Microsoft’s Monopoly Hangover

Microsoft announced something very impressive last week: revenue for the company’s 2017 fiscal year (which ended June 30) increased 5% year-over-year. That may not seem particularly meaningful until you realize 2016 was only the second year in the company’s history that revenue declined; the first included the worst economic slowdown since the Great Depression:

Moreover, all indications are that growth will continue, defeating the presumption that tech companies that start to decline do so inexorably. The most famous example that said inexorable decline need not be inevitable is IBM, which, in the early 90s, found itself in far more dire straits than Microsoft, only to recover under the leadership of Lou Gerstner:

Microsoft’s earnings report isn’t the only thing that has made me think of IBM lately; Two weeks ago, at Mirosoft’s annual partner conference, CEO Satya Nadella introduced a new offering called Microsoft 365. Nadella said:

Microsoft 365 is a fundamental departure in how we think about product creation. This is the coming together of the best of Office 365, Windows 10, Enterprise Mobility and Security…

We have decided that the time has come for us as a company and us as an ecosystem to talk about this in the terms that customers can get the most value. We want to bring these products together as an integrated solution. A complete solution that has got AI infused in it with intelligence, whether it is intelligence that is helping end users be more productive and creative and teamwork, or intelligence in security. It’s that complete solution for intelligent teamwork and security that we want to bring about with Microsoft 365.

A cynical take is that this is typical Microsoft, cribbing a successful naming scheme (‘365’) to rebrand a SKU that Microsoft actually announced a year ago. That’s true! A slightly more generous take is that Microsoft 365 is the latest implementation of the company’s decades-old bundling strategy, and, well, that’s true too!

The way Nadella framed the announcement though — associating customer value with integration — that is straight from Gerstner’s IBM playbook.

The IBM and Microsoft Monopolies

When Gerstner signed on as IBM CEO in the spring of 1993, the company had just recorded the biggest annual loss in American corporate history: -$4.97 billion. In his memoir about the turnaround he led at IBM, Who Says Elephants Can’t Dance?, Gerstner noted that 1993 was going just as badly:

At the end of May I saw April’s [numbers] and they were sobering. Profit had declined another $400 million, for a total decline of $800 million for the first four months. Mainframe sales had dropped 43 percent during the same four months. Other large IBM businesses—software, maintenance, and financing—were all dependent, for the most part, on mainframe sales and, thus, were declining as well.

Gerstner expanded on this point in various sections of his book:

Despite the fact that IBM, then and now, was regarded as a complex company with thousands of products…IBM was a one-product company—a mainframe company—with an array of multibillion-dollar businesses attached to that single franchise…It didn’t take a Harvard MBA or a McKinsey consultant to understand that the fate of the mainframe was the fate of IBM, and, at the time, both were sinking like stones.

IBM’s mainframe business was being hammered on two fronts: Unix-based alternatives offered modular lower-cost alternatives for back-end operations, while PCs were taking over many of the jobs mainframes used to do — and, in the long run, threatening to take over the data center itself. IBM was not only stuck with a product that was too expensive for a market that was simultaneously shrinking in size, but also an entire organization predicated on that product’s dominance.

This was Microsoft a few decades later: the company loved to brag about its stable of billion dollar businesses, but in truth they were all components of one business — Windows. Everything Microsoft built from servers to productivity applications was premised on the assumption that the vast majority of computing devices were running Windows, leaving the company completely out of sorts when the iPhone and Android created and captured the smartphone market.

The truth is that both companies were victims of their own monopolistic success: Windows, like the System/360 before it, was a platform that enabled Microsoft to make money in all directions. Both companies made money on the device itself and by selling many of the most important apps (and in the case of Microsoft, back-room services) that ran on it. There was no need to distinguish between a vertical strategy, in which apps and services served to differentiate the device, or a horizontal one, in which the device served to provide access to apps and services. When you are a monopoly, the answer to strategic choices can always be “Yes.”

That, though, is why it is so interesting to think about what happens — and the problems that arise — when the monopoly ends.

Post-Monopoly Problem One: Nature

The great thing about a monopoly is that a company can do anything, because there is no competition; the bad thing is that when the monopoly is finished the company is still capable of doing anything at a mediocre level, but nothing at a high one because it has become fat and lazy. To put it another way, for a former monopoly “big” is the only truly differentiated asset.

This was Gerstner’s key insight when it came to mapping out IBM’s future:

I am not sure that in 1993 I or anyone else would have started out to create an IBM. But, given IBM’s scale and broad-based capabilities, and the trajectories of the information technology industry, it would have been insane to destroy its unique competitive advantage and turn IBM into a group of individual component suppliers more minnows in an ocean.

In the big April customer meeting at Chantilly and in my other customer meetings, CIOs made it very clear that the last thing in the world they needed was one more disk drive company, one more operating system company, one more PC company. They also made it clear that our ability to execute against an integrator strategy was nearly bankrupt and that much had to be done before IBM could provide a kind of value that we were not providing at the time—but which they believed only IBM had a shot at delivering: genuine problem solving, the ability to apply complex technologies to solve business challenges, and integration.

So keeping IBM together was the first strategic decision, and, I believe, the most important decision I ever made—not just at IBM, but in my entire business career. I didn’t know then exactly how we were going to deliver on the potential of that unified enterprise, but I knew that if IBM could serve as the foremost integrator of technologies, we’d be delivering extraordinary value.

In Gerstner’s vision, only IBM had the breadth to deliver solutions instead of products; the next challenge would be changing the business model.

Post-Monopoly Problem Two: Business Model

The natural inclination for former monopolies, at least if Microsoft and IBM are any indication, is to stick with the monopoly-era business model. That meant doubling down on the device (or OS, as it were).

The problem with this approach is twofold:

  • First, as I just noted, the nature of the company is set: being big — which in this case means offering services to everyone — is much easier to accomplish than being better, a critical factor in selling a differentiated device in a competitive market.
  • Second, as long as the business-model is device-centric, there is a risk in destroying the services component of the business. In Microsoft’s case, that meant holding Office for iPad back to prop up Windows, for example, or building Azure (née Windows Azure) around Windows Server. IBM, in far more dire straights, was, as Gerstner noted, close to splitting up the company so that individual divisions could sell their respective devices on their own without corporate overhead.

The reality is that while changing business models is hard, for both Microsoft and IBM it was necessary to preserve what strengths they still had. This is why defenders of former-CEO Steve Ballmer miss the point when pointing out that Microsoft Azure and Office 365, the keys to Microsoft’s renewed growth, both got started under his watch. Look again at Gerstner’s account of IBM:

If you were to take a snapshot of IBM’s array of businesses in 1993 and another in 2002, you would at first see very few changes. Ten years ago we were in servers, software, services, PCs, storage, semiconductors, printers, and financing. We are still in those businesses today…

My point is that all of the assets that the company needed to succeed were in place. But in every case—hardware, technology, software, even services—all of these capabilities were part of a business model that had fallen wildly out of step with marketplace realities.

This is why I don’t give Ballmer too much credit for Office 365 and Azure: the products of Microsoft’s future were there, but the Windows-centric business model was constricting every part of the company to an ever-shrinking share of the overall market; Nadella’s greatest success has been taking off that straitjacket.1

Post-Monopoly Problem Three: Culture

Four years ago, while announcing a company-wide reorganization (that I thought was a bad idea), Ballmer wrote a memo called One Microsoft. This was the key paragraph:

We will reshape how we interact with our customers, developers and key innovation partners, delivering a more coherent message and family of product offerings. The evangelism and business development team will drive partners across our integrated strategy and its execution. Our marketing, advertising and all our customer interaction will be designed to reflect one company with integrated approaches to our consumer and business marketplaces.

I wrote in Services, Not Devices:

The crux of the problem is in that paragraph: no one is asking Microsoft to design its “customer interaction” to “reflect one company.” Customers are asking Microsoft to help them solve their problems and get their jobs done, not to make them Microsoft-only customers. The solipsism is remarkable.

The solipsism, at least if IBM was any indication, was also inevitable. Gerstner writes:

When there’s little competitive threat, when high profit margins and a commanding market position are assumed, then the economic and market forces that other companies have to live or die by simply don’t apply. In that environment, what would you expect to happen? The company and its people lose touch with external realities, because what’s happening in the marketplace is essentially irrelevant to the success of the company…

This hermetically sealed quality—an institutional viewpoint that anything important started inside the company—was, I believe, the root cause of many of our problems. To appreciate how widespread the dysfunction was, I need to describe briefly some of its manifestations. They included a general disinterest in customer needs, accompanied by a preoccupation with internal politics. There was general permission to stop projects dead in their tracks, a bureaucratic infrastructure that defended turf instead of promoting collaboration, and a management class that presided rather than acted. IBM even had a language all its own.

Sounds familiar!

Comic from Bonkers World

Gerstner’s response was to restructure IBM, change the company’s promotion and compensation policies, and most importantly, push IBM to better understand customers and then leverage its size to offer services they actually needed:

Our bet was this: Over the next decade, customers would increasingly value companies that could provide solutions— solutions that integrated technology from various suppliers and, more important, integrated technology into the processes of an enterprise. We bet that the historical preoccupations with chip speeds, software versions, proprietary systems, and the like would wane, and that over time the information technology industry would be services-led, not technology-led.

