or discouraged this outcome, and ended up being encouraging.

Back to Top

At the Outset

What policies had consequences for market participants? The Federal Communication Commission (FCC) played an important role from the beginning. It discouraged monopoly local telephone providers from interfering with the growth of new data services. Not a single U.S. phone company invested in the commercial Internet, or planned for it, but at crucial early moments they were required to do business with firms who did. This illustrates an important lesson for policy: the U.S. benefited as the location that had a much more robust set of commercial experiments than any other country.

Open organizations also played a key role. The Internet Engineering Task Force (IETF) helped govern improvement to the software protocols of the Internet. Workable standards required effort, discussion, iteration, testing, and support. To be sure, that did not happen by itself. The IETF did start with government support at DARPA and the National Science Foundation. There is a crucial difference between mandates and encouragement, however. Aside from the FCC rules limiting monopolies, the government mandated very little. For example, DARPA and the NSF did not instruct the participants at the IETF on what they needed to do (and, to be sure, none of the participants would have taken kindly to such instruction had policy makers been willing to try). The government helped pay for the beginning, and gave the efforts its blessing, and then stepped out of decision making. The IETF’s efforts were complemented by those of the World Wide Web Consortium (W3C). As is well known, Tim Berners-Lee established the W3C, and helped deploy it widely.

Like any economic activity, the commercial Internet possessed a value chain.

Most important, the IETF and W3C did not hoard the results from their efforts, nor were their lessons restricted to a small group of cognoscenti. Neither the IETF nor the W3C restricted how technology could be used, nor by whom. Instead, they helped to support a wide potential breadth of businesses after privatization. Both organizations also accumulated suggestions from many corners, scaling up their activities to adapt to the growing commercial Internet.

This openness permitted both incremental and radical change, even radical change that otherwise might have encountered roadblocks at private firms. This point is illustrated in my book with several examples from the development of the browser and the Apache webserver. Indeed, these governance norms have become an archetype themselves, and today similar behavior can be found in every open source organization, and related norms arise in many non-for profits, such as Wikipedia.

Back to Top

Support for a Surprise

Like any economic activity, the commercial Internet possessed a value chain. A value chain is a linked set of activities—typically offered by many firms—that together enable firms to sell goods and services to users. The value chain behind the commercial Internet and web did not resemble existing value chains in communications markets, which had been controlled by a small set of dominant players. Open systems played a key role in the evolution of this new value chain.

Who took advantage of the new opportunities? The simple answer starts with "outsiders," who specialized a wide number of different activities. At first, most commercial actions came from outsiders with distinct points of view. As an example, consider dial-up Internet service providers, many of which had been in the business of offering bulletin board services (BBS) before the commercial Internet grew. For the most part, the core of the computing and communications industry had treated BBSs as peripheral players. Eventually thousands of such ISPs provided service for the U.S. landscape.

Another key outsider came from university research. Netscape played a key role as catalyst, and this firm combined the outlook of an outsider with the financial backing of insiders. Programmers from the University of Illinois at Urbana Champaign, who had developed a browser and server software, worked with Jim Clark, an experienced entrepreneur, and his venture-capital backers, Kleiner, Perkins, Caufield, and Byers. This was just the first of many times that some commercial insiders remained open to the new outlook of outsiders and made mutually beneficial deals to develop new businesses.

Back to Top

Market-based Experiments

The latter part of the 1990s showed that the core architectural principles and engineering processes for operating the network could scale to a mass user base. As it turned out, the growth of investment and entrepreneurial entry persisted long after the initial rush diminished. In this case, once again, much of this commercial activity reflected economic archetypes for venture-based businesses.

In this respect, the events described illustrate another crucial lesson: Like other deployments of major technologies, the Internet was not valuable merely because it became available. The invention had to be adapted to many circumstances. Here, again, many firms from outside the core of the communications industry and computing industry perceived many of the opportunities for adaptation. For example, a large and independent ISP industry grew to cover many regions of the country. So too did a large number of independent contractors to provider related services, such as Web-page design and development.

Once the prototypes for the entire system were demonstrated, more commercial firms began to explore innovative applications in areas that employed frontier computing, such as back-office computing, or operations-enhancing enterprise computing in financial transactions, retailing and wholesaling, logistics, and media. The opportunities appeared large to many entrepreneurs, and while a few firms behind dot-com boom got more attention, a large number of quieter firms developed the Internet into something useful.

Established firms had to react to entrepreneurs, as many thought the entrants threatened to take leadership positions. For example, old-line firms such as IBM altered their array of services, and fundamentally altered their commercial focus. This widespread reaction and competition resembled another economic archetype, one of "creative destruction," yielding many new services for users, and extending the deployment of the technology to many new uses. Because the opportunities extended widely, once unleashed, the phenomenon extended across virtually every sector of the economy, and in virtually every urban location in the U.S.

