The trouble had started years before, when Yelp, which was at that point just a nascent start-up, signed a deal with Google, allowing the search giant to use some of its content, including the reviews of local services that users had posted on Yelp. This was a win-win, because it allowed Google, which at the time didn’t really do local search, to have access to more specific content that lived on Yelp’s site, and it increased traffic for Yelp, which needed more eyeballs. “It was better to be friends than enemies at that stage,” according to Yelp cofounder Jeremy Stoppelman, who started the firm with fellow tech entrepreneur Russel Simmons.4 Plus, they depended on Google to help them generate traffic. It was, after all, the main search highway on which most consumers began their travels, thanks to the network effect that favored the largest players.
From the beginning, Yelp and other such “vertical” search engines were in a different kind of search business than Google: They dealt in extremely precise and narrow types of content rather than the “universal” searches that were Google’s bread and butter. But as local search became more and more popular, Google decided to get deeper into that business. Plus, the rise of the smartphone, as well as the growth of Amazon, was putting pressure on Google to solidify its position in the larger tech marketplace. If the company wanted to be the operating system for people’s entire lives, it couldn’t afford to cede any piece of the search market—big or small. So, a couple of years after inking the deal with Yelp, Google decided to add a “local search” feature that allowed its own users to review and rate local services, just as Yelp’s did. Worried (and justifiably so) that Google was trying to copy his business model, Stoppelman decided not to renew their contract. Two years later, Google tried to buy Yelp for a whopping $550 million, but the deal fell apart.
That was when the gloves came off. At that point, “half of all smartphone searches had some kind of local intent behind them,” says Luther Lowe, Yelp’s policy director and one of the driving forces behind its efforts to get regulators to pay attention to Google’s anticompetitive practices.5 So to capture that traffic, Google, according to Lowe, began to display its own local search results above those of Yelp and other competitors, under the guise that this would create a more user-friendly interface, and actually create “better” results.6 Of course, as reams of complicated FTC documents make quite clear, better was defined, quite simply, as whatever was better for Google.
“They had a massive incentive to exploit their network dominance at that point,” says Lowe, “by siphoning people into their own Yelp clone.” In a memo that was accidentally leaked to The Wall Street Journal during an open records request,7 FTC staff noted estimates showing that creating a more equal playing field would have led to “annualized loss of $154 million” for Google on product queries. “In certain areas where Google already had existing vertical properties, such as shopping and local, Google saw a critical need to invest further and take measures to increase user traffic to those properties,” notes the document.8 As a result, mandates came down in executive meetings to find ways to fend off competition, regardless of what it meant for search quality. (“Larry thought [Google] should get more exposure,” reads one footnote.)9 And, eventually, that meant preferred placement at the top of search results. “When Google’s algorithms deemed local websites, such as Yelp or CitySearch, relevant to a user’s query, Google automatically returned Google Local at the top of the [result list].”10
As the leaked document makes clear, this “by any means necessary” approach to competition at Google originated at the highest rungs of the company. In one section, Marissa Mayer, who at that point ran Google’s local, maps, and location services division, argues for calculating search results in such a way as to favor the company’s own services.11 In another, Google chief economist Hal Varian admits that “from an antitrust perspective, I’m happy to see [comScore] underestimate our share.”12
According to Lowe, who has spent the past several years trying to prove to regulators on three different continents that Google unfairly gave preference to its own products, “The core motivation of Google is to be the middleman of all activity on the Internet.” By using the power of their network and ecosystem, Google knew it could effectively banish competitors like Yelp from the Internet altogether. For a few years, says Lowe, “if you opened an Android phone and it had a Google Places app preinstalled and you clicked on it, you might open a Yelp review of a restaurant or whatever, but with no link back or attribution whatsoever.”
That changed in 2011 when Yelp staff started appearing in front of state attorneys general and other regulators to tell the complicated story of how Google was handicapping competitors in areas where it wanted to grow its own business. Lowe recalls in particular one conference in Hawaii, where Yelp staff delivered a presentation so compelling that several of the AGs (some of whom have since pursued cases against Google) were forced to sit up and take notice. Google also had staff in the audience, and afterward, says Lowe, “they looked like they’d seen a ghost.” Google’s treatment of Yelp content improved somewhat after the conference, but by that time, says Lowe, “the damage was done, and Google had built enough of a following and collected enough of its own reviews that it didn’t need Yelp as much anymore.” Yelp survived, but has struggled to maintain its market share, and is, of course, a fraction of the size of Google. The network effect had done its work.
