CHAPTER 5

Platforms

Given that one, if not the, byword of today's technological landscape is “connectedness,” much attention has been focused on how systems of networked or related devices evolve. Whether the connection is obvious, as in mobile phones or Facebook friends, or less direct, as in unrelated people who all use Windows and Microsoft Office, the notion of platforms is central to any understanding of this world. Winning in a platform market can be a multibillion-dollar outcome but can involve speculative investments of the same scale. Such household names as Apple, Google, Intel, Nokia, and Sony need to be understood at least in part through the platform lens.

What do we mean by the term “platform”? A few definitions are in order. Platforms are foundational technologies, building blocks upon which an entire industry might develop. Michael Cusumano of MIT and his then graduate student Annabelle Gawer were plainspoken in their essential study of the topic: A platform is “an evolving system made of interdependent pieces that can each be innovated upon.”¹ That's concise but dense, and worth playing out: Platforms have to evolve, which presents challenging issues around innovation, return on investment, and other strategic decisions. Platforms are the product of multiple actors: Even Microsoft, a dominant company if ever there was one, relied on both a powerful network of software developers to build applications on the foundation technology and a myriad of hardware companies. Platforms are systems, not merely products: Their interdependent aspects can create prisoner's dilemma and other game theory scenarios, as when developers must choose where to focus their innovation and programming efforts. For these individuals and companies, betting on Microsoft in 1990 was a good move, as was aligning with the Apple App Store in 2008.

Harvard's Thomas Eisenmann and his colleagues Geoffrey Parker and Marshall Van Alstyne focus on the networks that rise up in the presence of a platform; these scholars account for the actors as much as the technologies:

Most of these aspects are straightforward, but for now, the role of rules should be highlighted. Whether in the form of government regulations, technical standards, or copyright or patent protection, these rules can become exceedingly complicated, expensive to create and enforce, and difficult to time with regard to the market. The right rule set can accelerate time to broad adoption; the wrong set can stall innovation in committees, litigation, or market uncertainty. Notable failed platforms include the 3DO video game, Super Audio Compact Disc (SACD) music format, and IBM's OS/2 computer operating system.

Recently, a strand of scholarship has been focusing on ways that a platform can be either one-sided or two-sided.³ In the former case, there are sellers and there are purchaser/users: In wireline telephony, the Bell company or companies were on one side and “subscribers” were on the other. Innovation was minimal, and the system, while closed, was exceptionally profitable and predictable. AT&T was a monopoly provider, from the handsets all the way through the infrastructure and voice services. By contrast, we now see examples of one-sided platforms with multiple firms, beginning with mobile telephony: Person A with a Samsung phone on T-Mobile can seamlessly call person B who has a Motorola phone operating on the Sprint network. Standards and protocols make such behavior possible.

Two-sided markets are exemplified by a credit card company: Before consumers will carry the card, they want to know that it will be accepted by many merchants. Merchants, for their part, want to know that enough consumers will use the card to justify the expenses of vendor adoption. Getting both sides of the platform to invest, especially in the early stages, can be challenging but extremely advantageous once the “flywheel effect” kicks in: One side of the interaction often subsidizes the other. More recently, Apple's nurturing of application developers on one side and end users on the other has worked extremely well: A large app selection enhances the iPhone's market appeal just as large user markets attract more developers. Shopping malls, health maintenance organizations, and video game console manufacturers also operate in two-sided markets.⁴

A wide range of strategic decisions is involved in platform success, none more important than opening or closing the platform. (This may or may not involve opening the source code: Linux is both open source and an open platform. The terms can be confusing.) eBay is a closed platform in the sense that user identities and reputational profiles work only within the auction site. For several years in the United States, Apple's iPhone worked only on AT&T's network. Apple imposes standards of content appropriateness and performance on third-party apps that Google's Android platform does not. Every DVD player's price includes royalties to the platform's creators, and all players are strictly defined in their features and performance. Closed platforms benefit from interoperability at the expense of innovation; for open platforms, the reverse is usually the case. Android users might need to run antivirus software on their smartphones but do not face the limits Apple places on the iPhone ecosystem.