There is nothing inevitable about the commercial development of technology and that also holds for innovation from the edges.

All this impatient investment encouraged a high tolerance for exploratory activity by commercial firms. Part of that tolerance was unsurprising in light of the scale of the new opportunities and the need to adapt. In addition, many entrepreneurs came from computing, where technological races were common, and they did not find it unusual to focus on exploratory activities that stretched the functionality of software. Many such firms employed a common strategy of "getting big fast."

The experiments had a palpable effect on perceptions about the direction of technical change in communications technologies. In 1994, hardly a boardroom in the U.S. considered the Internet a priority, and, yet, very few held that attitude in 1998. Views that had been regarded as outside the mainstream only a short time earlier were taken seriously by some of the stodgiest firms in some of the slowest moving industries. Within a few years a view that would have been considered radical a half-decade earlier had become mainstream.

Back to Top

Contrasting Experiences

There is nothing inevitable about the commercial development of technology, and that also holds for innovation from the edges. As already stressed, government policy can encourage or hinder it. In parallel, established firms may or may not cooperate, and the behavior of large powerful firms matters most for this latter observation. The contrast between the experience at IBM and Microsoft can illustrate both points.

Both firms profited from the rise of the commercial Internet and Web. No government had to compel either firm to sell into growing demand for their core products and services. IBM eventually sold more services as a technological intermediary and consultant. The commercial Internet and Web also increased the total number of PCs sold, and Microsoft benefited handsomely from the rising volume of software sold.

Their similarities ended there. While IBM got over its initial reluctance, Microsoft made only a temporary peace with non-proprietary standards. It was part of a long-term strategy to resist the emergence of alternative platforms, such as Netscape’s. Microsoft also used its existing contracts to encumber many of its business partners. That earned it an antitrust charge from the U.S. Department of Justice.

The ideology of the time also played a role, and not necessarily as a positive force. For a time, Internet exceptionalism became the prevailing view. This philosophy defied the economic archetypes of business history and boldly rejected traditional approaches to building and valuing businesses. Internet entrepreneurs and their financial backers were especially vocal proponents of this view. Arguably, that encouraged innovation, because it fostered risk taking. It also wasted the opportunity, by directing growth in many wasteful projects that otherwise should have been avoided. The voices of experienced entrepreneurs and investors and analysts went unheeded. (Were these lessons heeded? The lack of reform on Wall Street suggests not.)

The dot-com crash helped put an end to the positive public reputation of Internet exceptionalism. Further exploration took on an air of renewal. The creation of value from search engines, for example, built on the interplay between advances made by a couple of precocious graduate students, Larry Page and Sergey Brin. The commercialization of their business—again—combined the views of insiders and outsiders. This happened because markets remained open to a variety of perspectives from participants with many different origins.

The advances made in wireless Internet access contained many of the same aspects. What we today call Wi-Fi drew on ideas from such a combination of insiders and outsiders. Once again, its development illustrated the lessons of government policy that encourages experimentation and does not mandate outcomes. Wi-Fi resulted from the interplay among government policy for spectrum, the design of a standardization committee, and the aspirations of many private firms.

Back to Top


How did innovation from the edges contribute to deploying the Internet? Demand for new value created a situation in which different actors who held distinct opinions about the appropriate actions could test their views. That permitted an extraordinary display of market-oriented experimentation and decentralized decision making. Collectively, it learned how to make the technology value at an extremely fast pace, and across a wide set of applications—browsers, enterprise computing, server software, wireless access, and more.

More to the point, it was far more than ever would have emerged from developing a technology inside a central decision-making process. The old telephone monopoly could not have done the same and would not have. For the same reasons no well-meaning and prescient government planner could have mandated it. That illustrates a key policy lesson: Avoid putting discretion for experiments and market investment solely in the hands of large organizations.

The dot-com crash helped put an end to the positive public reputation of Internet exceptionalism.

There was also a lesson for conventional economics, which stresses that more competitive settings lead to lower prices and induce better service than delivered by a monopoly. While that did arise in the commercial Internet—for example, more ISPs lowered prices for Internet access—the experience in this history highlights an additional observation: Competitive markets fostered a diversity of innovative viewpoints. That, in turn, supported a diversity of experiments with new commercial opportunities that otherwise would never have been conducted inside a laboratory, and on a vast scale, yielding the vibrant Internet we all use today.

As we make policies for the next generation of technologies we must not forget these crucial lessons of the recent past. Neither government alone nor solely market magic enabled good inventions to reach users and improve lives. It took a wise combination, one that avoided monopoly and government mandates, and enabled innovation from the edges.

Back to Top

Back to Top

Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More