The case was ultimately dismissed, which was itself unusual given that the recommendation to bring the lawsuit came from the FTC’s own bureau of competition. Many sources I’ve spoken with feel that the decision came down to Google’s lobbying power in Washington; the company not only sent its top brass to lobby politicians directly, but also funded research favorable to its cause—some of which was done by academics such as Joshua Wright, who joined the FTC shortly before the case against Google was dropped.13
Kent Walker, Google’s chief legal officer, played down those concerns, without specifically denying them. “With regard to lobbying, it’s important to remember that the FTC professional staff, three different commissions, the Bureau of Competition, the Bureau of Economics, and General Counsel’s Office all reviewed this case and all found that Google was acting primarily on behalf of consumers with regard to the innovations we were making.”14 While Walker has always struck me as a decent guy—someone who is primarily just trying to clean up whatever legal messes the company gets itself into—I didn’t really buy his line. After all, it wasn’t until after Renata Hesse, a former Google counsel, had been appointed the acting DOJ antitrust chief (and Larry Page had met with FTC officials), that the case finally went away.
The issues themselves, though, remain. In fact, there are new bipartisan calls to reopen the case. Meanwhile, the company is now battling EU regulators in similar, ongoing investigations into whether it has used anticompetitive practices to winnow out smaller companies in search. In 2018, for example, the company paid a whopping €4.3 billion antitrust penalty to the European Union for abusing its power in the mobile phone market: nearly double what it was charged a year earlier for favoring its shopping service over competitors’ in its search results.
That case centered around two British technologists turned entrepreneurs named Adam and Shivaun Raff, a husband-and-wife team who launched an online price comparison site in the United Kingdom in 2006. The two were programming nerds who’d come up with an algorithm that was very good at particular shopping queries. (For example, what airline has the cheapest flight from Glasgow to Madrid on Tuesday? Or, what’s the best vacuum cleaner with a HEPA filter?) This sounds easy, but it’s not; specific, deep-search queries like this are actually much tougher to service than more universal ones. But the Raffs had cracked the code, and within forty-eight hours of launching their site, which they named Foundem, they were flooded with traffic from shoppers looking for everything from computers to appliances.
Then the traffic stopped, pretty
much cold turkey, according to Shivaun Raff, who approached me in 2017 following an article I had written in the FT on the way in which Big Tech used the network effect to achieve monopoly power. Or, to be more specific, the traffic from Google stopped. While Foundem would rank at the top of results from other search engines—including Yahoo and MSN Search—Google would bury them way down the list. Given that research shows that users pay attention to only the top five search results,15 Foundem was effectively banished from the Web.
The way in which Google was able to effectively banish both Yelp and Foundem from the top listings in searches, and thereby effectively stop them as competitors, has major similarities to the way other “essential facilities,” like the railroads and telecommunication lines of old, were able to hold up competitors and customers alike, providing access to their networks, or not, for whatever fee they liked, or in whatever way they liked.16 In 1900, for example, six U.S. rail companies owned or controlled 90 percent of the market for anthracite coal, resulting in high prices for buyers and massive profits for the railroads—which of course made it difficult for the independent coal companies to move product over their lines.17 The problem was eventually rectified through a “commodities clause” that separated platforms and commerce. This kind of separation eventually made its way into other areas such as banking, preventing bank holding companies from competing with their own clients in various industries (though they sometimes got around such things via regulatory loopholes).
The Internet is, of course, the railroad of our times—an essential piece of public infrastructure over which much of the world’s commerce and communication is now conducted. And the parallels between what nineteenth-century regulators called “the railroad problem” and the Internet problem of today are strikingly similar. As part of the research for this book, I cracked open a slim but surprisingly readable 1878 volume entitled Railroads: Their Origins and Problems, by Charles Francis Adams,18 a former railroad executive and regulator, who lays out the rise of the railways in both Europe and the United States, and the struggle to force them to serve the public at large rather than just a handful of industrialists. In a chapter called “The Railroad Problem,” Adams writes, “As events have developed themselves, it has become apparent that the recognised laws of trade operate but imperfectly at best in regulating the use made of these modern thoroughfares by those who thus both own and monopolise them.”
You could, of course, retitle the same chapter “The Internet Problem” and have a good summary of where we are today. Amazon captures over one-third of all U.S. online retail spending, a figure that was recently downgraded by the company itself from previous estimates of around half. Amazon attributes this revised estimate to changes in how third-party sales are accounted for, but it has nonetheless aroused suspicion that the company is fiddling with its metrics to get ahead of regulators trying to make an antitrust case.
Google represents 88 percent of the U.S. search engine market, and 95 percent of all mobile searches. Two-thirds of all Americans are on Facebook, which, having bought Instagram and WhatsApp, now owns four of the top eight social media apps. All of these companies, as well as Apple, the world’s first trillion-dollar company, have come under fire for using their enormous ecosystems to give their own products and services preference and keep competitors off their networks.19 It’s a problem inherent in both owning a platform and conducting business on it.