This is why the Microsoft 365 announcement and Nadella’s talk of integration is so interesting, and IBM plays a role in this story as well.

IBM’s Cloud Miss

I’ve previously written about how IBM, specifically Sam Palmisano, who succeeded Gerstner as CEO, missed the cloud. Remarking on Palmisan’s declaration that “You can’t do what we’re doing in a cloud” I wrote:

Something that is interesting about most cloud solutions is that few are really doing anything new. Rather cloud service providers are simply taking operations that were formerly done on premise and moving them to a cloud that is available for any enterprise to use. And, as Palmisano realized, the inherent lack of customization in such a model means that most cloud services are on a feature-by-feature basis inferior to on-premise software.

The reality, though, is that the businesses IBM served — and the entire reason IBM had a market — didn’t buy customized technological solutions to make themselves feel good about themselves; they bought them because they helped them accomplish their business objectives. Gerstner’s key insight was that many companies had a problem that only IBM could solve, not that customized solutions were the end-all be-all. And so, as universally provided cloud services slowly but surely became good-enough, IBM no longer had a monopoly on problem solving.

To put it bluntly, enterprises don’t need a systems integrator for their data center if they no longer have a data center. Once again IBM is stuck competing for a shrinking market, which is why the company’s revenue has now declined for 21 straight quarters.2

Microsoft’s Cloud Opportunity

Still, the fact that enterprises no longer have data centers doesn’t mean integration is no longer valuable; rather, the locus of needed integration has shifted to the cloud as well. The average enterprise customer uses 20~30 apps, data is often scattered on and off premise, or stuck in email or personal accounts, and while IT departments may be happy to no longer upgrade servers, managing identity and security across all of these services and on a whole host of new devices far more likely to be used outside a company’s intranet calls for the same sort of integrator Gerstner wanted IBM to be.

This seems to be the long-term goal of Microsoft 365. Microsoft said in a blog post:

[Microsoft 365] represents a fundamental shift in how we will design, build and go to market to address our customers’ needs for a modern workplace. The workplace is transforming—from changing employee expectations, to more diverse and globally distributed teams, to an increasingly complex threat landscape. From these trends, we are seeing a new culture of work emerging. Our customers are telling us they are looking to empower their people with innovative technology to embrace this modern culture of work.

With more than 100 million commercial monthly active users of Office 365, and more than 500 million Windows 10 devices in use, Microsoft is in a unique position to help companies empower their employees, unlocking business growth and innovation…

Microsoft 365 Enterprise:

  • Unlocks creativity by enabling people to work naturally with ink, voice and touch, all backed by tools that utilize AI and machine learning.
  • Provides the broadest and deepest set of apps and services with a universal toolkit for teamwork, giving people flexibility and choice in how they connect, share and communicate.
  • Simplifies IT by unifying management across users, devices, apps and services.
  • Helps safeguard customer data, company data and intellectual property with built-in, intelligent security.

Wait, inking?

Here’s the big concern I have about the Microsoft 365 rollout, and Microsoft generally: Nadella and team deserve plaudits for working through the first two post-monopoly problems. Microsoft has embraced its bigness and focused on services, and has the business model to match (although, it should be noted that it was Ballmer who was responsible for shifting most of Microsoft’s enterprise business to a subscription model years ago). That’s great!

I’m troubled, though, that I just articulated what I think is the Microsoft 365 strategy — or what it should be — far more clearly than either Nadella or Kirk Koenigsbauer, the corporate vice-president for the Office team that wrote this blog post. Indeed, Gerstner articulated the strategy best of all, and he wasn’t even talking about Microsoft or the cloud!

Then again, I’m not entirely sure a focus on cloud integration is Microsoft’s strategy after all: maybe the cynical take — that Microsoft is just stealing a successful name for yet another enterprise licensing bundle — is closer to the truth. It is striking that the primary reason Microsoft gives for Microsoft 365 is that it already has a lot of users.

Stepping back even further, Nadella loves to say “Our customers tell us” or some derivative thereof, but an actual articulation of customer use cases is consistently missing from his presentations. This keynote was not dissimilar to Nadella’s Build keynote, which featured a full 30 minutes of theory about the future of computing, that, while fascinating, seemed much more like a justification for Microsoft’s continued relevance as opposed to an articulation of demonstrated customer needs.

Can Culture Change?

The most bittersweet paragraph in Who Says Elephants Can’t Dance? is the final one:

I was always an outsider. But that was my job. I know Sam Palmisano has an opportunity to make the connections to the past as I could never do. His challenge will be to make them without going backward; to know that the centrifugal forces that drove IBM to be inward-looking and self-absorbed still lie powerful in the company. Continuing to drive change while building on the best (and only the best) of the past is the ultimate description of the job of Chief Executive Officer, International Business Machines Corporation.

Palmisano completely failed the challenge: what was the aforementioned reliance on IBM’s seemingly impregnable position as a systems integrator and dismissal of the cloud anything but the result of being “inward-looking and self-absorbed”? The same point applies to Palmisano’s obsession with profit-per-share: customers, Gerstner’s obsession, were totally forgotten.

That is why Gerstner’s IBM should be a potential inspiration to Microsoft, but Palmisano’s (and current CEO Ginni Rometty, who has hewed far more closely to Palmisano’s example than Gerstner’s) IBM a warning: culture is a curse, and for better or worse, a company can recover but never be fully cured.

  1. Update: A few folks have written in to note that I’m being a bit harsh on Ballmer. After all, not only did he start Azure, he fired the well-respected and successful head of the Server and Tools Business, Bob Muglia, because he wasn’t moving fast enough in the cloud. Muglia’s replacement? Satya Nadella []
  2. By the way, Gerstner predicted the public cloud in the first appendix of his book, which was published in 2003, four years before AWS was launched:

    Put all of this together—the emergence of large-scale computing grids, the development of autonomic technologies that will allow these systems to be more self-managing, and the proliferation of computing devices into the very fabric of life and business—and it suggests one more major development in the history of the IT industry. This one will change the way IT companies take their products to market. It will change who they sell to and who the customer considers its “supplier.” This development is what some have called “utility” computing.

    The essential idea is that very soon enterprises will get their information technology in much the same way they get water or electric power. They don’t now own a waterworks or power plant, and soon they’ll no longer have to buy, house, and maintain any aspect of a traditional computing environment: The processing, the storage, the applications, the systems management, and the security will all be provided over the Net as a service—on demand.

    The value proposition to customers is compelling: fewer assets; converting fixed costs to variable costs; access to unlimited computing resources on an as-needed basis; and the chance to shed the headaches of technology cycles, upgrades, maintenance, integration, and management.

    Also, in a post-September 11, 2001, world in which there’s much greater urgency about the security of information and systems, on-demand computing would provide access to an ultra-secure infrastructure and the ability to draw on systems that are dispersed— creating a new level of immunity from a natural disaster or an event that could wipe out a traditional, centralized data center.

    IBM misses him. []

Publishers and the Pursuit of the Past

Editor’s Note: This article was originally published as a Daily Update for Stratechery members. To receive the Daily Update please subscribe here.

David Chavern, the president and CEO of the News Media Alliance, a trade association representing 2,000 North American newspapers, has an op-ed in the Wall Street Journal that I’m going to take a longer-than-usual excerpt from:

The rapid growth of digital connectivity has pushed demand for information to unprecedented heights. Never in history have so many people consumed so much news. This is a boon for democracy. Although reporting is often an irritant to those in power, high-quality news and analysis is essential to any political system that depends on giving citizens the facts so they can draw their own conclusions.

The problem is that today’s internet distribution systems distort the flow of economic value derived from good reporting. Google and Facebook dominate web traffic and online ad income. Together, they account for more than 70% of the $73 billion spent each year on digital advertising, and they eat up most of the growth. Nearly 80% of all online referral traffic comes from Google and Facebook. This is an immensely profitable business. The net income of Google’s parent company, Alphabet, was $19 billion last year. Facebook’s was $10 billion.

But the two digital giants don’t employ reporters: They don’t dig through public records to uncover corruption, send correspondents into war zones, or attend last night’s game to get the highlights. They expect an economically squeezed news industry to do that costly work for them.

The only way publishers can address this inexorable threat is by banding together. If they open a unified front to negotiate with Google and Facebook—pushing for stronger intellectual-property protections, better support for subscription models and a fair share of revenue and data—they could build a more sustainable future for the news business.

But antitrust laws make such coordination perilous. These laws, intended to prevent monopolies, are having the unintended effect of preserving and protecting Google and Facebook’s dominant position. The digital giants benefit from legal precedent against collective action that has a chilling effect on publishers. Yet each newspaper or magazine on its own has only limited negotiating power.

Chavern’s solution is twofold: one, that regulators should do a better job of enforcing antitrust laws against Google and Facebook, and two, that Congress should grant publishers a safe harbor to negotiate collectively with Google and Facebook.

