In the Footsteps of Giants (Premium)

Next week, Google will face off against the U.S. Department of Justice (DOJ) in a milestone antitrust trial with strong echoes of the past. The DOJ is involved in an antitrust trial against a technology juggernaut? Now what does that remind us of?

United States v. Microsoft is well known with this audience, of course, but it’s been a while, and some of the details may be lost to the mists of time. I wrote about this antitrust trial in my Programming Windows series, and in the resulting book, Windows Everywhere. But here’s a recap of what went down, with an eye toward comparing that case and Microsoft’s response to what’s happening today with Google.

As Microsoft drove to prominence in the 1990s, it left a mountain of corpses in its wake, companies like Apple, Borland, Lotus, Novell, WordPerfect, and many others that were either destroyed or sidelined by the software giant as consumer and business customers settled on Windows and the related software and services ecosystem provided by its maker.

We’ve all heard the term “two sides to every story,” but in this case, there were often more than two sides. In some cases, Microsoft created (or bought and then enhanced) the better product. In some, the competition made their own mistakes, sabotaging any chance they had to beat Microsoft. In some, Microsoft preempted competing product announcements by announcing that it, too, was working on something similar even when it wasn’t, and would never actually ship such a product. And in many cases, it was a combination of these and other factors.

Regardless of the ethics of its business practices and the relative ineptitude of some of its competition, Microsoft eventually came to dominate the PC industry and thus personal computing. And its rule was absolute: in this crucial new market and a key driver of the economy of the 1990s, Microsoft was in effect bigger—more powerful—than the rest of the industry combined. There was Microsoft, and then there was a shrinking collection of also-rans.

Microsoft’s domination and potential for abuse wasn’t lost on U.S. antitrust regulators. When the U.S. Federal Trade Commission (FTC) heard that Microsoft and IBM had announced plans to split the market for PC-based operating systems between Windows and OS/2 in late 1990, for example, it launched an investigation into what it feared was an anti-competitive dividing of a market. What it found was much worse: Microsoft was using a wide range of anti-competitive business practices to harm competitors of all kinds, including IBM. And so in 1993, it finally voted whether to formally charge the company with sweeping antitrust charges. A vote that failed by a 3-to-2 margin, thanks only to the recusal of one commissioner who claimed an unstated conflict of interest. Microsoft was off the hook.

Until it wasn’t: the FTC handed its mountains of boxes of paperwork evidence over to the U.S. Department of Justice (DOJ), which then forced Microsoft into a consent decree that addressed some of the company’s abuses: it would limit its product bundling practices, in which it used a dominant product like Windows to prop up a new or unpopular product, and it agreed to stop charging PC makers for MS-DOS and Windows when they shipped PCs without either software. But one vaguely worded phrase in the agreement gave Microsoft the wiggle room it needed to continue its abuses: it could add new features to its products as long as they were truly integrated and not just added on to harm competitors.

This was, of course, the strategy Microsoft employed when cofounder and CEO Bill Gates finally realized the Internet threat in 1995: he directed every product group in the company to pivot to address the Internet. Key among those products, of course, was Windows, where Gates feared that the new market for web browsers that was led by Netscape Navigator could usurp Microsoft’s control of the industry. And so Microsoft acquired the code for a web browser and built Internet Explorer (IE). Which it didn’t release just as a standalone product, but instead integrated deeply into Windows.

The question, of course, was why.

On the one hand, Microsoft can and did argue that web browser integration in Windows would benefit consumers. Not only was IE free in a world in which Netscape charged for its product, but its integration in Windows would extend the platform’s capabilities, making the Internet a natural extension of the desktop-based capabilities that had preceded it.

But that isn’t why Microsoft integrated IE into Windows. As was proven via a mountain of evidence in the coming antitrust trial, Microsoft specifically to harm Netscape and prevent this new competitor from seizing part of its customer base. “Without unification we will lose to Netscape,” Gates wrote in one memo. It wasn’t enough to have the better product: Microsoft needed to kneecap Netscape to win.

Microsoft’s strategies did kneecap Netscape, as it turns out, and so the smaller company complained to the DOJ, which launched an investigation of Microsoft’s business practices, resulting in the historic United States v. Microsoft trial, a damning Finding of Facts from Judge Thomas Penfield Jackson, and the unexpected decision that he intended to break up Microsoft to punish it for its abuses. The key to this entire episode was the integration bit. Gates argued that “The future requires integration.” But all the DOJ—and Judge Jackson—could see was abuse, with Microsoft hobbling competitors big and small and working to ensure that Windows was closed to third-party integration. What Gates was really saying was that “the future requires integration of Microsoft software only,” they retorted.

