The innovator’s dilemma

A new icon popped up in my Google Suite a few days ago: two sinuous interwoven blue lines with the text “Currents” accompanying them. According to the Wikipedia article, “[Google Currents] is a software developed by Google for internal enterprise communication. Originally called Google+ for G Suite, Currents is the sole remnant of Google’s defunct social network Google+, which the company shut down entirely for personal and brand use on April 2, 2019”

The article links to Google’s other products in this domain. There was Google Wave, which tried to “combine the “paper trail” of e-mail, the immediacy of IM and IRC, and the collaborative text editing of a wiki”. It was launched to 100,000 users in September 2009, and shuttered a year later, after the initial release showed little interest from the general public. There was Google Buzz, a successor to Google Wave, which intended to be a social network built on top of email, where you could share “links, photos, videos, status messages” and other content either publicly or privately. Buzz was widely panned for pre-populating “a ready-made network of friends automatically selected by the company based on the people that each user communicated with most frequently through Google’s e-mail and chat services”. In other words, as a sort of Yakov Smirnoff joke, the social network found you, rather than the other way around. And, if that wasn’t dystopian enough, a New York Times article explained, “Sergey Brin, a Google co-founder, said that by offering social communications, which have primarily been used for entertainment purposes, Buzz would bridge the gap between work and leisure.” Buzz, too, was canned unceremoniously in December 2011, less than 2 years after its initial launch. It was superseded by Google+, the direct predecessor of Google Currents. Google+ was viewed as an existential threat by Facebook, who perhaps took it more seriously than they should have: “ComScore estimated that users averaged 3.3 minutes on the site in January 2012, and 7.5 hours on Facebook”. I also failed to mention some of Google’s other forays into social networking and general communication. There was the acquisition of a Finnish microblogging startup, Jaiku (2007-2011), as well as internal projects like Google Friend Connect (2008-2012), Orkut (2004-2014), Google Reader (2005-2013), and Google Allo (2016-2019). Funnily enough, Currents itself reuses the name of a previous app (2011-2013), which provided users access to a library of magazines. That is, Google’s history of failure is so extensive that it’s starting to run out of names.

I would argue that Google’s best innovations are behind it. What are the (internally built) products that Google is best-known for? Google Search, of course (1997-). Also Google Ads/Adwords (2000-), Google Maps (2005-), Google Earth (2005-), Gmail (2004-), Google Docs (2006-), and Google Chrome (2008-). Notice anything in common? All of these innovations happened more than a decade ago, and they haven’t evolved much since. (Google Cloud Platform is one important exception to this rule; although it launched in 2008, it has had to keep up pace with its Amazon and Microsoft competitors.) And this trend isn’t uncommon. Facebook has the same, remarkably shitty-looking interface as it did 10 years ago. Microsoft’s money-making products have an even older lineage, even if they are now on the cloud.

In fact, the way that big tech “innovates” now is by acquisition, not creation. This fact is uncontroversial.

When Google wants to expand into wearables, it licenses Fossil’s technology for $40 million and acquires Fitbit for $2.1 billion. The article I cited argues, “The general reasoning is this: Google has a serious hole when it comes to wearables and it hasn’t been able to develop its own way out of it, so it needs to buy its way out.” Google has an impressive track record of acquiring successful startups, including Android ($50 million in 2005), Youtube ($1.65 billion in 2006), Doubleclick ($3.1 billion in 2008), and Nest ($3.2 billion in 2014). In some cases, Google bought a competitor that was building a product superior to it (Google Video existed simultaneously with Youtube, but the latter became much more popular); in others, Google used acquisitions to breach a segment of the market it had not previously had access to. The same is true of Facebook (Whatsapp, Instagram, and OculusVR being its notable acquisitions), Amazon (Twitch, Whole Foods, etc.), Microsoft (LinkedIn, Github), and Apple (Beats, Shazam).

It all reminds me of a post I published over 3 years ago (!), about the late Clayton Christensen, a business professor and author of The Innovator’s Dilemma. If you recall,

The question he’d begun with, twenty years ago, was: Why was success so difficult to sustain? How was it that big, rich companies, admired and emulated by everyone, could one year be at the peak of their power and, just a few years later, be struggling in the middle of the pack or just plain gone? He had initially assumed that technology moved on, and the older players couldn’t keep up. But when he began to look into it he found that this wasn’t so: as technologies grew more sophisticated, whether by small improvements or radical leaps, the established companies, with their well-funded R. & D. departments, nearly always led the way.

The more interesting question now, I think, is the opposite. Why do 5 companies make up such a large percentage of the market? Why are their fortunes seemingly immune to, well, everything, from market downturns to threats from new sexy startups? (As a point of reference, Amazon’s stock is 50% higher than its pre-Covid peak. Apple’s and Microsoft’s are 15-20% higher, and Facebook and Google’s are about the same.)

