فصل 7

Chapter 7

Looking Forward

Platforms and the Future

The dramatic testimony of Mark Zuckerberg in front of the U.S. Congress and European lawmakers in the spring of 2018 signaled a new age for the business of platforms. Complaints about fake news, privacy violations, unchecked expansion, and the growing specter of antitrust and labor regulation put the most valuable companies in the world on notice. Platforms were no longer neutral matchmakers or intermediaries for transactions and innovations. Allowing anyone on any side of a platform completely free rein could be potentially dangerous to democracy, social well-being, and global economic stability. Whether they liked it or not, companies such as Apple, Amazon, Microsoft, Alphabet-Google, Facebook, Alibaba, Tencent, and others would need to accept the new roles that their platforms played in the world economy and modern society as a whole.

Even as policy makers and thought leaders were criticizing platform businesses, it is worth remembering that successful platforms have long lives. With millions or billions of participants engaged and connected, platforms tend to be more enduring than stand-alone product or service businesses. Microsoft, for example, introduced DOS and then Windows in the 1980s. Since the mid-1990s, it has faced attacks from antitrust authorities, security problems, and growing competition from open-source software and cloud computing. Nevertheless, Microsoft has remained dominant in PC operating systems and key applications for nearly four decades and is still enormously profitable. As we have seen with other examples throughout this book, successful platforms are difficult to unseat.

However, nothing lasts forever. Platforms, ecosystems, and the technologies that drive them will continue to evolve and change. Computing and communications platforms have faced continuous threats from new technologies over the past forty years. Some companies (ranging from Yahoo and MySpace to Nokia and BlackBerry) have seen their fortunes dramatically decline in short periods of time. Looking at the bigger picture, we can see that mainframes ultimately gave way to personal computers, the Internet, social media, mobile devices such as smartphones, and cloud computing. Old and new platforms coexist, though some have become more important than others. Looking forward from what we know today, artificial intelligence, machine learning, virtual and augmented reality, blockchain applications, and even quantum computing are likely to challenge currently dominant platforms, at least in some domains.

In this last chapter, we use the principles emphasized in this book to explore ongoing and future platform battlegrounds. We start by summarizing the book’s key arguments. We then identify four trends likely to impact today’s dominant players and the platform entrepreneurs and managers of tomorrow. We also explore how to evaluate emerging platforms and examine a few key competitions around next-generation technologies. In conclusion, we come back to the argument that large, successful platforms must increasingly engage in self-regulation and curation in order to adapt to a rapidly changing world. The end of truly open platforms seems to be upon us.

The Business of Platforms

Table 7-1 summarizes the core principles we have explored in this book. Understanding the business of platforms begins with the drivers of platform markets in general and digital competition more specifically: the potential for generating strong network effects, limiting multi-homing, restricting differentiated and niche competition, and building high barriers to entry. We also pointed out that platform thinking is not new: More than one hundred years ago, platforms like the telephone and the Yellow Pages, and those in several other industries, including railroads, electric power, radio, and television, depended heavily on network effects, low levels of multi-homing, and complementary innovations from ecosystem participants. But the platforms of today are largely digital, which is new. This technology, combined with the Internet, has enabled rapid, exponential growth on a global scale. It has also produced giant platform companies like Amazon, Alibaba, and Tencent with widely diversified businesses, often connected through user data. Some of these markets we can classify as winner-take-all, but most we cannot.

An important contribution of this book was highlighting the similarities and differences between two kinds of platforms. Most of the early platforms facilitated innovations, while the explosion of new platforms in the last decade or two has been driven by transactions. However, regardless of the platform type, managers had to deal with the same business challenges: Choose the key sides of the platform, solve the chicken-or-egg problem, design a business model (generate revenues and profit), and establish rules for using the platform as well as cultivating and governing the all-important ecosystem. As we discuss below, the commonalties or complementarities between transaction and innovation platforms have led to a growing number of hybrid companies, and this complex business model seems likely to become more common in the future.

Many industry commentators have positioned platforms as the Holy Grail of business strategy, as the one sure way to market dominance and enduring profits. This may be true if you end up as the “top dog” in your industry. But we also found that vast numbers of platform wannabes fail; in fact, there seemed to be far more losers than winners, just as the majority of start-up companies fail (some research indicates as many as 90 percent). Because of the complexity of platform businesses, it is especially easy to make mistakes. We found the biggest errors around mispricing on the most important market side, the failure to develop trust, mistiming entry, and simply ignoring the threat of competition. When firms get lulled into believing markets have tipped “permanently” in their favor, there is great danger in becoming complacent. Microsoft’s failure in browsers was a classic example of overestimating market power when you are ahead and its failure in smartphones was a classic example of underestimating the power of network effects when you are behind.