During this entire time, Microsoft adopted a belligerent stance, painting the DOJ as out of step with the fast-moving personal technology market. And this belligerence started at the top, with Gates slouching and stammering his way through a bewildering several hours of testimony in a videotaped deposition in which he argued over basic terms and lied about not remembering important events. Not helping matters, multiple Microsoft executives were caught lying on the stand during the trial, as more and more internal memos came to light, contradicting them again and again.

“We are not ‘write once, run anywhere’ kind of guys,” Microsoft’s Steve Ballmer said at one point, alluding to the slogan for Sun’s competing Java system, another platform the firm feared could undermine Windows. (This is now ironic given Microsoft’s current stance of openness and “meeting the customer where they are.”)

Ultimately, United States v. Microsoft ended with a whimper. Microsoft lost, Netscape was sold to America Online, and IE eventually beat Netscape Navigator, not just because it was bundled but because it became the better product in a short period of time. This paradox is a key debate in antitrust circles, where it becomes hard to identify the key factor in any competitive landscape. Did IE beat Navigator because it was better? Because it was free, which both helped consumers and hurt the competition? Because it was more integrated with Windows? Because Netscape made missteps of its own? Or was it simply a little bit of each? And if so, can one even weigh whether Microsoft’s business practices tilted the playing field?

None of it mattered. Microsoft reversed the integration of IE and Windows because it was technically flawed to begin with, and it never fully integrated IE technology into the Windows shell because users didn’t want it. IE beat Netscape, yes, but then Microsoft got complacent, and future competitors like Mozilla Firefox, which rose out of Netscape, and then Google Chrome, went on to dominate IE.

The biggest change, however, was the unforeseen one. Microsoft, hobbled by its battles with the DOJ—and then later with the EU’s European Commission—lost talent and willpower as the 1990s turned into the 2000s. Gates handed over the CEO reins to Ballmer, struggled to make sense of .NET and his new Chief Software Architect role, and then became just one of many dozens of key executives to fade into the background and leave the company. Microsoft, which had vowed to never take its eye off the ball and become the next IBM did just that. And it weathered a “lost decade,” as I call it, in which its relative power and ability to forestall competition fell and then disappeared.

It was during this time that new competitors like Amazon, the renewed Apple, Facebook, and Google created new personal technology products, competitors that the Microsoft of the 1990s would have handily destroyed or sidelined. The wounded and directionless Microsoft of this era helped to create the personal computing market as we know it today, one that is dominated by mobile devices and online services, and not by PCs and Windows. Microsoft, bigger than it ever was, is now just one of many players in a more heterogeneous world. And among those competitors, all of which are now flexing their market power muscles in ways reminiscent of the Microsoft of old, it is Microsoft, oddly, that seems the least abusive.

And it is in the world that Google, which dominates the online services market because of its control of Internet search and its resulting ad-based revenues windfall, now finally faces its own antitrust reckoning in the U.S., and as with United States v. Microsoft, it is the DOJ that has brought this about. This time, of course, it’s United States v. Google.

Are these stories even similar?  Google, after all, started off a research project—I originally wrote “an almost research project-like entity,” but it was literally a research project—by two Stanford PhD students, Larry Page and Sergey Brin, in 1996. So it’s the most California thing ever, in contrast to the cold opportunity positioning of Microsoft, with the pair believing that they could use an algorithm to better determine the relevance of web pages as the result of Internet searches. This system, called page ranking, led to the creation of Google the company, and, as with Apple, it was funded by a single outside investor initially.

Google would have been a moderately successful business under the initial plan, which involved selling ads associated with search keywords, and we can speculate that Microsoft or some other company might have swallowed it up at some point, making millionaires out of the founders. But its hiring of Eric Schmidt as CEO in 2001 was pivotal: Schmidt had worked at Sun Microsystems and Novell, two Microsoft victims, previously, and he brought an anti-Microsoft bias to Google just as it was experiencing explosive growth. And then Google went public—in a very Googly way—making its founders billionaires. And Google started snapping up companies with its newfound wealth, entering new markets.  YouTube, DoubleClick, and Waze were key among its earliest acquisitions, with each triggering a Google dominance of the respective markets.

But none are more valuable than DoubleClick, the dominant online advertising service. (And not coincidentally a company that Microsoft had wanted to acquire to thwart Google but failed.) The combination of DoubleClick and Google Search resulted in the Google we know today: in its most recent financial quarter, 79 percent of Google’s $74.6 billion in revenues came from advertising. Google is, in effect, just an advertising company, relegating most of its paid personal technology products and services to largely noncompetitive hobbies. Without advertising, Google would be a non-event. And it would certainly not be of interest to antitrust regulators.

Google is also unique in that its business practices have received the most scrutiny, by far, among the most dominant Big Tech firms, which makes it similar to the Microsoft of the past. It faces multiple antitrust cases in the EU, alone, but it is also the only Big Tech firm facing an actual antitrust trial in the U.S. A trial that starts next week because Google, like Microsoft before it, has refused to settle or accommodate regulatory concerns in any way. Google’s argument, like Microsoft’s before it, is that it simply builds a better product.