That a safe harbor is necessary to negotiate collectively was made apparent in the Apple ebooks case; while reasonable people can disagree as to whether or not Apple was rightly found to be guilty, no one disputes that the publishers colluded to raise prices, a per se violation of the Sherman Antitrust Act.1

Aggregation Theory Redux

That publishers feel the need to pursue the same strategy is driven by the reality of Aggregation Theory. To quickly recap, in a world of abundance, as long as the aggregator controls demand, suppliers have no choice but to commoditize themselves according to the strictures of the aggregator, which is another way of saying that suppliers have no choice but to engage in perfect competition. This is extremely ruinous for publishers in particular:

  • In a perfect market, the price of an undifferentiated commodity will be its marginal cost
  • The marginal cost is the cost to produce one more item (as opposed to the total cost, which includes the fixed costs necessary to create the item; those fixed costs, though, are already spent whether or not an additional item is created)2
  • The marginal cost of a digital item is zero, which means in a perfect market the inevitable price of a digital item is zero

In short, aggregators are market makers, and the markets they make — thanks to the aforementioned lack of distribution costs and transactions costs — are pretty darn close to perfect, particularly in the case of purely digital goods. The problem is that publishers have huge fixed costs, even if you ignore old world infrastructure like printing presses and delivery trucks; specifically, publishers have to pay salaries, but those costs, by virtue of being fixed, not marginal, have zero impact on a publication’s pricing power in a perfect market.3

The end result is dire for newspapers in particular: in commodity markets the winning companies have superior cost structures, which means they can sustainably sell at the market-clearing price; newspapers, though, by virtue of being built for a world of print, will never have the cost structure or mentality to succeed in the long run.

Thus this solution: Chavern and the big publishers want permission from Congress to escape the perfect competition fostered by Aggregation Theory via collusion. The theory seems to be that, were the 2,000 newspapers party to this proposal able to present a unified front, they could force concessions from Google and Facebook that would make their businesses viable.

News Versus Advertising

There’s just one problem with this analysis — and that problem extends to Chavern’s proposal as a whole: it’s based on a myth. Specifically, newspapers have never succeeded by selling news, a point I made explicitly in The Local News Business Model:

By owning printing presses and delivery trucks (and thanks to the low marginal cost of printing extra pages), newspapers were the primary outlet for advertising that didn’t work on (or couldn’t afford) TV or radio — and there was a lot of it. Maximizing advertising, though, meant maximizing the potential audience, which meant offering all kinds of different types of content in volume: thus the mashup of wildly disparate content listed above, all focused on quantity over quality. And then, having achieved the most readership and the ability to expand to fit it all, the biggest newspaper could squeeze out its competitors.

Too many newspaper advocates utterly and completely fail to understand this; the truth is that newspapers made money in the past not by providing societal value, but by having quasi-monopolistic control of print advertising in their geographic area; the societal value was a bonus. Thus, when Chavern complains that “today’s internet distribution systems distort the flow of economic value derived from good reporting”, he is in fact conflating societal value with economic value; the latter does not exist and has never existed.

This failure to understand the past leads to a misdiagnosis of the present: Google and Facebook are not profitable because they took newspapers’ reporting, they are profitable because they took their advertising. Moreover, the utility of both platforms is so great that even if all newspaper content were magically removed — which has been tried in Europe — the only thing that would change is that said newspapers would lose even more revenue as they lost traffic.

This is why this solution is so misplaced: newspapers no longer have a monopoly on advertising, can never compete with the Internet when it comes to bundling content, and news remains both valuable to society and, for the same reasons, worthless economically (reaching lots of people is inversely correlated to extracting value, and facts — both real and fake ones — spread for free).

A Better Solution for Publishers

I just said that “good reporting” has never had economic value; this wasn’t quite right. Rather, the articles that result from “good reporting” don’t have economic value: once published on the Internet they have zero marginal cost in a world of perfect competition. This is why publishers have to shift their mindset about their product and their market.

Specifically, publishers should be selling “good reporting”, that is, the commitment to the regular production of content that the buyer would like to see. This is a slight distinction but a critical one: successful subscription products do not sell content but rather the production of the content, and that production, unlike the articles themselves, can be differentiated and sold as a scarce product. That, though, means knowing who the buyer is: it’s not advertisers, but rather readers.

Moreover, this approach addresses the Facebook and Google problem: the issue with Aggregation Theory from a supplier perspective is that the aggregator owns the consumer relationship; however, because Facebook and Google are advertising companies, they are not even competing in the subscription market. They, like advertisers, only care about content that has already been produced, not how it came to be.

In fact, this is the single most ridiculous part of this proposal: one of the issues Chavern wishes to collectively bargain with Facebook and Google about is “better support for subscription models”. In other words, Chavern wishes to bring in Facebook and Google as an aggregator in the one market — subscriptions — where newspapers actually have a viable business model.

It’s easy to envision how this could play out: Google and Facebook set up subscription offerings for publishers, eventually create the bundle of the future, and, by virtue of owning the consumer, skim off most of the profits, leaving publishers desperately pursuing page views to get their minuscule share of revenue. Sound familiar?

The fundamental issue is this: there is a business model that works for publishers, but it requires a dramatic shift in mindset and the long hard slog of building a business. That means understanding what customers want, building a product that appeals to them, reaching them, moving them down a sales funnel, and retaining them. What it does not mean is the suffocating sense of entitlement and delusion that underlies not just this proposal but the majority of commentary from newspapers themselves that expects someone — anyone! — to give journalists money simply because they are important.

That’s the thing: journalism is important. It is so important that the sooner publishers let go of a long-gone world where publications earned advertising simply by existing, and actually build publications that can not just survive but thrive on the Internet, the better off society will be. And, in that fight, this move by the News Media Alliance is actively hurting the cause.

  1. Apple was ruled to have been a co-conspirator which meant they were per se guilty; the company argued their relationship with the publishers was a vertical one, which would have led to a rule of reason analysis that I suspect would have exonerated the company. I wrote more about the case here []
  2. Technically all costs, at least in the long-run, are marginal costs; however, in the short-run, the determination of whether or not to produce one more item is based purely on the cost of that item alone; a useful overview of the difference is here []
  3. Salaries are fixed costs in terms of producing one more article; they are much closer to variable costs, though, than something like a printing press []

Ends, Means, and Antitrust

The European Commission levied a record €2.42 billion ($2.73 billion) fine on Google yesterday for having “abused its market dominance as a search engine by giving an illegal advantage to another Google product, its comparison shopping service.” Commissioner Margrethe Vestager said in a press release announcing the decision:

“Google has come up with many innovative products and services that have made a difference to our lives. That’s a good thing. But Google’s strategy for its comparison shopping service wasn’t just about attracting customers by making its product better than those of its rivals. Instead, Google abused its market dominance as a search engine by promoting its own comparison shopping service in its search results, and demoting those of competitors.

“What Google has done is illegal under EU antitrust rules. It denied other companies the chance to compete on the merits and to innovate. And most importantly, it denied European consumers a genuine choice of services and the full benefits of innovation.”

It is tempting when these decisions come down to start with the ends: specifically, does the outcome in question agree with one’s pre-existing views on such matters as regulation generally, antitrust specifically, and even nationalism (or continentalism, as it were)? The means matter, though, especially in this decision: there are three meaningful questions in this case that cut to the heart of antitrust regulation of digital companies:

  • What is a digital monopoly?
  • What is the standard for determining illegal behavior?
  • What constitutes a competitive product?

The European Commission’s decision was impressive on some of these questions, and very problematic on others; all sides of the antitrust debate should be wary of standing in resolute opposition or support.

What is a Digital Monopoly?

This is perhaps the most consequential aspect of this case, and I think the European Commission got it exactly right. Last year in Antitrust and Aggregation I explained why the unique dynamics of the Internet push towards dominant players that look very different from the monopolies of the past:

Aggregation Theory is about how business works in a world with zero distribution costs and zero transaction costs; consumers are attracted to an aggregator through the delivery of a superior experience, which attracts modular suppliers, which improves the experience and thus attracts more consumers, and thus more suppliers in the aforementioned virtuous cycle. It is a phenomenon seen across industries including search (Google and web pages), feeds (Facebook and content), shopping (Amazon and retail goods), video (Netflix/YouTube and content creators), transportation (Uber/Didi and drivers), and lodging (Airbnb and rooms, Booking/Expedia and hotels).

The first key antitrust implication of Aggregation Theory is that, thanks to these virtuous cycles, the big get bigger; indeed, all things being equal the equilibrium state in a market covered by Aggregation Theory is monopoly: one aggregator that has captured all of the consumers and all of the suppliers. This monopoly, though, is a lot different than the monopolies of yesteryear: aggregators aren’t limiting consumer choice by controlling supply (like oil) or distribution (like railroads) or infrastructure (like telephone wires); rather, consumers are self-selecting onto the Aggregator’s platform because it’s a better experience.