One of the most daunting challenges is figuring out how established non-digital companies can adapt to platform competition. We argued that platforms are a serious threat to many traditional businesses, but there are alternatives. Old dogs can indeed learn new tricks, however difficult that may be. We suggested a simple framework for traditional businesses: Build, buy, or belong to a platform. When you are small, and when are you trying to prevent a market from tipping, established firms can leverage the power of platforms by joining an existing player. When you are large and time-to-market is essential, buying a platform can provide the skills and technology you may not have, and help traditional firms get over the hump and into the platform business. Building a platform from scratch is the hardest strategy and not for the faint of heart. But if you can pull it off, the rewards can be substantial.

Finally, we discussed how platforms have become double-edged swords: Platforms have enabled some firms to achieve great economic, social, and political power, which can be easily abused, sometimes unwittingly. With platforms increasingly coming under intense government and media scrutiny, it is critical to remind managers and entrepreneurs not to be a bully. There is nothing illegal or ethically wrong with following the logic of platforms and network effects into a dominant position. But once you become dominant, then different rules should apply. Platform companies need to anticipate the growing likelihood of antitrust intervention. In addition, platforms have enabled the sharing or gig economy, which has a logical extension: Every worker can become a temporary contractor. But, as we have seen in the backlash to Uber and other gig-economy platforms, that is not a viable labor strategy for the long term. Furthermore, fraud, violations of privacy, poor quality goods, and other platform “complications” have the potential to torpedo trust, which is fundamental to platform success. Most platforms are digital intermediaries or innovation facilitators; they do not have personal relationships with their users and complementors. Consequently, we believe that the largest, most powerful platforms will increasingly need to balance openness with economies of scale and scope, and self-regulate or curate to avoid potentially punishing conflicts with policy makers.

Platforms and the Future

The platforms we talked about in this book all started sometime in the past century or earlier. We know a lot about the histories of the company strategies and operations. But another question to explore is how emerging platforms are likely to evolve over the next decade and beyond. We see at least four major trends that could change the way we think about the business of platforms in the future.

First, digital competition will turn more and more platform firms into hybrids. In the old world (1980s and 1990s), innovation and transaction platforms were distinct businesses. Connecting buyers and sellers, advertisers and consumers, or users of the different social networks appeared to be fundamentally different from stimulating outside firms to create complementary innovations in the form of their own products and services that make the platforms increasingly valuable. But in the last decade a growing number of successful innovation platforms have integrated transaction platforms into their business models, and transaction platforms have sought to open APIs and encourage third parties to create their own complementary innovations. Rather than lose control over distribution, innovation platforms want to manage the customer experience (think Apple’s App Store). And owners of transaction platforms recognize that not all innovation can or should be internal (think Facebook’s platform). Prominent examples include Google’s decision to buy and push Android, Amazon’s decision to create multiple innovation platforms around AWS and Alexa, and the decisions of Uber and Airbnb to allow developers to build services on top of their transaction platforms. Another example is Snapchat, which allowed users to post messages or photos that quickly disappeared. It struggled to make a profit as a pure transaction platform, especially after Facebook’s Instagram copied many of its most popular features. To inspire more innovation and user activity, Snapchat decided in June 2018 to open its platform and user database to application developers. The underlying driver was digital competition. Unlike in the traditional economy, where companies require expensive physical investments to build out the business model, in the digital world, companies can grow rapidly with a clever use of data, software, and platform strategy.

Second, we see next-generation platforms driving innovation to a new level. Advances in artificial intelligence, machine learning, and big-data analytics enable organizations to do more things with less investment, including building businesses that were impossible in years past. Although AI is still in a nascent phase, Google, Amazon, Apple, IBM, and other firms are no longer treating their technology as proprietary. Instead, they have turned AI capabilities into platforms that third parties can access and build upon for their own applications.

Third, the logic of platform thinking, which is driven by network effects, multisided markets, and the potential for winner-take-all-or-most outcomes has led to growing market power concentrated in a small (but rising) number of firms. In the 1960s and 1970s, IBM represented the pinnacle of platform power. In the 1980s and 1990s, it was Intel and Microsoft. In the last two decades we are contending with the market power of Apple, Google, Amazon, Facebook, Alibaba, and Tencent, among others.