Here, again, we must face the central paradox of antitrust. That is, this statement can be true—Google Search is objectively the best Internet search engine, and by a wide margin—but that doesn’t obviate the fact that it could also be harming consumers and/or competitors by using illegal business practices to artificially tilt the playing field in its favor, forestalling any future innovation that might have otherwise occurred.

I will again argue that Google’s relative size—compared to Microsoft in the late 1990s—also raises unique challenges. When Microsoft dominated personal technology, the addressable market was small, perhaps tens of millions of a small number of hundreds of millions of users. But Google, like Apple, Amazon, and Facebook, today has a market of several billions. 3 billion people use Android every single day, and Google bragged at its most recent I/O event that six of its offerings have over 2 billion users each while another 15 have over 500 million users each.

And then there’s Search. Google didn’t provide new Search usage numbers at I/O this past year, perhaps related to the coming antitrust trial, but it dominates the market with a 92 percent usage share. It’s so big that the other players listed, each with a very low single-digit usage share, may as well not exist at all. It’s like comparing Microsoft Office usage to that of LibreOffice. Which most have never even heard of.

But Google Search today faces an uncertain potential new competitor in OpenAI, just as Windows in 1995 faced Netscape Navigator. Here, Google has another advantage, however: unlike the Microsoft of that era, which was caught with its pants down, unaware of how the world was changing, Google has always invested in AI. Indeed, Google was only one of two Big Tech firms, the other being Microsoft, making major and regular investments in AI, and then sprinkling the capabilities that resulted into its products and services. Google is at least somewhat uniquely positioned to handle this coming transition away from pure search to what I’ll call AI-based search. (Even Microsoft was not: it had to invest heavily in OpenAI to be competitive.)

The rise of AI will certainly be a cornerstone of Google’s argument, mirroring Microsoft in the 1990s, that new competitors are always waiting in the wings and ready to challenge its dominance. That its market power has done nothing to stop this from happening, and that it will need to compete with OpenAI and others and win on the merits, or not. In other words, this market is healthy and could be shaken up at any time.

The DOJ, meanwhile, will also dust off its Microsoft-era playbook and argue that Google’s dominance, no, Google’s monopoly in Search was arrived at largely organically, which is legal, but that Google then maintained and extended this monopoly illegally, harming both consumers and competitors. For example, it infamously pays Apple an estimated $12 billion each year to keep its search engine the default on the iPhone, because doing otherwise could lead to a competitor stealing share. (Microsoft, you may recall, argued this year in the wake of its Bing AI announcements, that every one percent of usage share in Internet search was worth billions of dollars each year.) And since Apple executives will take the stand during the Google trial, it is remotely possible that we’ll find out exactly how much that iPhone deal is really worth.

And to be clear, Google has abused its market power in Search. It was found to use its insider access to what’s popular in trending services to direct results to its own services over those of rivals, and it has even created new services of its own as needed as a result of these trends. (Amazon does the same thing with its online store searches.)

One of the more interesting things about the Microsoft story is how it rebounded from its belligerent stance of the 1990s in more recent years. As I wrote in Nice Guy (Premium), Microsoft senior vice president and general counsel Brad Smith was hired in 2001 to get the software giant out of its antitrust messes, and he convinced its leadership “to soften its approach and make peace with regulators at home and abroad. He was hired, and he did just that, and he’s led Microsoft’s legal efforts ever since.” His conciliatory approach is what makes the Activision Blizzard acquisition even remotely possible. And he’s employed a similar strategy in dealing with the EU’s problems with Teams/Microsoft 365 bundling.

Smith’s counterpart at Google is Kent Walker. But Walker is no Smith, and he is not interested in meeting the DOJ or any other regulator halfway. Instead, Walker is presenting a throwback defense that can come off with the same belligerent undertones as the Microsoft from the 1990s.

“American law should be about promoting benefits for consumers,” he told The New York Times this week, echoing that one side of the antitrust paradox coin. “If we move away from that and make it harder for companies to provide great goods and services for consumers, that’s going to be bad for everyone.”

And there it is: in this shortsighted viewpoint, Google’s behavioral issues and illegal monopoly maintenance—which, again, work to prevent outside innovation that could benefit users—should be ignored because Google today makes the better product, which, um, benefits users.

That doesn’t mean that the counterpoint outweighs his argument. Instead, both views can be true. The question is whether they are, in this case. And if so, which is weighed more heavily as a matter of law.

There are too many unknowns here to even debate this issue. Google will, of course, argue that it doesn’t harm competition at all, and that AI is here and could change everything. Hell, Google could conceivably still settle this case at the final hour, showing a maturity that Microsoft struggled to attain for many years. But it does seem that this case will go to trial, and I’m fascinated to see what we learn as a result.

It all starts next Tuesday.