That bit about self-selection is the most obvious reason to critique this decision: how can Google have a monopoly when users — over 90% of the population in most European countries, according to the European Commission’s Factsheet — could choose to use another search engine simply by typing a URL, or opening another app? The Factsheet has the answer:

There are also high barriers to entry in these markets, in part because of network effects: the more consumers use a search engine, the more attractive it becomes to advertisers. The profits generated can then be used to attract even more consumers. Similarly, the data a search engine gathers about consumers can in turn be used to improve results.

This is exactly right, and in my view a real breakthrough in antitrust regulation. In the physical world, limited by scarcity, economic power comes from controlling supply; in the digital world, overwhelmed by abundance, economic power comes from controlling demand, and that control stems from a virtuous cycle that, for the reasons explained in the excerpt above, accrues to the dominant player in a space. In other words, to note that end users could go elsewhere is to ignore the reality that users are not dummies, and that network effects are the foundation of digital monopolies.

What is the Standard for Determining Illegal Behavior?

The United States and European Union have, at least since the Reagan Administration, differed on this point: the U.S. is primarily concerned with consumer welfare, and the primary proxy is price. In other words, as long as prices do not increase — or even better, decrease — there is, by definition, no illegal behavior.

The European Commission, on the other hand, is explicitly focused on competition: monopolistic behavior is presumed to be illegal if it restricts competitors which, in the theoretical long run, hurts consumers by restricting innovation. From the Factsheet:

Market dominance is, as such, not illegal under EU antitrust rules. However, dominant companies have a special responsibility not to abuse their powerful market position by restricting competition, either in the market where they are dominant or in separate markets. Otherwise, there would be a risk that a company once dominant in one market (even if this resulted from competition on the merits) would be able to use this market power to cement/further expand its dominance, or leverage it into separate markets…

As a result of Google’s illegal practices and the distortions to competition, Google’s comparison shopping service has made significant market share gains at the expense of rivals. This has deprived European consumers of the benefits of competition on the merits, namely genuine choice and innovation.

The European Commission approach, relative to the U.S. when it comes to determining illegality, has both pluses and minuses: the good thing is that it is an approach that is actually applicable to digital markets, particularly those monetized by advertising. Given the fact that Google is free for consumers, it is basically all but impossible for the company to be found guilty of antitrust behavior by the U.S., as the FTC determined a few years ago. The truth is that proxies are always problematic, including price, and there is no better example than the absurdity of the U.S. Justice Department successfully suing Apple for building a competitor to Amazon, the actual e-book monopolist.

On the other hand, isn’t consumer welfare the entire point? Sure, a narrow focus on price is perhaps a bad proxy, but if dominant services are winning by being better — which is my argument in Aggregation Theory — why should regulators busy themselves with demanding worse alternatives be given the right to succeed?

This is a point where many of those focused on the antitrust ends go wrong: in their crusade against big companies, they fail to grapple with the reality that on the Internet being big comes from being the best, leaving their arguments vulnerable to the critique that they are, in fact, anti-consumer, or at a minimum, anti-competence. To put it another way, in contrast to the previous point, regulators are treating people like dummies, assuming they can’t figure out how to find a competitive service, when in fact the truth is they don’t want to.

What Constitutes a Competitive Product?

This is by far the most concerning part of the European Commission’s decision, for two reasons.

First, if I search for a specific product, why would I not want to be shown that specific product? It frankly seems bizarre to argue that I would prefer to see links to shopping comparison sites; if that is what I wanted I would search for “Shopping Comparison Sites”, a request that Google is more than happy to fulfill:

Screen Shot 2017-06-28 at 6.40.22 PM

The European Commission is effectively arguing that Google is wrong by virtue of fulfilling my search request explicitly; apparently they should read my mind and serve up an answer (a shopping comparison site) that is in fact different from what I am requesting (a product)?

The second reason is even more problematic: “Google Shopping” is not actually a search product; it is an ad placement:

Screen Shot 2017-06-28 at 6.56.13 PM

You can certainly argue that the tiny “Sponsored” label is bordering on dishonesty, but the fact remains that Google is being explicit about the fact that Google Shopping is a glorified ad unit. Does the European Commission honestly have a problem with that? The entire point of search advertising is to have the opportunity to put a result that might not otherwise rank in front of a user who has demonstrated intent.

The implications of saying this is monopolistic behavior goes to the very heart of Google’s business model: should Google not be allowed to sell advertising against search results for fear that it is ruining competition? Take travel sites: why shouldn’t Priceline sue Google for featuring ads for hotel booking sites above its own results? Why should Google be able to make any money at all?

This is the aspect of the European Commission’s decision that I have the biggest problem with. I agree that Google has a monopoly in search, but as the Commission itself notes that is not a crime; the reality of this ruling, though, is that making any money off that monopoly apparently is. And, by extension, those that blindly support this decision are agreeing that products that succeed by being better for users ought not be able to make money.

Long-time readers of Stratechery know that I have shifted my position on antitrust over time. At the end of the day, I tend to agree with the European Union that competition is an end worth pursuing in and of itself, and that antitrust regulation is fundamentally different from the sort of red tape that limits entrepreneurship. Indeed, it is directly opposed: red tape regulation entrenches incumbents and limits new entrants, while antitrust regulation limits incumbents and enables new entrants.

Moreover, I believe that Google has been a bad actor: the company’s scraping of content from Yelp, TripAdvisor, and Amazon was clearly anticompetitive, and a much better example of illegally favoring Google’s results over superior competition. And, more broadly, I started writing about the Google-Facebook duopoly in online advertising earlier than most; I have been building the case that that is where the problems of monopoly might most clearly be seen.

In short, I agree with the ends as far as the European Commission’s ruling is concerned: Google is a monopoly, and they act badly. In this case, though, I simply can not tolerate the means: I can not see any compelling case in which consumer welfare is better served by offering answers they didn’t actually ask for, and attacking an ad unit feels a lot more like an attempt to hurt a company as opposed to helping competition.

More broadly, antitrust advocates have to appreciate that, when it comes to digital monopolies, there is a very fine line to walk between opposing products that are better for consumers and promoting competition: I do think competition is ultimately pro-consumer, but simply presuming that “big” is bad when “big” comes from a superior customer experience is little more than a shortcut to political irrelevance.1

  1. In other words, wait for the Android case! More on the long-term repurcussions for Google tomorrow in the Daily Update []

Amazon’s New Customer

Back in 2006, when the iPhone was a mere rumor, Palm CEO Ed Colligan was asked if he was worried:

“We’ve learned and struggled for a few years here figuring out how to make a decent phone,” he said. “PC guys are not going to just figure this out. They’re not going to just walk in.” What if Steve Jobs’ company did bring an iPod phone to market? Well, it would probably use WiFi technology and could be distributed through the Apple stores and not the carriers like Verizon or Cingular, Colligan theorized.

I was reminded of this quote after Amazon announced an agreement to buy Whole Foods for $13.7 billion; after all, it was only two years ago that Whole Foods founder and CEO John Mackey predicted that groceries would be Amazon’s Waterloo. And while Colligan’s prediction was far worse — Apple simply left Palm in the dust, unable to compete — it is Mackey who has to call Amazon founder and CEO Jeff Bezos, the Napoleon of this little morality play, boss.

The similarities go deeper, though: both Colligan and Mackey made the same analytical mistakes: they mis-understood their opponent’s goals, strategies, and tactics. This is particularly easy to grok in the case of Colligan and the iPhone: Apple’s goal was not to build a phone but to build an even more personal computer; their strategy was not to add on functionality to a phone but to reduce the phone to an app; and their tactics were not to duplicate the carriers but to leverage their connection with customers to gain concessions from them.

Mackey’s misunderstanding was more subtle, and more profound: while the iPhone may be the most successful product of all time, Amazon and Jeff Bezos have their sights set on being the most dominant company of all time. Start there, and this purchase makes all kinds of sense.

Amazon’s Goal

If you don’t understand a company’s goals, how can you know what its strategies and tactics will be? Unfortunately, many companies, particularly the most ambitious, aren’t as explicit as you might like. In the case of Amazon, the company stated in its 1997 S-1:’s objective is to be the leading online retailer of information-based products and services, with an initial focus on books.

Even if you picked up on the fact that books were only step one (which most people at the time did not), it was hard to imagine just how all-encompassing would soon become; within a few years Amazon’s updated mission statement reflected the reality of the company’s e-commerce ambitions:

Our vision is to be earth’s most customer centric company; to build a place where people can come to find and discover anything they might want to buy online.

“Anything they might want to buy online” was pretty broad; the advent of Amazon Web Services a few years later showed it wasn’t broad enough, and a few years ago Amazon reduced its stated goal to just that first clause: We seek to be Earth’s most customer-centric company. There are no more bounds, and I don’t think that is an accident. As I put it on a podcast a few months ago, Amazon’s goal is to take a cut of all economic activity.

This, then, is the mistake Mackey made: while he rightly understood that Amazon was going to do everything possible to win in groceries — the category accounts for about 20% of consumer spending — he presumed that the effort would be limited to e-commerce. E-commerce, though, is a tactic; indeed, when it comes to Amazon’s current approach, it doesn’t even rise to strategy.