Lastly, we see virtually all large platform companies evolving from free markets to curated businesses. As we discussed in Chapter 6, many managers, entrepreneurs, and technical experts once believed that platforms would only bring “good” into the world: They would connect people, products, and services at ever-decreasing prices, and free the world from the frictions and imperfections of traditional marketplaces or modes of communication. But as we have suggested throughout this book, the new world and traditional worlds have to coexist. Not all actors in the digital world are do-gooders. Partisan politics, spies, terrorists, counterfeiters, money launderers, and drug dealers all have found ways to use platforms to their advantage. Platform companies are also profit-seeking enterprises, and this motive can sometimes lead to abuses of power and technology from the perspective of users and ecosystem partners. Once platforms get large enough to impact social, political, and economic systems, then they will increasingly need to reflect on the purpose of their operations as well as evolve from hands-off to more hands-on curation. Although it is a cliché, for the world’s biggest platforms, growing power means growing responsibility.

New Platform Battlegrounds

With all these issues in mind, we can look to several platform battlegrounds currently under way to help us think about what comes next and the increasing role platforms are likely to play in the future. Some early-stage platforms may evolve into proprietary products and services. And some current products and services or path-breaking technologies may turn into new types of platforms. In the remainder of this chapter, we discuss two relatively new platform battlegrounds and their possible evolution, if artificial intelligence impacts them the way we predict: voice wars and autonomous vehicle ride sharing. Then we will look at two emerging and future battlegrounds: quantum computing and gene editing.

CURRENT/ONGOING

Perhaps the most important new technology in the battle for platforms over the next decade is artificial intelligence and machine learning. For many industries, AI has disruptive potential. Two of the most obvious and powerful applications for AI are voice recognition and driverless cars. Both involve a dramatic change in platform ecosystems.

VOICE WARS: RAPID GROWTH BUT CHAOTIC PLATFORM COMPETITION

Although artificial intelligence has been around for decades, one branch has made exceptional progress: machine learning (using special software algorithms to analyze and learn from data) and the subfield of deep learning (using hardware and software to build massively parallel processors called neural networks to mimic how the brain works). Applications of these technologies have led to dramatic improvements in certain forms of pattern recognition, especially for images and voice. Apple got the world excited about a voice interface when it introduced Siri in 2011. For the first time, consumers had access to a natural conversation technology that actually worked (at least some of the time). Despite its first-mover advantage, however, Apple’s strategy for Siri was classic Apple: The company designed Siri as a product that complemented the iPhone, not as an innovation or transaction platform that could generate powerful network effects.

When Amazon introduced the Echo speaker device with Alexa software in late 2014, it set in motion a war for platform domination among Google, Apple, Microsoft, Alibaba, Tencent, and a host of start-ups. Amazon’s strategy was to create a new platform powered by a combination of Amazon Web Services, speech recognition, and high-quality speech synthesis. CEO Jeff Bezos sought to bundle the technology with an affordable piece of dedicated hardware. Immediately identifying the potential for cross-side network effects, Amazon launched its Alexa Skills Kit (ASK)—a collection of self-service APIs and tools that made it easy for third-party developers to create new Alexa apps. This open-platform strategy accelerated the number of Alexa skills from roughly 5,000 in 2016 to over 50,000 in 2018. Amazon offered a wide variety of skills, such as playing games like Jeopardy!, ordering an Uber ride, and asking about the weather and news. Our favorite app is that we can ask Alexa to remotely start a car, which will cool down the car in the summer and warm it in the winter. If you don’t show up, the car shuts down after ten minutes.

The combination of very low prices (which were most likely below cost for Amazon and Google) and extraordinary ease of use led to an explosion in sales of these intelligent assistant devices. As an early mover, Amazon quickly captured the largest market share. But Amazon’s success spurred Google, Apple, Samsung, and various Chinese companies into action. By late 2017, voice was morphing into a classic platform battle: Amazon and Google heavily discounted products to build their installed base, with each side racing to add applications and functions. The goal was to drive network effects. Amazon, for example, released its Echo Show, an Alexa speaker with a visual display. Amazon hoped that it would spread from family member to family member and to friends. Then, anyone with an Echo Show could make visual phone calls to each other, much like Apple’s FaceTime. All the major players were also licensing their technologies to consumer electronics companies, car companies, and enterprise software companies, hoping they would incorporate their voice solutions, often for free.

The platform challenge was that multi-homing was easy: Any customer could own or use the Google, Amazon, Microsoft, and Apple voice interfaces. There were no significant switching costs (yet) for consumers. There were also many opportunities for differentiation and niche competition: Apple focused on the quality of music; Amazon on media and e-commerce; Google on search-related inquiries; and Microsoft on enterprise needs.

