How Big Companies Use
Technology To Stifle
Competition

Lesen Sie die deutsche Version hier.

Three decades ago, business magazines declared that low cost personal computers and software were creating a «New Economy». Small startups could affordably access advanced technology, allowing them to compete with the big firms. They could innovate, grow, and eventually disrupt long established dominant firms. Those were exuberant times for startup firms. I know, because I ran one of them, founding one of the first desktop publishing firms. Our company and others provided tools that allowed startups to enter the publishing market. Hundreds of new magazines and newspapers took this opportunity.

But that period of exuberance did not last. My company was acquired, and the publishing market was soon beset with competition from the Internet. More generally, across industries startups soon encountered strong headwinds to growth. Consider the story of Nuance Communications. In 2005, years before Apple’s Siri and Amazon’s Alexa, Nuance merged with another startup, Scan Soft Inc., to attack a burgeoning opportunity in speech recognition. The new Nuance Communications developed powerful speech processing software and grew rapidly for almost a decade—an average of 27 percent per year in sales. Then suddenly, around 2014, it stopped growing. Revenues in 2019 were roughly the same as revenues in 2013. Nuance ran into strong headwinds, as large computer firms that were once its partners became its competitors.

Nuance’s story is far from unique. Across all major industries and technology domains, startups are facing unprecedented obstacles to growth. New companies are springing up to exploit innovative opportunities just as they have in the past. These companies can now tap into an extraordinary flood of venture capital. Yet all is not healthy in the startup economy. Innovative startups are growing much more slowly than comparable companies did in the past.

Surprisingly, a major culprit is technology—proprietary information technology deployed by large firms that dominate their industries. While we usually think of technology as creating disruption, now some technologies are suppressing industrial turnover, which has declined sharply over the last two decades. And this has broad negative implications for the economy, including slower innovation. Researchers have shown that the slower growth of innovative firms has substantially slackened productivity growth. And this is of great concern to economists because slower productivity growth means that incomes grow less.

Nuance is stopped

Nuance began in 1994, as a spin off from SRI, a Stanford laboratory that had developed speech recognition technology for the US government. Scan Soft was a Xerox spinoff. Before the two merged in 2005, speech recognition was constrained by computer processing power. Systems recognized only limited vocabularies, though they nevertheless proved useful in commercial applications such as the transcription of medical records with specialized vocabularies or telephone customer support centers.

By the late 2000s, things changed. With more-powerful computers, Nuance was able to develop a major innovation, «large vocabulary continuous speech recognition.» Now you could say anything about any topic and the technology could accurately transcribe it in real time. Nuance used this technology in an app called Dragon Dictation. When Apple introduced the iPhone 3GS at their 2009 Worldwide Developers Conference, they featured the Dragon Dictation app in their display. Once Apple validated the product, Samsung and all the other phone manufacturers wanted it. So, too, did Google, Amazon, and Microsoft. Nuance grew rapidly, both by signing up these major customers and also millions of individual customers who purchased the iPhone app. The app became the number one business productivity application in the iTunes store. In 2011, Apple introduced Siri, which was based on Nuance technology. Nuance revenues grew to $1.7 billion in 2013.

But this growth was short-lived. Along with Nuance, the Big Tech firms realized that voice was poised to become a prime channel for humans to interact with computers and cloud services. Voice recognition was no longer just about dictation, but about searching for information, shopping, selection of music or video entertainment, controlling appliances, and much more. Compared to keyboards or computer mice, it was handsfree, much faster, it didn’t require typing skills, and it is a much more natural way for humans to communicate.

The Big Tech firms started plowing big investments and talent into this opportunity. While Apple initially purchased the Nuance engine to power Siri, it has since invested in developing its own systems. Amazon began plowing large investments into its Alexa voice assistant. Today Amazon has over 10,000 engineers working on Alexa products, more than ten times the number of core R&D employees Nuance employed at its peak. Google followed quickly with its Home Assistant. And Big Tech has successfully raided Nuance’s talent pool, bringing top people into their folds.

But more important than their financial resources, the Big Tech companies also have the advantage of large customer bases, complementary products, and vast amounts of data at their disposal, enabling them to continually improve their voice recognition systems. Today the data moat is a deep one: there are 300 million Alexa devices installed; Google handles 5.6 billion searches each day on average and half of Google users report using voice for search. And Amazon broadened the range of tasks that Alexa performs by creating an ecosystem where third-party developers add new «skills» much as third-party apps enhance smartphones. There are now over 100,000 of these skills, ranging from playing music on specific radio stations to telling jokes. In addition, Amazon has licensed the Alexa far-field technology to appliance manufacturers to control dishwashers, clothes washers and dryers, and vacuum cleaners. All these activities take Alexa’s voice recognition capabilities far beyond dictation.

Nuance could not compete on this battlefield. It retreated to focus on market niches such as health care. In 2021, Microsoft bought Nuance. This was not a bad outcome, but it reflected the reality that Nuance’s growth had stopped.