And, conversely, why is the startup ecosystem foundering? Why does there seem to be so little innovation burbling out of American capitalism? Even the Wall Street Journal had a discussion about the latter set of questions, entitled, “Six Theories on Why Fast-Growing Startups Seem to Be Disappearing”.

The answers to the two sets of questions are related. It is perhaps better to think of Apple, Microsoft, Amazon, Google, and Facebook as holding companies for a large collection of tenuously connected enterprises. (Google’s incorporation as “Alphabet” makes this concept explicit.) For each, its business strategy is two-fold. For its core products, its golden geese, it will use its vast cash reserves to buy up any possible competition, and its network and market power to eliminate anyone who can’t be bought. Witness, for instance, Facebook’s purchase of Instagram for $1 billion, or Amazon’s predatory price war against Diapers.com (which eventually convinced its founders to sell to Amazon).

First, in 2009, Amazon sent a senior vice president to have lunch with the founders of the startup behind Diapers.com, called Quidsi. He warned them that Amazon was thinking about getting into the diaper business and suggested they think about selling. As Stone tells it, this was not a friendly suggestion. It was more like the kind of offer you can’t refuse. From Businessweek:

Soon after, Quidsi noticed Amazon dropping prices up to 30 percent on diapers and other baby products. As an experiment, Quidsi executives manipulated their prices and then watched as Amazon’s website changed its prices accordingly. Amazon’s pricing bots—software that carefully monitors other companies’ prices and adjusts Amazon’s to match—were tracking Diapers.com.

Facebook also famously tried to acquire Snap, makers of Snapchat, for $3 billion. After it was spurned, Facebook essentially copied all of Snapchat’s major features in Instagram, after which Snapchat’s user growth slowed dramatically.

And, for those areas of the market where they don’t make money, but might be adjacent to their existing business, these tech “holding companies” will use targeted acquisitions to enter the space. Amazon got into groceries; Microsoft into professional networking; Facebook into virtual reality, and Google into health tech.

Regardless of whether the potential acquisition target represents a threat or an opportunity, most incentives in the startup ecosystem align towards acquisition rather than IPO. Employees and founders want to get paid cash rather than suffer the vagaries of the market. The company doing the acquiring, buffeted by plentiful cash reserves, desires a long-term investment. And politicians and judges alike have a warped sense of the terms “antitrust” and “competition”, and, more broadly, whether consumers are actually benefiting from these mergers. The absence of countervailing forces (perhaps aside from the venture capitalists themselves) leads to ever more consolidation, further entrenchment of the leading technology companies, a dearth of successful independent startups, and, overall, a sclerotic business landscape, one in which Google engineers spend their days building products that no one will use, while Google businesspeople wait to snap up companies that might. If this is the future of innovation in Silicon Valley, it looks mightily bleak.

Quoting from the Wired article I cited about Facebook,

Privately investors grumble that they won’t invest in social media companies anymore. There will never be another WhatsApp, they argue, because Facebook will buy or bury it before it’s able to get to that size. Indeed, the number of investments in internet and mobile social companies has been steadily declining since 2014, according to data provider CB Insights. The firm projects just $693 million in global investments into the category this year, less than half of 2014’s $1.4 billion. Meanwhile some founders see selling out to Facebook—no matter how early—as inevitable.

There is a further way that large tech companies have strangled startups, and one that is less remarked upon. I work at a mobile game studio. Every single one of the companies I have mentioned in this essay makes money off of us. For every in-app purchase in our game, Google takes 30%, if the purchase occurred on an Android device, while Apple takes 30%, if it occurred on one of its devices. (A genuinely innovative new email provider, Hey, ran into this problem while trying to sell subscriptions on mobile.) Facebook and Google (the former especially) are our primary channels for acquiring new customers. We spend tens of thousands of dollars on user acquisition every day. (It is also remarkable how shitty Facebook’s advertising tools are; this is a consequence of being the only game in town.) We also pay tens of thousands of dollars monthly to Amazon in cloud computing costs. Finally, although these are much smaller costs, we rely on Google’s suite of business tools (Gmail, etc.), and we pay Microsoft indirectly through LinkedIn fees and Github charges. My experience is not atypical. Large tech companies do not have to innovate because money simply flows in their direction, much like a landlord collecting a rent check every month. Indeed, some have termed the modern incarnation of capitalism a “rentier economy”. Microeconomics 101 is wrong about many things, but it is correct that taxes create deadweight losses. In this case, it is large tech companies collecting this tax, and startups—or, really, American capitalism as a whole—bearing the losses.

In a weird mirror image of right wing anxieties about taxation, I occasionally wonder: what happens when this tax, this rent, becomes so onerous that it rebounds against the entities extracting it? In recent weeks, New York landlords and real-estate agents have discovered that there isn’t an endless supply of yuppies to mine for monthly checks. I hope, although I do not expect, that the heavyweights of the NASDAQ will soon learn a similar lesson.

Advertisement