Each player also had a different business model. Amazon was building a hybrid platform, with third parties creating applications and customers doing transactions with their Echo devices. Indeed, the average Amazon household spent $1,000 annually on the site; Prime members spent $1,300; and households with an Echo spent $1,700. Apple initially tried to make money on its hardware (which explained its high prices and low initial market penetration). No company in 2018 seemed to have a clear path to making a profit directly from this technology.

As we finished this book, it was too early to tell how the voice wars will play out. The market was still like the Wild West—more chaos than order. Between 2017 and 2018, improvements in machine learning and deep learning were creating better voice experiences across all competitors. Google appeared to be the technical leader in AI, with many applications in search, advertisements, and machine translation, among others. Apple, which lagged behind in early benchmarks, was improving quickly, as were the benchmarks for Microsoft’s Cortana and Amazon’s Alexa. In 2018, Google had the advantage of hundreds of millions of devices (Android smartphones) that have Google’s voice capabilities embedded. Amazon had the advantage of the largest smart-speaker‒installed base, with tens of millions of devices sitting in users’ homes, especially in the United States. Ultimately, we expect voice to be a classic platform battle, where the winner(s) will depend on who can build up the largest installed base of users, who can create the best ecosystem for producing innovative applications, and who (if anyone) can lock in their customer base, limit multi-homing in the future, and create a sufficiently compelling solution to reduce competition from niche players and differentiation in the market.

RIDE SHARING AND SELF-DRIVING CARS: FROM PLATFORM TO SERVICE

While AI will spawn a range of new platforms, it will also enable new capabilities that may destroy existing platforms. One of the most exciting AI applications has been the emergence of self-driving cars. Ironically, this new technology may replace some of the most widely used platforms in the world: Uber, Lyft, Didi Chuxing, and other ride-sharing businesses. Despite the strong cross-side network effects, the ride-sharing platform revolution could actually disappear.

The business challenge for ride-sharing platforms is simple: They tend to lose money, and lots of it. The cost of attracting and paying drivers as well as keeping ride prices low has squeezed profit margins. In addition, many drivers multi-home (serve both Uber and Lyft, or conventional taxi companies). Therefore, Uber, Lyft, Didi Chuxing, and other ride-sharing companies have announced that their long-term strategy is to move away from being a pure platform, matching riders with drivers, toward a model of “transportation as a service,” in which they own or lease all their own vehicles, including both automobiles and bicycles or scooters. Tech companies like Google and most of the major automobile manufacturers like GM and Toyota were also investing aggressively in the same direction. Despite a long history of selling products, even the most conservative car companies see AI as the route toward becoming a service company. As Lyft CEO Logan Green said in 2018, “We are going to move the entire [car] industry from one based on ownership, to one based on subscription.” The emergence of autonomous vehicle technology promises to remove human drivers, which could dramatically drive down the marginal cost of transportation services. Amortizing the R&D and fleet costs of self-driving cars is likely to be very high. But the economics could improve because there are no driver payments and cars will be utilized more intensively, dramatically reducing the cost per mile. GM estimated that, when it launches its service in 2019, rides would initially cost $1.50 per mile, 40 percent less than current ride-hailing services. Some estimates suggested that the cost per mile of a self-driving vehicle could fall as low as 35 cents per mile, down from an average of $2.86 per mile in 2018. Observers see the combination of new technology and better economics forcing Uber (and other ride-sharing platforms) to “either figure out a way to buy or at least manage an enormous fleet (possibly by going public to foot the bill), or face annihilation by others who will.” Facing this threat, Uber began investing in autonomous vehicle technology in 2014. Uber’s cofounder and then CEO Travis Kalanick stressed the importance of winning the race: “The minute it was clear to us that our friends in Mountain View [i.e., Google] were going to be getting in the ride-sharing space, we needed to make sure there is an alternative [self-driving car]. Because if there is not, we’re not going to have any business.” Kalanick added that developing a self-driving car “is basically existential for us.” Uber announced in November 2017 that it would buy 24,000 self-driving vehicles from Volvo, giving it a fleet to test and later deploy in an autonomous ride-hailing service.