Declining disruption

But what happened to Nuance is not just a retelling of the old story of large firms out-investing startups. Across a wide range of industries, dominant firms are employing large scale information systems to outflank their competitors, including innovative startups. They are using large, proprietary software systems to better manage complexity and thus differentiate themselves from rival firms. And this has allowed them to increase their market dominance and to avoid being disrupted. While these systems bring important new benefits to society, they also undermine a critical engine of growth, namely, the expansion of innovative startups. Economists have identified this slowdown as a major contributor to the decline in aggregate productivity growth.

In retail, Walmart’s inventory management and logistics software allows it to stock its stores with a far greater selection of products at lower cost, to tailor each store to local needs, and to respond quickly as demand changes and hot products emerge. Top manufacturers such as Boeing and Toyota are able to offer products with many more features by basing their systems on complex software. Using large data systems, top financial companies tailor credit cards and home equity loans to individual consumers on a massive scale and then to target the marketing of these products. Even the top waste management companies, health insurers, and pharmacy benefit managers are making large investments in proprietary software to outflank their competition. In aggregate, these investments by firms in their internal software (excluding companies whose product is software) now total over $240 billion per year. This represents a huge shift in investment, up from $19 billion in 1985, and it is dominated by large firms. Ranking firms by sales, the top four firms in each industry increased their investment in own-developed software eight-fold since 2000, far more than even second tier firms. Moreover, these investments in software were accompanied by complementary investments in computers, communications equipment, and organizational changes. In many cases, the top firms used the software to enable new business models and built internal «platforms» to realize them.

And these investments by top firms have paid off. Since the 1980s, the top four firms in each industry have increased their market shares by 4-5 percent in most sectors. My research shows that investments in proprietary software caused most of this increase.

This greater industry dominance is accompanied by a corresponding decline in the risk of disruption for the top firms. Ever since Clayton Christensen’s 1997 book «The Innovator’s Dilemma» corporate managers have been obsessed with the disruption of top firms. At the time Christensen wrote his book, disruption had been rising. But since about 2000—when top firms started their investment spree in proprietary systems—the risk of industrial disruption has dropped sharply. In a given industry, the chance that a high-ranking firm (as measured by sales) will drop out of one of the top four spots within four years has declined by half, from over 20 percent to around 10 percent. Here, too, investments by dominant firms in their internal systems largely accounts for the change. While new technologies disrupt entire industries—think newspapers or DVDs—some technologies are now suppressing the disruption of dominant firms.

How does this happen and why does it apparently affect so much of the economy? It is because these business systems address a major shortcoming of modern capitalism. Beginning in the late nineteenth century, innovative firms found that they could often achieve dramatic cost savings by producing at a large scale. The shift dramatically reduced consumer prices, but it came with an important limitation: to achieve large volumes, products and services needed to be standardized. Henry Ford famously declared that customers could have «any color so long as it is black.» Retail chain stores achieved their efficiencies by providing a limited set of products to their thousands of stores. Finance companies offered standard mortgages and loans. Products had limited feature sets; stores had limited selection and were slow to respond to changing demand; many consumers could not get credit or only obtained it on terms that were costly and not suited for their needs.

But software changes the equation, partly overcoming these limitations. Software reduces the costs of managing complexity. With the right data and the right business organization, software allows businesses to tailor products and services to individual needs, to offer greater variety or more product features. And this allows them to beat rivals, dominating their markets. Walmart stores offer far greater selection than Sears or K-Mart stores, and they respond faster to changing customer needs. Sears was long the king of retail; now Walmart is, and Sears is in bankruptcy. Toyota responds quickly, producing new models when it detects new consumer trends; smaller car companies cannot afford the billions of dollars needed to design new car models. Similarly, only Boeing and Airbus can manage to build new highly complex jumbo jets. The top four credit card companies have the data and the systems to effectively target card offers to individual consumers, gaining maximum profit and market share; they dominate the market.

These software-enabled platforms allowed top firms to cement their dominance of markets. It also slowed the growth of rivals, including innovative startups. Slowing startup growth is the flipside of declining disruption.

Startup growth slows

A variety of evidence shows that startup growth has slowed down substantially. Looking at venture-backed startups, the time required to receive funding has lengthened substantially. The median time from the founding of a startup to the time it receives seed round funding grew from 0.9 years in 2006 to 2.5 years in 2020. The median time to late-stage venture funding rose from 6.8 years to 8.1 years. For firms that were acquired, the average time from first financing to acquisition tripled from a little over two years in 2000 to 6.3 years in 2018. For firms that went public, the comparable times rose similarly. A 2019 study of high-quality tech startups found that after 2000, they were less likely to grow sufficiently for a high-value acquisition or IPO.