But the monopoly argument is one that neither Google nor any other Big Tech giant is willing to admit has merit. According to Kent Walker, the failures of competitors Yelp or Foundem had little to do with anything that the search giant did. “There’s an awful lot of good econometric data that shows that the decline of various services came for a whole variety of reasons unrelated to the evolution of Google search results,” he says. Not only did Google’s own results get better, says Walker, but other giants like Amazon were rising. That may well be the case, but it’s quite telling that the search competition was increasingly among giants (indeed, Amazon has since become Google’s main rival in search) who had the power to keep smaller companies entirely out of the market.20
The Raffs, for their part, believe that Google was purposefully trying to shut them down. After traffic tailed off, the Raffs, who knew people within Google and were well-connected within the global tech community, began reaching out to people at the company, to no avail. It was no secret that Google had been working for years on its own Google Shopping service (formerly known as Froogle),21 but the Raffs didn’t immediately assume there was sabotage involved.
“For three and a half years, we went through various channels, official and unofficial. We just never got any meaningful response about what was happening,” Shivaun Raff says. Meanwhile, they began seeing similar stories on programmers’ websites, tales of entrepreneurs who’d gotten too close to the company’s core business model and were effectively put out of business. Some described the all-too-familiar experience of disappearing from search results, or claimed that Google had put pressure on their customers or clients; others recounted lawsuits that took such a financial toll that they drove smaller players out of business.22 The Raffs themselves soldiered on, eventually building up a user base of more than 2 million people, but they were draining their savings, and without the Google traffic, it was a struggle just to meet the day-to-day needs of their business. “Google is the gateway,” says Shivaun. “If you are excluded from that ecosystem, you die.”
Eventually, they decided to “stop being British,” as Shivaun put it to me, and take their case to the regulators, which is how Foundem became the lead complainant in the European Commission’s Google Search antitrust case, launched in 2009. It was led by the tough-as-nails EU competition chief, Margrethe Vestager, who eventually found against the firm in 2017. In compliance with EU law, Google was given eighteen months to figure out a way to rejigger its algorithms to eliminate bias in search. But in late 2018, the Raffs sent a letter to the commissioner, telling her that they were unpersuaded that the Google “compliance mechanism,” which depended once again on its own black box algorithmic formulas, was working. “It has now been more than a year since Google introduced its auction-based ‘remedy’ and the harm to competition, consumers, and innovation caused by Google’s illegal conduct has continued unabated,” they wrote. “We therefore respectfully urge you to commence non-compliance proceedings against Google.”23
It’s quite possible that the EU will do just that in the coming months or years. But it will likely be too late for the Raffs, who have had to turn back to consulting to make their livings, and are running Foundem more as an act of defiance—and determination to win their legal battle—than to make real money. “We’re not stopping,” says Raff. “They are going to have to change the way they do things.”
The Power of the Ecosystem
Google itself isn’t worried about competition from upstarts but from other behemoths, namely Amazon, which is the only company that is really giving the search engine a run for its money these days. Many people who are shopping for products, as opposed to looking for general information, now start their searches on Amazon, which means that advertising money is also migrating to the e-tail giant. WPP, the world’s largest purchaser of advertising (it shops on behalf of major companies globally), spent $300 million on “behalf of its clients on Amazon search ads last year,” and about 75 percent of that was pulled from budgets that would have been spent on Google-related advertising. No wonder the Googlers are looking nervously over their shoulders to see what Jeff Bezos is doing.
All of this shines an uncomfortable light on how the supposedly decentralized Internet economy has spawned a handful of ruthless oligopolies that have begun to use their power to undermine start-up growth, job creation, and labor markets. Over the past two decades, more than 75 percent of U.S. industries have seen an increase in concentration of both wealth and influence. If you compare the numbers wit
h the post–World War II period, when U.S. growth was strongest, the contrast is striking. In 1954, the top sixty companies accounted for less than 20 percent of U.S. GDP, according to the Brookings Institution. Today, the top twenty companies make up more than 20 percent.24
Why is this happening? One reason, of course, is global competition, which has put more pressure on U.S. businesses, and pushed companies away from the more equitable postwar pie sharing between workers, corporations, and local communities. Another is the shift in antitrust law. But another, less explored, reason is the network effect that goes along with the platform technology business model.25 Concentration is happening everywhere, but it’s most pronounced within the information economy. (According to the McKinsey Global Institute, industries such as tech, pharma, and finance, which are based on data and intellectual property that can be monopolized and moved anywhere around the world, are the most prone to concentration.)26
It is the economic and political challenge of our time. Jason Furman, the former head of the Council of Economic Advisers, believes concentration is creating barriers to entry in many key markets.27 Academic David Autor has linked the corporate consolidation to workers making less.28 New research from the McKinsey Global Institute has found the same, and has noted in particular the way in which technology has driven down the labor share of the overall economic pie.29 There is also evidence that a small group of “superstar” companies are pulling way ahead of others, not only in terms of profits but also productivity.30 In other words, the biggest companies, particularly in the most digitally connected parts of the economy (tech, finance, and media), are incredibly productive. Everyone else, not so much. The upshot is that economic growth as a whole has suffered.31
Don't Be Evil Page 16