Lyft has taken a different approach. Rather than develop its own self-driving technology, it is trying to form partnerships through its “Open Platform Initiative,” which resembled Google’s Open Handset Alliance for Android smartphones. Lyft’s platform initiative brings together several automakers, including GM, Land Rover, and Ford, to integrate their autonomous vehicle projects into one ride-hailing network. Initially, the Open Platform Initiative is offering partners access to ride data for testing purposes, but ultimately plans to make their self-driving vehicles available on its ride-hailing platform. Lyft’s chief strategy officer noted in late 2017 that “we’re focused on partnering with the auto industry because frankly, we think we can’t do this alone and need each other to be successful.” Lyft cofounder John Zimmer even predicted that “autonomous vehicle fleets will quickly become widespread and will account for the majority of Lyft rides within 5 years.” Lyft’s strategy may signal the emergence of a different type of transaction platform, where Lyft connects riders to self-driving vehicles from a variety of manufacturers. But many of its partners have also invested in ride-hailing technology and could launch self-driving taxi services of their own. In response, even Lyft invested in a self-driving research center to develop its own autonomous vehicle technology, indicating that it, too, might move away from the open-platform model. Who will win and who will lose remains uncertain. Moreover, self-driving car services may not come so quickly and may never be as profitable as high-volume transaction platforms like eBay, Priceline, and Expedia, and even Airbnb, which are asset-light and primarily match buyers and sellers or users and suppliers. Nonetheless, future consumers are very likely to benefit from more and cheaper ride-sharing services as long as these businesses have enough capital to survive until the businesses are profitable.

Emerging/Future

Now let’s look at two other emerging platform battlegrounds: quantum computing and gene-editing technology. The nascent businesses and ecosystems are even more dependent on advances in science and technology, but the platforms should become more relevant in the future and remain so for many decades to come.

QUANTUM COMPUTERS: AN INNOVATION PLATFORM FOR NEXT-GENERATION COMPUTING In 1981, Nobel laureate Richard Feynman challenged the physics and computing communities to build a computer mimicking how nature actually works—a quantum computer. Universities and then companies started research. By 2015, McKinsey consultants estimated there were 7,000 researchers working on quantum computing, with a combined budget of $1.5 billion. By 2018, dozens of universities, approximately thirty major companies, and more than a dozen start-ups had notable R&D efforts. The state of the technology today resembles conventional computing in the late 1940s and early 1950s. We have laboratory devices and some commercial products and services, but mostly from one company. We have incompatible computer technologies, mostly in the research stage and with different strengths and weaknesses. All the machines require specialized skills to build and program. Companies still work closely with universities and national laboratories. There is no consensus as to what is the best technology or design. Nonetheless, we believe quantum computers represent a revolutionary innovation platform for specialized applications, with the potential to generate new types of transaction platforms for “quantum computing as a service” and secure quantum communications.

Quantum computers are built around circuits called quantum bits, or “qubits.” One qubit can represent not just 0 or 1 as in traditional digital computers, but both 0 and 1 simultaneously. Qubits give quantum computers the potential to perform astounding calculations, far beyond the reach of conventional digital computers. As few as 300 qubits can represent information equal to the estimated number of particles in the known universe. To perform calculations, however, qubits need to exploit some unique properties described by quantum mechanics, and that makes building and using large quantum computers difficult.

There are several competing technologies, all with the potential to make quantum computers that are more stable, scalable, and flexible than current devices, most of which reside in research laboratories. The 2018 business and patents leader is D-Wave, a private company spun out of the University of British Columbia in 1999 to accumulate patent rights in exchange for research grants. It has been funded mainly by venture capital, corporate investors such as Goldman Sachs, and, more recently, Jeff Bezos and the Central Intelligence Agency. Google and IBM, as well as start-ups such as Quantum Circuits, are deploying a different approach, using electrons or nuclei. Xanadu, a Toronto start-up, uses photons to build its quantum circuits. Microsoft has yet another design, which it plans to build within five years and make commercially available via the cloud. There are potentially strong network effects associated with the number and quality of programming tools and applications for the competing quantum computers, but the ecosystems remain at a very early stage. The most important applications are likely to be mathematical problems such as in combinatorial optimization that require massively parallel computations. For example, in 2012, Harvard researchers used a D-Wave computer to simulate protein molecule unfolding (which is useful in drug discovery). More recently, Northrop Grumman has been using D-Wave to model software systems to detect errors. And Volkswagen has been using D-Wave to optimize traffic routes for thousands of vehicles simultaneously, potentially useful for self-driving vehicles. Perhaps the “killer app” will be quantum encryption and secure communications. These applications utilize an algorithm discovered in 1994 by Peter Shor, formerly of Bell Labs and now at MIT. Shor demonstrated how to use a quantum computer to factor very large numbers, which makes it possible to create unbreakable cryptographic keys. Governments (the United States and China in particular) as well as companies (AT&T, Raytheon, Alibaba, Huawei, NEC, and Toshiba, among others) are pursuing these applications. China is especially advanced. Will quantum computing become a successful new platform business? Network effects are weak so far because the application ecosystems are still nascent and divided among several platform contenders. As of this writing, D-Wave has the lead in applications and the largest patent portfolio, followed by IBM and Microsoft. IBM leads in recent annual patent filings. At universities, the patent application leaders are MIT, Harvard, Zhejiang (China), Yale, and Tsinghua (China). At the country level, the United States led with about eight hundred total patents, three to four times the numbers from Japan and China. For the business to progress faster, however, quantum computers need more researchers able to apply these patents. They, in turn, need access to bigger quantum computers so they can build better programming tools and test real-world applications. IBM, D-Wave, Google, and Microsoft were aggressively moving in this direction and making their quantum computers available as a cloud service.