In contrast, any 120-volt electrical appliance should work in any U.S. state. ISO* standards govern shipping containers, which will fit on any ship in any port. Closed platforms such as video games keep all profit within a small number of partner companies, while open platforms (such as 802.11 or Wi-Fi, not to mention the Internet itself) can grow exceptionally large, sometimes exhibiting literally global network effects along with coordination and compatibility issues.

Platforms can be controlled by one or more companies in two dimensions: who controls the platform, usually through patents, and who delivers it to the market? Sony owns the PlayStation intellectual property (IP), licenses the rights to build games for it to outside firms, and handles console design and manufacturing in-house. Industry consortia such as the one behind Orbitz may have multiple entities that own the IP, even though there is a single channel to market. The DVD standard, by contrast, is owned by a small number of firms and delivered to the market by hundreds.⁵

Until Linux proved otherwise, conventional wisdom held that platforms needed proprietary control: Microsoft Windows, the AT&T phone network, and the compact disc owned by Sony and Philips are familiar examples. The following exchange between the widely respected MIT economist Lester Thurow and Computer Reseller News, a major trade publication, illustrated a widely held skepticism:

At the same time that Linux and other fully open platforms are enjoying great success, Apple is building a series of mobile platforms that, while in some ways are closed, are also extremely profitable. According to Morgan Stanley's Mary Meeker, the iPhone and iPod Touch surpassed 50 million units shipped in nine quarters after launch.⁷ Netscape Navigator reportedly had 38 million downloads in 18 months, but that could include double and triple counting. In addition, Netscape's Internet distribution model allowed it a substantial advantage over conventional logistics while Apple physically moved all those devices.

The iPhone and Android platforms have spurred a vast ecosystem of software developers. Hundreds of thousands of applications for each platform are available. While about 400,000 are free and the average selling price is $1.44 for paid apps, GPS add-ons from TomTom sell for $50. Other top sellers include mobile editions of both conventional board games such as Uno and electronic games like Madden.

Thus, the iPhone neatly illustrates the interdependent aspects of a platform. Those applications are helping drive truly staggering demands on bandwidth. Researchers at the giant networking company Cisco Systems estimate that global mobile bandwidth demand will increase 66 times in the four years following 2009. Based on AT&T's experience, that number is fully believable: Mobile data traffic increased 4,962% (essentially 50 times) in the three years following the iPhone launch, but time will tell. The wide dissatisfaction with iPhone performance was often blamed on AT&T's network, but provisioning that kind of growth would tax any organization. In addition, as much as the iPhone has stressed the cellular network, the picture would be far worse if Wi-Fi, which is essentially ten times faster, had not picked up so much of the load.

Strategic Levers

The key themes of Gawer and Cusumano's case study of Intel are interdependency and innovation, both of which involve players operating outside the traditional vertically integrated firm yet that need to be managed. While the tech sector is obviously characterized by these dynamics, the authors point out that more and more industries have this structure, in part because of the increasing software content outside computers. As the Intel story unfolds, many of the lessons—about openness, about competition, about internal friction—do in fact translate far beyond the world of semiconductor design and fabrication.

Gawer and Cusumano assert that firms have four basic “levers” they can pull to influence the direction of a platform that typically is not owned by any one firm.⁸ These levers are:

1. Scope of the firm. This is a macroscopic view of the buy/build decision: What gets done inside, or outside, or in both places? How are changes to these priorities and competencies decided and navigated?
2. Product technology. In particular, how are decisions made and executed with regard to architecture, interfaces, and intellectual property? How modular is a product or subunit? How open are the interface technologies* used to integrate third-party innovation?
3. Relationships with external complementors. Does a platform leader follow Intel, which in the 1990s professed not to want to drive companies out of business? Or is the model closer to Microsoft, which repeatedly did so by swallowing software functionality introduced by competitors, typically into the operating system?
4. Internal organization. Intel is a huge, powerful company with the usual fiefdoms and internal competitiveness. How are these delimited and ruled in such a way as to support innovation of interdependent pieces? How are culture and process managed? Perhaps most critically, how are long-term industry efforts (such as the USB standard) accounted for in quarterly financial reporting and performance measurements?