Amazon’s Strategy

As you might expect, given a goal as audacious as “taking a cut of all economic activity”, Amazon has several different strategies. The key to the enterprise is AWS: if it is better to build an Internet-enabled business on the public cloud, and if all businesses will soon be Internet-enabled businesses, it follows that AWS is well-placed to take a cut of all business activity.

On the consumer side the key is Prime. While Amazon has long pursued a dominant strategy in retail — superior cost and superior selection — it is difficult to build sustainable differentiation on these factors alone. After all, another retailer is only a click away.

This, though, is the brilliance of Prime: thanks to its reliability and convenience (two days shipping, sometimes faster!), plus human fallibility when it comes to considering sunk costs (you’ve already paid $99!), why even bother looking anywhere else? With Prime Amazon has created a powerful moat around consumer goods that does not depend on simply having the lowest price, because Prime customers don’t even bother to check.

This, though, is why groceries is a strategic hole: not only is it the largest retail category, it is the most persistent opportunity for other retailers to gain access to Prime members and remind them there are alternatives. That is why Amazon has been so determined in the space: AmazonFresh launched a decade ago, and unlike other Amazon experiments, has continued to receive funding along with other rumored initiatives like convenience store and grocery pick-ups. Amazon simply hasn’t been able to figure out the right tactics.

Amazon’s Tactics

To understand why groceries are such a challenge look at how they differ from books, Amazon’s first product:

  • There are far more books than can ever fit in a physical store, which means an e-commerce site can win on selection; in comparison, there simply aren’t that many grocery items (a typical grocery store will have between 30,000 and 50,000 SKUs)
  • When you order a book, you know exactly what you are getting: a book from Amazon is the same as a book from a local bookstore; groceries, on the other hand, can vary in quality not just store-to-store but, particularly in the case of perishable goods, day-to-day and item-to-item
  • Books can be stored in a centralized warehouse indefinitely; perishable groceries can only be stored for a limited amount of time and degrade in quality during transit

As Mackey surely understood, this meant that AmazonFresh was at a cost disadvantage to physical grocers as well: in order to be competitive AmazonFresh needed to stock a lot of perishable items; however, as long as AmazonFresh was not operating at meaningful scale a huge number of those perishable items would spoil. And, given the inherent local nature of groceries, scale needed to be achieved not on a national basis but a city one.

Groceries is a fundamentally different problem that needs a fundamentally different solution; what is so brilliant about this deal, though, is that it solves the problem in a fundamentally Amazonian way.

The First-And-Best Customer

Last year in The Amazon Tax I explained how the different parts of the company — like AWS and Prime — were on a conceptual level more similar than you might think, and that said concepts were rooted in the very structure of Amazon itself. The best example is AWS, which offered server functionality as “primitives”, giving maximum flexibility for developers to build on top of:1

The “primitives” model modularized Amazon’s infrastructure, effectively transforming raw data center components into storage, computing, databases, etc. which could be used on an ad-hoc basis not only by Amazon’s internal teams but also outside developers:

stratechery Year One - 274

This AWS layer in the middle has several key characteristics:

  • AWS has massive fixed costs but benefits tremendously from economies of scale
  • The cost to build AWS was justified because the first and best customer is Amazon’s e-commerce business
  • AWS’s focus on “primitives” meant it could be sold as-is to developers beyond Amazon, increasing the returns to scale and, by extension, deepening AWS’ moat

This last point was a win-win: developers would have access to enterprise-level computing resources with zero up-front investment; Amazon, meanwhile, would get that much more scale for a set of products for which they would be the first and best customer.

As I detailed in that article, this exact same framework applies to

Prime is a super experience with superior prices and superior selection, and it too feeds into a scale play. The result is a business that looks like this:

stratechery Year One - 275

That is, of course, the same structure as AWS — and it shares similar characteristics:

  • E-commerce distribution has massive fixed costs but benefits tremendously from economies of scale
  • The cost to build-out Amazon’s fulfillment centers was justified because the first and best customer is Amazon’s e-commerce business
  • That last bullet point may seem odd, but in fact 40% of Amazon’s sales (on a unit basis) are sold by 3rd-party merchants; most of these merchants leverage Fulfilled-by-Amazon, which means their goods are stored in Amazon’s fulfillment centers and covered by Prime. This increases the return to scale for Amazon’s fulfillment centers, increases the value of Prime, and deepens Amazon’s moat

As I noted in that piece, you can see the outline of similar efforts in logistics: Amazon is building out a delivery network with itself as the first-and-best customer; in the long run it seems obvious said logistics services will be exposed as a platform.

This, though, is what was missing from Amazon’s grocery efforts: there was no first-and-best customer. Absent that, and given all the limitations of groceries, AmazonFresh was doomed to be eternally sub-scale.

Whole Foods: Customer, not Retailer

This is the key to understanding the purchase of Whole Foods: from the outside it may seem that Amazon is buying a retailer. The truth, though, is that Amazon is buying a customer — the first-and-best customer that will instantly bring its grocery efforts to scale.

Today, all of the logistics that go into a Whole Foods store are for the purpose of stocking physical shelves: the entire operation is integrated. What I expect Amazon to do over the next few years is transform the Whole Foods supply chain into a service architecture based on primitives: meat, fruit, vegetables, baked goods, non-perishables (Whole Foods’ outsized reliance on store brands is something that I’m sure was very attractive to Amazon). What will make this massive investment worth it, though, is that there will be a guaranteed customer: Whole Foods Markets.

stratechery Year One - 270

In the long run, physical grocery stores will be only one of Amazon Grocery Services’ customers: obviously a home delivery service will be another, and it will be far more efficient than a company like Instacart trying to layer on top of Whole Foods’ current integrated model.

I suspect Amazon’s ambitions stretch further, though: Amazon Grocery Services will be well-placed to start supplying restaurants too, gaining Amazon access to another big cut of economic activity. It is the AWS model, which is to say it is the Amazon model, but like AWS, the key to profitability is having a first-and-best customer able to utilize the massive investment necessary to build the service out in the first place.

I said at the beginning that Mackey misunderstood Amazon’s goals, strategies, and tactics, and while that is true, the bigger error was in misunderstanding Amazon itself: unlike Whole Foods Amazon has no particular desire to be a grocer, and contrary to conventional wisdom the company is not even a retailer. At its core Amazon is a services provider enabled — and protected — by scale.

Indeed, to the extent Waterloo is a valid analogy, Amazon is much more akin to the British Empire, and there is now one less obstacle to sitting astride all aspects of the economy.

  1. To be clear, AWS was not about selling extra capacity; it was new capability, and Amazon itself has slowly transitioned over time (as I understand it is still a hybrid) []

Podcasts, Analytics, and Centralization

Tucked into the last day of WWDC was a session on podcasting, and it contained some big news for the burgeoning industry. Before getting into the specific announcements, though, the session itself is worth a bit of analysis, particularly the opening from Apple Podcasts Business Manager James Boggs:

First we want to talk for a moment about how we think about modern podcasts. Long-form and audio. We get excited about episodic content that entertains, informs, and inspires. We get excited and many of our users have gotten excited too.

I went on to transcribe the next 500 or so words of Boggs’s presentation, which included various statistics on downloads, catalog size, and reach; a listing of Apple “partners” organized by media and broadcast organizations, public media, and independents; and even started in on Boggs’s review/promotion of individual podcasts like “Up and Vanished” and “Masters of Scale” before I realized Boggs was never going to actually say “how [Apple] think[s] about modern podcasts.” I won’t make you read the transcript — take my word when I say that there was nothing there.

Still, that itself was telling; Boggs’s presentation perfectly reflects the state of podcasting today: Apple is an essential piece, even as they really don’t have anything to do with what is going on (but naturally, are happy to take credit).

A Brief History of Podcasts

Probably the first modern podcast was created by Dave Winer in 2003, although it wasn’t called a “podcast”: that was coined by Ben Hammersley in 2004, and the inspiration was Apple’s iPod. Still, while the media had a name, the “industry”, such that it was, was very much the wild west: a scattering of podcast creators, podcatchers (software for downloading the podcasts), and podcast listeners, finding each other by word-of-mouth.

stratechery Year One - 267

A year later Apple made the move that cemented their current position as the accidental gorilla of the industry: iTunes 4.9 included support for podcasts and, crucially, the iTunes Music Store created a directory (Apple did not — and still does not — host the podcast files themselves). The landscape of podcasting was completely transformed:

stratechery Year One - 268

Centralization occurs in industry after industry for a reason: everyone benefits, at least in the short term. Start with the users: before iTunes 4.9 subscribing and listening to a podcast was a multi-step process, and most of those steps were so obscure as to be effective barriers for all but the most committed of listeners.