Quantum computers will probably never replace conventional computers. Nor do we see this as a winner-take-all-or-most market. Quantum computers are likely to remain special-purpose devices that exploit quantum phenomena for certain types of massively parallel computations. They are not well suited to everyday computing tasks that require speed, precision, low cost, and ease of use. Multi-homing on different types of quantum computers is also likely to persist, keeping potential application ecosystems divided and weakening network effects. In particular, D-Wave computers cannot run Shor’s algorithm and so are not useful for cryptography or quantum communications. IBM, Google, and Microsoft, as well as several start-ups, are designing more general-purpose devices, although these remain experimental or small-scale.

Quantum computing as a platform is also likely to face serious challenges in regulation because of cryptography applications. On the one hand, quantum computers may be able to break secure keys generated by the most powerful conventional computers, which now protect much of the world’s information and financial assets. On the other hand, quantum computers themselves can potentially generate unbreakable quantum keys as well as facilitate secure quantum communications. Combined with hacker tools for entering computer systems and accessing cryptocurrencies, quantum computers clearly have the potential for social mischief as well as good. They can help solve currently impossible computation problems but also facilitate unbreakable data silos to hide illegal or unethical activities. The leading companies will have to figure out policies to regulate themselves as well as to work with governments, which are likely to play a major role in overseeing at least some quantum computer applications and services.

CRISPR: AN INNOVATION PLATFORM FOR GENE EDITING

Gene editing—altering DNA to modify the characteristics of plants, animals, and even people—was already a global market worth over $3 billion in 2017 and expected to double in the next five years. In 2018, there were also over 2,700 clinical trials under way for human gene therapies. Much of the know-how remains in the research or pre-commercialization stages, though some technologies are spawning nascent innovation platforms and ecosystems, similar to what we have seen in quantum computers and other industries. One particularly promising technology is CRISPR, or “clustered regularly interspaced short palindromic repeats.” CRISPR refers to small pieces of DNA that bacteria use to recognize viruses. What scientists observed years ago is that specialized segments of RNA and associated enzymes in one organism can modify genes (DNA sequences) in other organisms. For example, this happens naturally when the immune system in bacteria fight against an invading virus. In 2012, several scientists discovered they could use CRISPR sequences of DNA as well as “guide RNA” to locate target DNA and then deploy CRISPR-associated enzymes as “molecular scissors” to cut, modify, or replace genetic material. The potential applications include diagnostic tools and treatments for genetic diseases as well as genetic reengineering more broadly. An August 2016 article in National Geographic magazine described CRISPR’s potential: CRISPR places an entirely new kind of power into human hands. For the first time, scientists can quickly and precisely alter, delete, and rearrange the DNA of nearly any living organism, including us. In the past three years, the technology has transformed biology. . . . No scientific discovery of the past century holds more promise—or raises more troubling ethical questions. Most provocatively, if CRISPR were used to edit a human embryo’s germ line—cells that contain genetic material that can be inherited by the next generation—either to correct a genetic flaw or to enhance a desired trait, the change would then pass to that person’s children, and their children, in perpetuity. The full implications of changes that profound are difficult, if not impossible, to foresee. Editing a human embryo’s germ line is not simply a hypothetical possibility. In December 2018, reports surfaced that a “rogue” Chinese scientist already had used CRISPR to disable a gene in twin unborn babies that would make them resistant to HIV. The scientist reportedly couldn’t disable both copies of the gene in one of the embryos but implanted it anyway. The babies were born normally, but the disabled genes could make them, and their offspring, susceptible to other diseases. This apparently secret experiment (the scientist has not yet published data to confirm what he did) would have been illegal in the United States and some other countries. It has also created considerable consternation among the scientific community because of the apparent lack of global controls on the use of a powerful technology. Nonetheless, gene editing will continue to evolve. The technology will provide opportunities for companies to pursue product solutions, such as to build stand-alone diagnostic tools or gene therapies for problematic diseases and conditions. This is possible because DNA resembles a programming language and data-storage technology that can be adapted to different contexts. Some institutions and companies have already created products, tools, and components that other firms are building upon. Like today’s quantum computers, however, there are limitations. Each use of CRISPR requires specialized domain knowledge, such as the genome of a particular organism and disease, and then tailoring to the application, such as to design a diagnostic test or therapeutic product for a specific disease or to reengineer a plant to fight off insects. But, along with rising numbers of CRISPR researchers, platform-like network effects and multisided market dynamics are also appearing and helping the ecosystem grow. In particular, more research publications have led to improvements in tools and reusable component libraries, which have attracted more researchers and applications, which in turn have inspired more research, tool development, applications, venture capital investments, and so on.