But the clearest evidence of a slow-down is how fast firms grow when they become more productive. The key feature of dynamic economies, what economist Joseph Schumpeter called «creative destruction», is that more productive firms—firms with better products or lower costs or better business models—grow faster than less productive ones, eventually displacing less productive incumbent industry leaders. The growth of productive firms drives the disruption of leaders. But since 2000, on average, firms with a given productivity grow only half as fast as firms with that same level of productivity did in the 1980s and 1990s. When productive firms grow more slowly, they are less likely to «leapfrog» industry leaders and displace them. Last year, research by myself and my colleague Erich Denk directly linked this slowdown to the greater industry dominance of large firms and their investments in software and other intangibles.

Another view, expressed forcefully by Congressional investigators in hearings and in a report published in 2020, attributes the decline in economic dynamism to a different source: the weakening of government antitrust policy since the 1980s. In this account, large firms have been permitted to acquire their rivals, reducing competition. Acquisitions have made these firms more dominant, especially in Big Tech, supposedly leading to a decline in the rate of entry of new tech firms and a decline in venture capital funding for early-stage firms. But in fact, the entry rate of firms in tech industries is only down modestly from the exceptional surge of the dot-com boom and early-stage venture capital financing is at record levels, with twice as many financings today as in 2006 and four times the dollars invested. The problem isn’t that large firms are preventing startups from entering markets or from getting funding; the problem is that dominant firms are employing large scale technologies that make it harder for startups to grow. Moreover, big firms like Walmart and Amazon have grown mainly by executing superior business models, not by buying rivals. Indeed, the rate of acquisitions by dominant firms has declined since 2000.

Of course, dominant firm acquisitions do sometimes affect startups. Some researchers have identified so-called «kill zones», where Big Tech makes acquisitions to shut down competition, and venture capital is hard to find. But other researchers find that startups often respond by just moving their innovative activity to a different application. Moreover, the prospect of acquisition by a large firm often incentivizes startups. Indeed, despite what happened to Nuance, the entry rate of startups in speech recognition and natural language processing has quadrupled since 2005 and 55 percent of these startups have received venture capital investments.

A problem for society

The slowdown in the growth of innovative startups is not just a problem for a few thousand firms in the tech sector. It reaches all major sectors of the economy, and it affects the health of the entire economy. Researchers at the U.S. Census have shown that the slower growth of productive firms accounts for perhaps as much as a third of the slowdown in aggregate productivity growth. Aggregate productivity measures the amount of output the economy produces per person and it serves as a rough index of economic well-being. The entire economy grows more productive in two ways: individual firms can produce more output per worker or more productive firms can grow faster, driving up the average. The researchers found that firms are still improving their own productivity at about the same rate as in the past, but the slowdown is substantially coming from slower growth of productive firms. Critically, startup firms are growing less rapidly on average and fewer of them grow very fast. The headwinds blowing against Nuance and other innovative startups are responsible for significant economic problems. In my book «The New Goliaths», I also relate this development to growing economic inequality, greater social division, and the declining effectiveness of government.

What will it take to fix this problem? Stronger antitrust enforcement of mergers and acquisitions might help, but the changes in economic dynamism are driven more by changing technology. A more basic problem is that these major new technologies are proprietary; they are only accessible to a small number of huge corporations. No matter how much technologies advance, their benefit will be limited and problems will arise if access to them is limited. In the past, new technologies have spread widely; they were licensed broadly, or firms independently developed alternatives enabling greater competition and innovation. Government sometimes helped this process. Bell Labs developed the transistor but was forced by antitrust authorities to license the technology broadly, creating the semiconductor industry. Similarly, IBM created the modern software industry when it unbundled software, selling it separately from computer hardware; they did this in response to antitrust pressure. Today we are seeing some similar but voluntary openings of technology. Amazon opened up its proprietary IT infrastructure to create the cloud industry, which has had a strong effect on the prospects of small startup firms. Antitrust policy can be used to encourage or compel large firms to open their proprietary platforms. Other policies can also encourage a greater diffusion of technology, such as the treatment of employee noncompete agreements and intellectual property policy.

However, achieving the right policies with regard to antitrust, employee mobility, and intellectual property requires a careful balancing—we don’t want to undercut innovation incentives. We don’t want to discourage firms from developing the technology in the first place because this technology is very beneficial to society. In general, policy needs to strike a balance between providing rewards to firms that advance the technology and spreading the use of that technology broadly so that it does not benefit just a few. But the evidence strongly suggests that current policies are out of balance. Across developed countries, the spread of technology has slowed as evidenced by a growing gap between the productivities of the largest firms and the rest, who lack access to the best technology. And employee mobility has dropped substantially. Moreover, firms have earned immense profits from opening up their technologies, both when they have done so voluntarily, as with Amazon’s AWS, or when they have been pressured by antitrust authorities, as with IBM. These examples suggest that there may be many other opportunities that are both privately rewarding and socially beneficial that are not being pursued because they require bold action by firm leaders. Indeed, even though unbundling proved to be enormously profitable for IBM, firm executives studied the opportunity, but failed to act until they were coaxed by antitrust authorities.

Finding the right policies will be difficult, it will take time, and it will involve intense political jockeying. But the starting point is to recognize that today’s economy is one where technology has taken on a new role and is being used to suppress disruption and competition.