  • Find a podcast
  • Get a podcatcher
  • Copy the URL of the podcast feed into the podcatcher
  • Copy over the audio file from the podcatcher into iTunes
  • Sync the audio file to an iPod
  • Listen to the podcast
  • Delete the podcast from the iPod the next time you sync’d

iTunes 4.9 made this far simpler:

  • Find a podcast in the iTunes Store and click ‘Subscribe’
  • Sync your iPod
  • Listen

Recounting this simplification may seem pedantic, but there is a point: this was the most important improvement for podcast creators as well. Yes, the iTunes Music Store offered an important new discovery mechanism, but it was the dramatic improvement to the user experience that, for the vast majority of would-be listeners, made podcasts even worth discovering in the first place. Centralized platforms win because they make things easier for the user; producers willingly follow.

Interestingly, though, beyond that initial release, which was clearly geared towards selling more iPods, Apple largely left the market alone, with one important exception: in 2012 the company released a standalone Podcasts app for iOS in the App Store, and in 2014 the app was built-in to iOS 8. At that point the power of defaults did its job: according to the IAB Podcast Ad Metrics Guidelines released last fall, the Apple Podcast App accounts for around 50% of all podcast players across all operating systems (iTunes is a further ~10%).1

The Business of Podcasting

It’s not clear when the first podcast advertisement was recorded; a decent guess is Episode 67 of This Week in Tech, recorded on September 3, 2006 (Topic: “Does the Google CEO’s place on Apple’s board presage a Sun merger?”). The sponsor was surprisingly familiar — Visa (“Safer, better money. Life takes Visa.”), and Dell joined a week later.

Over the ensuing years, though, the typical podcast sponsor was a bit less of a name brand — unless, of course, you were a regular podcast listener, in which case you quickly knew the brands by heart: Squarespace, Audible, Casper Mattress, Blue Apron, and recent favorite MeUndies (because who doesn’t want to hear a host-read endorsement for underwear!). Companies like Visa or Dell were few and far between: a study by FiveThirtyEight suggested brand advertisers were less than five percent of ad reads.

The reason is quite straightforward: for podcasts there is neither data nor scale. The data part is obvious: while podcasters can (self-)report download numbers, no one knows whether or not a podcast is played, or if the ads are skipped. The scale bit is more subtle: podcasts are both too small and too big. They are too small in that it is difficult to buy ads at scale (and there is virtually no quality control, even with centralized ad sellers like Midroll); they are too large in that the audience, which may be located anywhere in the world listening at any time, is impossible to survey in order to measure ad effectiveness.

That is why the vast majority of podcast advertisers are actually quite similar: nearly all are transaction-initiated subscription-based services. The “transaction-initiated” bit means that there is a discrete point at which the customer can indicate where they heard about the product, usually through a special URL, while the “subscription-based” part means these products are evaluating their marketing spend relative to expected lifetime value. In other words, the only products that find podcast advertising worthwhile are those that expect to convert a listener in a measurable way and make a significant amount of money off of them, justifying the hassle.2

The result is an industry that, from a monetization perspective, looks a lot like podcasting before iTunes 4.9; there are small businesses to be built, but the industry as a whole is stunted.

Apple Podcast Analytics

This is the context for what Apple actually announced. Jason Snell had a good summary at Six Colors:

New extensions to Apple’s podcast feed specification will allow podcasts to define individual seasons and explain whether an episode is a teaser, a full episode, or bonus content. These extensions will be read by the Podcast app and used to present a podcast in a richer way than the current, more linear, approach…

The other big news out of today’s session is for podcasters (and presumably for podcast advertisers): Apple is opening up in-episode analytics of podcasts. For the most part, podcasters only really know when an episode’s MP3 file is downloaded. Beyond that, we can’t really tell if anyone listens to an episode, or how long they listen—only the apps know for sure. Apple said today that it will be using (anonymized) data from the app to show podcasters how many people are listening and where in the app people are stopping or skipping. This has the potential to dramatically change our perception of how many people really listen to a show, and how many people skip ads, as well as how long a podcast can run before people just give up.

The new extensions are a nice addition, and a way in which Apple can enhance the user experience to the benefit of everyone. As you might expect, though, I’m particularly interested in the news about analytics. Problem solved, right? Or is it problem caused? What happens when advertisers realize that everyone is skipping their ads?

Advertisers: Not Idiots

In fact, I expect these analytics to have minimal impact, at least in the short run. For one, every indication is that analytics will only be available to the podcast publishers, although certainly advertisers will push to have them shared.3 More pertinently, though, all of the current podcast advertisers know exactly what they are getting: X amount of podcast ads results in Y number of conversions that result in Z amount of lifetime value.

Indeed, contrary to what many folks seem to believe, advertisers, whether they leverage podcasts, Facebook, Google, or old school formats like radio or TV, are not idiots blindly throwing money over a wall in the vague hopes that it will drive revenue, ever susceptible to being shocked, shocked! that their ads are being ignored. Particularly in the case of digital formats advertisers are quite sophisticated, basing advertising decisions off of well-known ROI calculations. That is certainly the case with podcasts: knowing to a higher degree of precision how many ads are skipped doesn’t change the calculation for the current crop of podcast advertisers in the slightest.

What more data does do is open the door to more varied types of advertisers beyond the subscription services that dominate the space. Brand advertisers, in particular, are more worried about reaching a guaranteed number of potential customers than they are tracking directly to conversion, and Apple’s analytics will help podcasters tell a more convincing story in that regard.

In truth, though, Apple’s proposed analytics aren’t nearly enough: advertisers still won’t know who they are reaching or where they are located, and while brand advertisers may not have the expectation of tracking-to-purchase no one wants to throw money to the wind either. The problem of surveying effectively to measure things like brand lift is as acute as ever, and it simply isn’t worth the trouble to do a bunch of relatively small media buys with zero quality control.

Apple’s Opportunity

This, though, is why Apple’s centralized role is so intriguing. Remember, the web was thought to be a wasteland for advertising until Google provided a centralized point that aggregated users and could be sold to advertisers. Similarly, mobile was thought to monetize even worse than the (desktop) web until Facebook provided a centralized point that aggregated users and could be sold to advertisers. I expect a similar dynamic in podcasts: the industry will remain the province of web hosting and underwear absent centralization and aggregation, and the only entity that can accomplish that is Apple.

One can envision the broad outlines of what the business for a centralized aggregator for podcasts might look like:

  • The centralized aggregator would likely offer hosting to podcast creators, not only to secure the user experience and get better analytics (including on downloads through other apps) but also to dynamically insert advertisements. Those advertisements would also be available to smaller podcasts that are currently not worth the effort to advertisers.
  • Advertisers would get their own dashboard for those analytics and, more importantly, the opportunity to buy ads at far greater scale across a large enough audience to make it worth their while. Ideally, at least from their perspective, they would actually be able to target their advertising buys as well.
  • Users would, at least in theory, benefit from a far broader array of content made possible by the growth in revenue for the industry broadly.

There are already companies trying to do just this: I wrote about E.W. Scripps’ Midroll and their acquisition of podcast player Stitcher last year. The problem is that Stitcher only has around 5% of listeners, and it is the ownership of users/listeners, not producers/podcast from which true market power derives. Apple has that ownership, and thus that power; the question is will they use it?

Surely the safe bet is “no”. iAd, Apple’s previous effort at building an advertising business, failed spectacularly, and Apple’s anti-advertising rhetoric has only deepened since then. That’s a problem not only in terms of image but culture: Apple seems highly unlikely to be willing to put in the effort necessary to build a real advertising business, and given how small such a business might be even in the best-case scenario relative to the rest of the company, that’s understandable.4

To be sure, should Apple decline to seize this opportunity it will be celebrated by many, particularly those doing well in the current ecosystem. Podcasting is definitely more open than not, with no real gatekeepers in terms of either distribution or monetization. That, though, is why the money is so small: gatekeepers are moneymakers, and while podcasts may continue to grow, it is by no means inevitable that, absent a more active Apple, the money will follow.

Disclosure: Exponent, the podcast I host with James Allworth, does have a (single) sponsor; the revenue from this sponsorship makes up a very small percentage of Stratechery’s overall revenue and does not impact the views in this article

  1. For what it’s worth, Exponent has a much different profile: Apple Podcasts has about 13% share, while Overcast leads the way with 26% share, followed by (surprisingly!) Mobile Safari with 23% []
  2. This shows why Casper mattresses are the exception that proves the rule: mattresses are not a subscription service, but they are much more expensive than most products bought online, which achieves the same effect as far as lifetime value is concerned []
  3. I’m less worried about the fact other podcast players may not offer similar analytics: the Apple Podcast app will be used as a proxy, although this may hurt podcasts that have a smaller share of downloads via the Apple Podcast app (as total listeners may be undercounted absent similar analytics from other apps) []
  4. It’s Google’s challenge in building a real hardware business in reverse []

Apple’s Strengths and Weaknesses

The San Jose location of WWDC, Apple’s annual developer conference, felt a bit odd, but Apple sought to strike a familiar tone: the artwork on and around the San Jose McEnery Convention Center featured a top-down view of humans, and a familiar message:


The idea of Apple existing at the intersection of technology and liberal arts was central to the late Steve Jobs’ conception of Apple and, without question, a critical factor when it came to Apple’s success: at a time when technology was becoming accessible to consumers and their daily lives Apple created products — one product, really, the iPhone — that appealed to consumers not only because of what it did but how it did it.