An important player in the nascent CRISPR ecosystem is a nonprofit foundation called Addgene, founded in 2004 by MIT students. It funds itself by selling plasmids, small strands of DNA used in laboratories to manipulate genes. Since 2013, it has been collecting and distributing CRISPR technologies to help researchers get started on their experiments. The Addgene tools library consists of different enzymes and DNA or RNA sequences useful to identify, cut, edit, tag, and visualize particular genes. There are also numerous start-ups, some of which have already gone public. CRISPR Therapeutics (founded in 2013) is trying to develop gene-based medicines to treat cancer and blood-related diseases, and is collaborating closely with Vertex and Bayer. Editas Medicine (2013) and Exonics Therapeutics (2017) are tackling diseases such as cancer, sickle cell anemia, muscular dystrophy, and cystic fibrosis. Beam Therapeutics (2018) plans to use CRISPR to edit genes and correct mutations. Mammoth Biosciences (2018) is following more of a platform strategy and developing diagnostic tests that could be the basis for new therapies. The company is broadly licensing its patents and encouraging other firms to explore therapies based on its testing technology. In fact, Mammoth’s goal is to create “a CRISPR-enabled platform capable of detecting any biomarker or disease containing DNA or RNA.” In a recent public statement, the company summarized its strategy to cultivate an applications ecosystem: Imagine a world where you could test for the flu right from your living room and determine the exact strain you’ve been infected with, or rapidly screen for the early warning signs of cancer. That’s what we’re aiming to do at Mammoth—bring affordable testing to everyone. But even beyond healthcare, we’re aiming to build the platform for CRISPR apps [italics added] and offer the technology across many industries. Broad commercialization of CRISPR is still years away. The technology is also better at screening, cutting, and rewriting rather than inserting DNA. And only recently have medical centers and companies applied to start CRISPR-related clinical trials. There are also alternative technologies with different strengths and weaknesses. In particular, TALEN (transcription activator-like effector nuclease), another gene-cutting enzyme tool, seems to be more precise than CRISPR and more scalable for some non-laboratory applications, though it can be more difficult to use. In general, CRISPR is in the lead as a potential gene-editing technology platform, with several universities and research centers, start-up companies, and established firms actively publishing papers, licensing and applying for patents, and sharing their tools and depositories of genetic components. Most researchers are also focusing on CRISPR-Cas9, a specific protein that used RNA to edit DNA sequences.

One concern we have is that the business models of biotech start-ups and pharmaceutical companies depend on patent monopolies, making the industry ultra-competitive and locking applied research into protected silos. The result is potentially a “zero-sum game” mentality. This contrasts to the more cooperative (but still highly competitive) spirit of “growing the pie” together that we generally see with basic science and which we saw in the early days of the personal computer, Internet applications, and even smartphone platforms such as Google’s Android. Of course, most CRISPR scientists openly shared and published their basic research. And although the U.S. Patent and Trademark Office already has granted hundreds of patents related to CRISPR, patent holders usually offered free licenses to academic researchers, even those still under litigation.

Ethical and social issues might also hinder widespread use of gene editing, especially if more “rogue” and potentially dangerous misuses of CRISPR occur. The debates are clearly more serious than what we discussed in Chapter 6 regarding the abuse of social media platforms. The broader controversy involving CRISPR centers on how much genetic engineering we, as a society, should allow. Experts already disagree about the safety of genetically altered plants and animals that contribute to the human food supply. Scientists can deploy similar technology to change human embryos and cells, and we might someday control how this is done to treat genetic diseases or potential disabilities. But should we allow parents to edit their children’s genes—for example, to prevent diseases like HIV that they might or might not contract, or to select for blue versus brown eyes, or a higher IQ? In sum, platform dynamics has been influencing industries and technologies outside of personal computers, Internet applications, and smartphones. Again, though, it is not so clear how to use the power of the platform wisely and safely, and what types of government monitoring and self-regulation are most appropriate. These issues seem likely to become even fiercer topics of debate as CRISPR and other gene-editing technologies evolve into more widely used platforms for medical, food, and other applications.