The Intel Architecture Lab was a critically important piece of computing history during the boom of the 1990s, a place that embodied the paradox of “coopetition”: Secrets were shared (in the form of forthcoming Intel architectures and specifications), interoperability was assured (usually) on equal terms, and new markets were invaded (in the case of the incursion of the PCI^† bus on IBM territory). The stated goal was to sell more processors not by stealing share as a primary tactic but by increasing the vitality of the entire PC industry.

Related to this type of position, platform owners must be aware of a particular competitive dynamic known as platform envelopment.⁹ As opposed to traditional firm-based competition (Coke versus Pepsi), platform ecosystems can win or lose somewhat independently of any given company's strategy or execution. 3-D television is one example. For existing platforms, the business model often can be undermined by the incorporation of a platform inside a larger system: Microsoft was fought in court by Real Networks over the nature of Windows-based media players that replicated Real's functionality. Western Union's telegraph business was rendered obsolete by e-mail, a business outcome for which no strategic response (lowering prices, mergers or acquisitions, entry into new markets) would have been adequate. Thus, competition in the age of platforms concerns new kinds of strategic constraints and possibilities.

Finally, because of the highly connected and interconnected nature of the technology landscape, corporate competition may now be less central than platform competition. Consider the long list of contemporary platforms, some not directly competing with any others yet all a product of the Internet age:

• Adobe Flash
• Adobe pdf
• Amazon Kindle
• Amazon Web Services
• Apple iTunes/App Store
• eBay
• Facebook
• Google AdWords/AdSense
• Google Android
• Google Chrome
• Google Maps
• GPS
• LinkedIn
• Linux
• Microsoft Xbox
• Microsoft Windows
• Nintendo Wii
• Salesforce.com
• PayPal
• SAP
• Sony PlayStation
• Twitter
• Wi-Fi
• YouTube

Looking Ahead

The companies, entities, and standards just listed will be key to understanding the next 25 years. As opposed to being the age of the great railroads, or AT&T, or Wal-Mart, American business success will increasingly be defined by platform dynamics: innovation, revenue capture, ecosystem health and development, and lock-in and lock-out. Next we turn to the issue of what kinds of markets will emerge to engage these platforms.

Notes

1. Annabelle Gawer and Michael Cusumano, Platform Leadership: How Intel, Microsoft, and Cisco Drive Industry Innovation (Boston: Harvard Business School Press, 2002), p. 2.

2. Thomas R. Eisenmann, Geoffrey Parker, and Marshall Van Alstyne, “Opening Platforms: How, When, and Why?” Harvard Business School Working Paper 09-030, p. 3., 2008.

3. J. Rochet and J. Tirole, “Platform Competition in Two-Sided Markets,” Journal of the European Economic Association 1, no. 4 (2003): 990–1029.

4. On two-sided markets, see David S. Evans and Richard Schmalensee, Catalyst Code (Boston: Harvard Business School Press, 2007) as well as Thomas Eisenmann, Geoffrey Parker, and Marshall Van Alstyne, “Strategies for Two-Sided Markets,” Harvard Business Review (October 2006). http://hbr.org/2006/10/strategies-for-two-sided-markets/ar/1.

5. Eisenmann et al., “Opening Platforms,” p. 5.

6. John Roberts, “Thurownomics 101,” Computer Reseller News, September 22, 2000, www.crn.com/news/channel-programs/18834629/thurownomics-101.htm.

7. Mary Meeker, Scott Devitt, Liang Wu, “Economy + Internet Trends,” presentation to Web 2.0 Summit, October 20, 2009, San Francisco www.morganstanley.com/institutional/techresearch/internet_ad_trends102009.html.

l8. Gawer and Cusumano, Platform Leadership, p. 40.

9. Thomas Eisenmann, Geoffrey Parker, and Marshall Van Alstyne, “Platform Envelopment,” Harvard Business School Working Paper 07-104., 2007.