That said, it was telling that this artwork and the sentiment it signified were not referenced in the keynote itself; after a humorous skit about a world without apps, Tim Cook delivered platitudes about how Apple and its developers were on a “collective mission to change the world”, and immediately launched into what he said were six important announcements. It was not dissimilar to Sundar Pichai’s opening at Google I/O: when the announcements that matter are grounded on the realities of a company’s core competencies and position in the market, vision can feel extraneous.

Apple’s Announcements

Cook’s first four announcements spoke to those core capabilities and the position they afford Apple (or don’t, as the case may be) in the markets in which it competes:

tvOS: It was generous of Cook to give tvOS top billing: the only announcement of note was the upcoming availability of Amazon Prime Video on Apple TV. That itself is a reminder of Apple’s diminished position in the space: winning in TV is not about hardware or software, much less the integration of the two, but rather content. The brevity of this announcement — there wasn’t even the traditional executive hand-off — spoke to Apple’s status as an also-ran.

watchOS: This garnered more time, and the headline feature was the Siri watch face. The watch face, which implied a broadening of Siri’s brand from voice to context-based general assistant, seeks to anticipate and deliver the information you need when you need it. The model is Google Now; the difference is that Siri is now housed in an attractive and increasingly popular watch that works natively with an iPhone, while the equivalent Google service requires not simply a different watch but a different phone entirely. It is a testament to Apple’s biggest advantage: thanks to the iPhone the company already owns the “best” customers,1 frequently rendering moot Google’s superiority in managing information.

macOS: This actually encompassed two of Cook’s six promised announcements;2 the separation of MacOS and Mac computers was, I suspect, born of Apple’s desire to convince developers and other pro users that the company was not abandoning their favorite platform. Moreover, the addition of hardware announcements, after several years in which WWDC was software only, resulted in a very different feel to this keynote: after all, hardware is exciting, even if, in the long run, it is software that actually matters. That feeling, though, goes to the very core of what Apple sells: superior hardware differentiated — and thus sold at a handsome margin — by exclusive software.

As is always the case with the modern incarnation of Apple, though, the announcements that truly mattered centered around iOS.

iOS 11

The iOS-related announcements, despite being only one of Cook’s “Big Six”, could have been their own keynote; given the importance of mobile generally and iOS specifically that would have been more than justified. Taken as a whole the iOS segment in particular highlighted what Apple does best, what it struggles with, and what reasons there are to be both optimistic and pessimistic about the company’s fortunes in the long run.

Strength: Defaults

Controlling one of the two dominant mobile operating systems grants Apple the power of defaults. That means iMessage is both an iPhone lock-in and a channel to introduce new services like person-to-person Apple Pay. Siri can be accessed both via voice and the home button, and, just similar to the WatchOS update, is increasingly integrated throughout the operating system. Photos and Maps are used by the majority of iPhone customers, even if alternatives offer superior functionality.

Weakness: Limited Reach

At the same time, iMessage will never reach the dominance of a service like WeChat because it is limited to Apple’s own platforms — as it should be! iMessage is the canonical example of how strengths and weaknesses are two sides of the same coin: it is iMessage’s exclusivity that allows it to be a lock-in, and it is that same exclusivity that limits the standalone value.

Strength: Hardware Integration

Peppered throughout Apple’s presentation were seemingly small features like new compression algorithms that depend on Apple controlling everything from Messages to the camera to the processor that makes it all work. The most impressive example was ARKit: in one fell swoop Apple leaped ahead of the rest of the industry in the race to realize the promise of augmented reality. The contrast to Facebook was striking: while the social network is seeking to leverage its control of content distribution to lure developers to build on Facebook’s “camera”, Apple is not only offering the same opportunity (the results of which can, of course, be shared on Facebook or Instagram), but also delivering a superior set of APIs that, by virtue of being part of that vertical stack, are both more powerful and accessible than anything a 3rd-party application can deliver.

Weakness: Services

While Apple bragged about Siri’s natural language capabilities and alluded to a limited number of new “intents” that can be leveraged by apps, it is not an accident that there were no slides about accuracy, speed, or developer support: Siri is well behind the competition in all three. More fundamentally, all of Apple’s services are intrinsically limited by the fact that they exist to sell Apple hardware: those services, and the teams that work on them, will never be the most important people in the company, and their development will be constrained by the culture of Apple itself.

Strength: Privacy

Apple not only touted its privacy credentials, it also showed off new features to actively limit things like autoplaying videos and advertising networks that follow you across sites. As a user both are very welcome; strategically, both features follow from the fact that Apple makes money on its hardware, while companies like Google, Facebook, and other online businesses rely on advertising and the collection of data.

Weakness: Data

Collecting data is useful for more than advertising, though. Here Google is the obvious counter: certainly the search company wants to better target advertisements, but the benefits gained from data go far beyond overt monetization. It is data that drives Google’s superior machine learning capabilities and the customer-friendly features that follow in apps like Google Photos. Interestingly, Apple made moves in this direction, syncing things like facial recognition data and Messages across devices, favoring convenience over a very slight increase in the risk to privacy. To be clear, the data will still be encrypted, both in transit and at rest, but that is my point: encryption means that Apple cannot leverage the data it will now store to make its services better.

Strength: The App Store

The strategic role of 3rd-party apps has shifted over time: once a differentiator for iOS, Android has largely reached parity, and apps are now table stakes. They are also a big moneymaker: Apple has been pushing the narrative on Wall Street that it is a services company, fueled by the $30 billion the company has collected from app sales and especially in-app purchases in free-to-play games; 30% of that total has come in the last year alone. Make no mistake, this is a compelling narrative: iPhone growth may be slowing in the face of saturation and elongated update cycles, but that only means there is that large of a base from which to earn App Store revenue.

Weakness: Developer Economics

The success of free-to-play games and the associated in-app purchases has come at a cost, specifically, management blindness to the fact that the rest of the developer ecosystem isn’t nearly as healthy, and that the App Store is no longer a differentiator from Android. The fundamental problem remains that for productivity apps in particular it is necessary to monetize your best customers over time; Apple has improved the situation, particularly with the addition of subscription pricing and de facto trials (basically, starting a subscription at $0), but hasn’t made any moves to support trials or upgrade pricing for paid apps, despite the fact that is the proven successful model for productivity applications on the Mac. I have long argued that bad developer economics is the fundamental reason that the iPad hasn’t fulfilled its potential; yesterday’s iPad software enhancements were welcome and will help, but I suspect letting developers set their own business models would be even more transformative.

Strength and Weakness: Business Model

This point is part and parcel with all of the above: Apple’s strengths derive from the fact it sells software-differentiated hardware for a significant margin, which allows for exclusive apps and services set as defaults, deep integration from chipset to API, a focus on privacy, and total control of the developer ecosystem. And, on the flipside, Apple only reaches a segment of the market, is less incentivized and capable of delivering superior services, has less data, and can afford to take developers for granted.


Apple’s final announcement encapsulated all of these tensions. The long-rumored competitor to Amazon Echo and Google Home was, fascinatingly, framed as anything but. Cook began the unveiling by referencing Apple’s longtime focus on music, and indeed, the first several minutes of the HomePod introduction were entirely about its quality as a speaker. It was, in my estimation, an incredibly smart approach: if you are losing the game, as Siri is to Alexa and Google, best to change the rules, and having heard the HomePod, its sound quality is significantly better than the Amazon Echo (and, one can safely assume, Google Home). Moreover, the ability to link multiple HomePods together is bad news for Sonos in particular (the HomePod sounded significantly better than the Sonos Play 3 as well).

Of course, superior sound quality is what you would expect from a significantly more expensive speaker: the HomePod costs $350, while the Sonos Play 3 is $300, and the Amazon Echo is $150. From Apple’s perspective, though, a high price is a feature, not a bug: remember, the company has a hardware-based business model, which means there needs to be room for a meaningful margin. The Echo is the opposite: because it is a hardware means to the service ends that is Amazon, it can be priced with much lower margins and, as has already happened, be augmented with even cheaper devices like Echo Dots (or, in the case of the Echo Show, offer more functionality for a price that is still more than $100 cheaper than the HomePod).

The result is a product that, beyond being massively late to market (in part because of iPhone-induced myopia), is inferior to the competition on two of three possible vectors: the HomePod is significantly more expensive than an Echo or Google Home, it has an inferior voice assistant, but it has a better speaker. That is not as bad as it sounds: after all, the iPhone is significantly more expensive than most other smartphones, it has inferior built-in services, but it has a superior user experience otherwise. The difference — and this is why the iPhone is so much more dominant than any other Apple product — is that everyone already needs a phone; the only question is which one. It remains to be seen how many people need a truly impressive speaker.

This, broadly speaking, is the challenge for Apple moving forward: in what other categories does its business model (and everything that is tied up into that, including the company’s product development process, culture, etc.) create an advantage instead of a disadvantage? What existing needs can be met with a superior user experience, or what new needs — like the previously unknown need for wireless headphones that are always charged — can be created? To be clear, the iPhone is and will continue to be a juggernaut for a long time to come; indeed, it is so dominant that Apple could not change the underlying business model and resultant strengths and weaknesses even if they tried.