Final Thoughts

Industry-wide platforms and global ecosystems for innovations and transactions have already changed many aspects of our personal and working lives. Many more changes and new technologies will come. We are referring to more than technologies like voice-powered AI assistants and self-driving cars, or quantum computing and gene editing. More broadly, the explosion of transaction platforms has enabled nearly every type of exchange imaginable in today’s world. Platform entrepreneurs have made “anything as a service” possible. We are heading into a future where we will buy and own fewer products (cars, bikes, vacation homes, household tools, consumer electronics, etc.), and we will contract for more services directly with each other. We will likely manage this sharing through peer-to-peer transaction platforms and technologies such as blockchain, to ensure secure and transparent exchanges. To a large degree, transaction platforms originated in ancient bazaars as well as nineteenth-century advertising businesses like the Yellow Pages and shopping catalogues. However, no one predicted such a rise in the popularity, diversity, and global reach of modern transaction platforms, which initially appeared as applications that made personal computers and smartphones more valuable as innovation platforms.

One ongoing challenge for all platform companies is the centralization of power. Yes, the Internet once promised to deliver a “flat world,” where distributed computing and communication networks provided equal access to digital information and economic opportunities. While this is partly true, the opposite trend has also emerged. Platform dynamics has led to more centralization of economic and social activity in a relatively small number of companies, and they seem to be getting increasingly large and powerful with each passing day. In response, we see growing demands from both users and governments to regulate or break up some of the biggest platforms. This movement brings to mind how the muckrakers at the turn of the twentieth century called for the dissolution of Standard Oil and other monopolies.

Today, there are growing calls for governments to rein in platform businesses and for entrepreneurs, managers, and boards of directors at the leading platform companies to take more responsibility for their social, political, and economic power. Purely “open” platforms, with no rules overseeing access, actions, or content, have made some platforms seem like the lawless American Wild West, where “might” dominated “right.” Consequently, we believe that platform companies need to respond to these calls by self-regulating and curating. They need to set limits for who can do what on the platform. Both self-regulation and curation are central to good governance and will become more important in the future, even though they may weaken network effects, financial returns, and growth opportunities for some platform businesses.

We encourage platform managers and entrepreneurs to have great ambitions. We hope this book can help them. However, different rules should apply once a platform business gains a certain size and influence, or achieves the ability to link markets in heretofore unseen ways, such as through new types of data-driven economies of scale and scope. It is the joint responsibility of leaders from government, society, and business to figure out the new rules and make sure competition remains transparent and fair to the extent possible. If not, platform businesses that openly abuse their power are likely to fail in the long run, or at least fail to achieve their potential to do good things. There is also the constant danger of individuals misusing their access to these global platforms. Governments, universities, and companies need to work together, and work harder, to figure out how to curb platform abuses.

Our call for self-regulation and curation also has implications for the kinds of leaders and managers that platform companies will need in the future. It takes courage to make decisions that raise costs, reduce advertising revenue, or suppress network effects and growth potential. Nonetheless, the present reality requires platform companies to expand their strategic visions and definitions of success. Misusing platform power and technology is not a clever way to build an enduring ecosystem or contribute to a stable society, which is good for business. We must measure success in terms that go beyond sales, profits, and market values, although these metrics remain essential to a sustainable business model. We hope that company executives and boards of directors at the top platform companies, as well as leaders in government and academia, will recognize that times have changed. We need to choose a next generation of leaders with a better understanding of how platforms and digital technologies can impact society and the global economy.

In this book, we tried not to exaggerate the importance of platform businesses, present and future. Nor did we try to overly simplify what a platform company must do in order to survive and thrive. Rather, we applied some cold logic and hard data—and several decades of experience. Our goal has been to help managers and entrepreneurs build platform businesses that can stand the test of time and win their share of battles with both digital and conventional competitors.

The bottom line is that platforms have the potential for both good and evil. This is why we say that platforms are double-edged swords. Every major company we cited in this book has been the subject of government investigations, local regulatory oversight, and intense media scrutiny. No one has been spared. Microsoft, Alphabet-Google, Apple, Intel, Facebook, Cisco, Qualcomm, Uber, Airbnb, Alibaba, Tencent, and many other firms, small and large, have faced legal, taxation, or regulatory challenges.

At the same time, the data suggests that industry platforms offer a more efficient way to organize the innovation process and many other types of economic activity. Platforms have already delivered revolutionary change. But we now live in a world where the “business of platforms” has become intimately tied to digital competition, innovation, and power—for better and for worse. It is up to us whether platforms in the future improve the world or undermine it. We are optimistic but cautious.