  1. Speaking strictly of which customers generate the most monetary value either through their purchases or advertising targeting []
  2. iPad hardware and optimizations all fell under the iOS umbrella []

Faceless Publishers

When I first worked for a (student) newspaper, the job of a publisher seemed odd to me; as far as I and my editorial colleagues were concerned, the publisher was the person the editor-in-chief, who we viewed as the boss, occasionally griped about after a few too many drinks, usually with the assertion that he (in that case) was a bit of a nuisance.

That attitude, of course, was the luxury of print: whatever happened on the other side of the office didn’t have any impact on the (in our eyes) heroic efforts to produce fresh content every day. We were the ones staying in the office until the wee hours of the night, writing, editing, and laying out the newspaper that would magically appear on newsstands the next morning, all while the publisher and his team were at home in bed.

The moral of this story is obvious: the publisher represented the business side of the newspaper, and the effect of the Internet was to make the job and impact of editorial easier and that of a publisher immeasurably harder, in large part because many of a publisher’s jobs became obsolete; it is the editorial side, though, that has paid the price.

The Jobs a Publisher Did

In the days of print, publishers provided multiple interlocking functions that made newspapers into fabulous businesses:

  • Brand: A publisher had a brand, specifically, the name of the publication; this was the primary touchpoint for readers, whether they were interested in national news, local news, sports, or the funny pages.
  • Revenue Generation: Most publishers drove revenue in two ways: some money was made through subscriptions, the selling, administration, and support of which was handled by dedicated staff; most money was made from advertising, which had its own dedicated team.
  • Human Resources: Editorial staff were free to write and complain about their publishers because everything else in their work life was taken care of, from payroll to travel expenses to office supplies.

What tied these functions together was distribution: a publisher owned printing presses and delivery trucks which, combined with their established readership and advertising relationships, gave most newspapers an effective monopoly (or oligopoly) in their geographic area on readers and advertisers and writers:

stratechery Year One - 264

Each of these functions supported the other: the brand drove revenue generation which paid for editorial that delivered on the brand promise, all underpinned by owning distribution.

Publishing’s Downward Spiral

It is hardly new news, particularly on this blog, to note that this model has fallen apart. The most obvious culprit is that on the Internet, distribution, particular text and images, is effectively free, which meant that advertisers had new channels: first ad networks that operated at scale across publishers, and increasingly Facebook and Google who offer the power to reach the individual directly.


I wrote about this progression in Popping the Publishing Bubble, and the intertwined functionality of publishers explains the downward spiral that followed: with less revenue there was less money for quality journalism (and a greater impetus to chase clicks), which meant a devaluing of the brand, which meant fewer readers, which led to even less money.

What made this downward spiral particularly devastating is that, as demonstrated by the advertising shift, newspapers did not exist in a vacuum. Readers could read any newspaper, or digital-only publisher, or even individual bloggers. And, just as social media made it possible for advertisers to target individuals, it also made everyone a content creator pushing their own media into the same feed as everyone else: the brand didn’t matter at all, only the content, or, in a few exceptional cases, the individual authors, many of whom amassed massive followings of their own; one prominent example is Bill Simmons, the American sportswriter.

Vox Media + The Ringer

I wrote about Simmons two years ago in Grantland and the (Surprising) Future of Publishing, and noted that media entities needed to think about monetization holistically:

Too much of the debate about monetization and the future of publishing in particular has artificially restricted itself to monetizing text. That constraint made sense in a physical world: a business that invested heavily in printing presses and delivery trucks didn’t really have a choice but to stick the product and the business model together, but now that everything — text, video, audio files, you name it — is 1’s and 0’s, what is the point in limiting one’s thinking to a particular configuration of those 1’s and 0’s?

In fact, it’s more than possible that in the long-run the current state of publishing — massive scale driven by advertising on one hand, and one-person shops with low revenue numbers and even lower costs on the other — will end up being an aberration. Focused, quality-obsessed publications will take advantage of bundle economics to collect “stars” and monetize them through some combination of subscriptions (less likely) or alternate media forms. Said media forms, like podcasts, are tough to grow on their own, but again, that is what makes them such a great match for writing, which is perfect for growth but terrible for monetization.

My back-of-the-envelope calculations estimated that Simmons’ Ringer podcast network was likely generating millions of dollars, and in an interview with Recode earlier this year, Simmons confirmed that is the case, claiming that podcast revenue was more than covering the cost of creating not just podcasts but the website that, at least in theory, created podcast listeners.

Still, given Simmons’ ambitions, it would certainly be better were the site more than a cost center, which makes the company’s most recent announcement particularly interesting. From the New York Times:

The Ringer, a sports and culture website created by Bill Simmons, will soon be hosted on Vox Media’s platform but maintain editorial independence under a partnership announced on Tuesday. Mr. Simmons, a former ESPN personality, will keep ownership of The Ringer, but Vox will sell advertising for the site and share in the revenue. The Ringer will leave its current home on Medium, where it has been hosted since it began in June 2016.

Jim Bankoff, Vox’s chief executive, said in a phone interview that the partnership was the first of its type for the company and would allow it to expand its offerings to advertisers. Mr. Simmons said in a statement: “This partnership allows us to remain independent while leveraging two of the things that Vox Media is great at: sales and technology. We want to devote the next couple of years to creating quality content, innovating as much as we can, building our brand and growing The Ringer as a multimedia business.”

Simmons is exactly right about the benefits he gets from the deal: instead of building duplicative technology and ad sales infrastructure, The Ringer can simply use Vox Media’s. This is less important with regards to the technology (Vox’s insistence that Chorus is a meaningful differentiator notwithstanding) but hugely important when it comes to advertising. It’s not simply the expense of building an infrastructure for ad sales; the top line is even more critical: it is all but impossible to compete with Google and Facebook for advertising dollars without massive scale.

Make no mistake, Simmons is the sort of writer that many advertisers would be happy to advertise next to (his podcast has had an impressive slate of brand names, in addition to the usual mainstays like Squarespace and Casper mattresses); the problem is that when it comes to the return-on-investment of buying ads, the “investment” — particularly time — is just as important as the “return”: a brand looking to advertise directly on premium media is far more likely to deal with Vox Media and its huge stable of sites than it is to do a relatively small deal with a site like The Ringer.

Indeed, the bifurcation in the Internet’s impact on editorial and advertising — the former is becoming atomized, the latter consolidated — explains why the implications for Vox Media are, in my estimation, the more important takeaway from this deal.

Vox Media’s Upside

To date Vox Media has been a relatively traditional publisher, albeit one that has executed better than most: the company has built strong brands that attract audiences which can be monetized through advertising, and that revenue, along with venture capital, has been fed into an impressive editorial product that builds up the company’s brands.

The Ringer, though, is not a Vox Media brand: it is Simmons’ brand, a point he emphasized in his statement, and that’s great news for Vox. The problem with editorial is that while the audience scales, production doesn’t: content still has to be created on an ongoing basis, and that means high variable costs.

Infrastructure, though, does scale: Vox Media uses the same underlying technology for all of its sites, which is exactly what you would expect given that software can be replicated endlessly. Crucially, the same principle applies to advertising: one sales team can sell ads across any number of sites, and the more impressions the better. Presuming The Ringer ends up being not an outlier but rather the first of many similar deals,1 then that means that Vox Media has far more growth potential than it did as long as it was focused only on monetizing its owned-and-operated content.

Publishers of the Future

The new model portended by this deal looks something like this:

stratechery Year One - 265

In this model the most effective and scalable publisher is faceless: atomized content creators, fueled by social media, build their own brands and develop their own audiences; the publisher, meanwhile, builds scale on the backside, across infrastructure, monetization, and even human-resource type functions.2 This last point makes a faceless publisher more than an ad network, and crucially, I suspect the greatest impact will not be (just) about ads.

Earlier this month I wrote about the future of local news, which I argued would entail relatively small subscription-based publications. Said publications would be more viable were there a faceless publisher in place to provide technology, including subscription and customer support capabilities, and all of the other repeatable minutiae that comes with running a business. Publishers still matter, but much of what matters can be scaled and offered as a service without being tied to a brand and a specific set of content.

I suspect this is part of the endgame for publishing on the Internet: free distribution blew up the link between editorial and publishing and drove them in opposite directions — atomization on one side and massively greater scale on the other. And now, that same reality makes possible a new model: a huge number of small publications backed by entities more concerned with building viable businesses than having memorable names.

Disclosure: I have previously spoken at the Vox Media-owned Code Media conference and was previously a guest on The Bill Simmons Podcast; I received no monetary compensation for either appearance

  1. There is already a parallel to The Ringer within Vox Media: the company’s vast network of team-specific sites that sit under the SBNation umbrella []
  2. This is where Medium went wrong: the company made motions towards this model — which is why The Ringer is hosted there — but has decided to pursue a Medium subscription model instead []