Will Spotify Ruin Podcasting?


Welcome to BIG, a newsletter about the politics of monopoly. If you’d like to sign up, you can do so here. Or just read on…

Today I’m going to write about podcasting, with some observations on how a media market that is far healthier than online publishing functions. But first, a quick update on the Democratic Party cartel . Most people saw the fiasco of the Iowa caucuses as a political story, which it obviously is. But it’s also a monopoly story. The corporation, ACRONYM, which ruined the caucuses with their Shadow app, has power because of their relationship with big donors. This nugget in Politico drives home the point.

Critics of ACRONYM declined to speak on the record because they are concerned about maintaining party unity and retribution from the group’s donors.

Imagine that, fellow practitioners are so afraid of retaliation that they can’t openly criticize a self-dealing con artist who openly ruined the Iowa caucuses, a sort of 737 Max moment for Democrats. This is the same fear and coercion we feel in any concentrated industry, only applied to the business of Democratic politics.

And now…

Photo credit to https://nicolassolop.com/.

A Functional Market

From 2000-2006, the web was an open place. If you built a highly trafficked web property, you could finance yourself by selling advertising. It wasn’t just that the internet itself was a decentralized network, so were financing streams behind content. There were many publishers and paths for readers to get to those publishers through different search engines or aggregators, and ad networks matched advertisers to those properties. The result was a flourishing of voices and new media projects.

It was a three tiered system, with production, distribution, and advertising operating in vertically separated layers. There were certainly industry players who breached these layers - the New York Times sold its own ads and distributed its own product, but no one had dominant power in any one layer.

That is very much what podcasting looks like today. About a third of Americans listen to podcasts regularly, and there are 750,000 active podcasters, including independent successes like Chapo Trap House, The Bill Simmons Podcast and Joe Rogan, as well investment by major newspapers like the New York Times, and capital-infused podcast-only networks. Nonprofits have podcasts, so do industry associations.

There are real and serious problems in podcasting, as there are in any industry, leading to union drives and sometimes to attempts to ruin the industry by overfunded Wall Street goons. As more ad money floods into the space and branded podcasts become more important, there are debates over labor relationships, as well as censorship and advertising especially within large branded podcast institutions, like Pineapple Street Studios or Radiotopia or Spotify. Fear of Apple, which has a powerful spot in podcast distribution through its podcast app as well as its app store, is rife.

But these are debates within a growing industry in which various stakeholders debate values and rules. It’s nothing like the nuclear winter of the non-audio internet content industry.

In podcasting, the major directories for distribution use an open standard called RSS to list podcasts, a standard originally developed by open web advocates like Aaron Swartz. Apple has the dominant podcast app. I’ve heard there are some issues with how Apple deals with ratings, but so far, Apple operates as a benevolent despot, largely not collecting data and not privileging its own content. There are also a host of advertising networks, as well as some subscription options, so financing is relatively distributed.

Podcasting is a three tiered system, with production, distribution, and advertising in vertically separated layers. There are corporations who produce, distribute, and sell podcast ads, but the markets in all three layers are open. As a result, there’s relative ease of entry. It’s hard to build an audience, but if you do, you can get access to a financing channel. And the net result is a lot of diverse voices and entrepreneurship. This dynamic is changing somewhat, as branded podcast corporations increasingly build formulas to launch new shows, but relative to near all other media ecosystems, with the exception of the mid-2000s open internet, it’s open. And the debates within the industry about contested relationships between advertisers, podcast hosts, corporate management, and workers happen within a context of a relatively flexible system.

There may be different interest groups within the industry, but the industry as a whole has an interest in preserving this structure. It was very much like the open web of the mid-2000s. So there are many lessons podcast industry players can learn from what happened to that space, and how to prevent it from happening to podcasting.

Why the Open Web and Publishing Began Dying

Anyone trying to understand modern media or monopoly has to spend a lot of time understanding Google and Facebook, because they are the pace-setters in our economy. Every corporate leader, from agricultural to podcasting, sees what they have done, and tries to reproduce their success in their own industry.

While Google and Facebook are framed as tech companies, they are in fact advertising companies and middlemen in the flow of information. Google gets roughly 80% of its revenue from ads, for Facebook it’s a little over 98%. And what they did from 2004-2014 was to redirect the flow of ad money from publishers to themselves. So that’s what’s happening throughout the economy, especially in the media industry; middlemen everywhere are trying to find ways to redirect the flow of other people’s revenue to themselves.

Google and Facebook did this through two key techniques. The first was to acquire gatekeeping power in distribution. Google is gatekeeper in search, online video, and maps, whereas Facebook is a gatekeeper in social networking. To get to users, you have to go through Google and Facebook.

The second was to use this gatekeeping power to vertically integrate into dominant advertising platforms. Google and Facebook both sell large amounts of advertising, and they both bought up adtech companies in a merger spree from 2004-2014. They force partners to hand over data - and data is a key input into advertising - as a condition of getting access to their networks.

Such data wasn’t just personal information about the users, but more importantly, proprietary business data of publishers. The Wall Street Journal has a list of subscribers and readers, and it would never share this information with the New York Times, because they are competing to sell space to the same advertisers. But both Google and Facebook got all of this proprietary information, and more, from the publishers using their networks.

With most users going through either Google properties or Facebook properties, these corporations could then de-commodify the ad value of a publisher ad slot. After these corporations became dominant, there simply was no need to advertise in, say the New York Times to show an ad to a New York Times reader if Google could show that ad to that same wealthy reader cheaper somewhere else. The relationship between the New York Times and the reader is based on trust earned through capital invested in reporting the news. Google then takes that relationship in the form of structured data, and makes money that should have gone to the New York Times by advertising to New York Times readers against non-New York Times content. This looks more like a form of theft, not efficiency.

Facebook and Google dominance is across an entire layer - search or social - and also vertically upstream into the advertising, and increasingly the distribution through apps and browsers.

It’s important to recognize that these corporations shape their business models around public policy decisions that enable monopolization and targeted ads. Both Facebook and Google acquired power a series of anti-competitive mergers, like Google’s purchase of YouTube and DoubleClick, Facebook’s purchase of Instagram and WhatsApp. Similarly, during debates over privacy and data in the 1990s and 2000s, libertarians won the argument that data markets should remain largely free of public rules. And so these two platforms, who were running communications networks that should have been regulated as public utilities, could discriminate and insert ads in front of those who needed to use their networks to reach customers.

The result of these public policy decisions was Google and Facebook, the death of the open web, and increasingly, independent publishing on the internet.

So what lessons does that hold for the podcasting industry?

The Spotify Challenge

This week, Spotify announced an acquisition of The Ringer, one of the biggest networks in podcasting, built by sports writer Bill Simmons, with 30 different podcasts that have hundreds of millions of downloads a month. The CEO of Spotify discussed what they see in terms of value with this acquisition, saying “we bought the next ESPN.” It’s an open question what Spotify is going to do with this acquisition, but the strategy of the corporation seems fairly clear.

Spotify is rolling up the internet audio market, which their CEO, Daniel Ek, thinks will grow from $100 billion in revenue a year to somewhere near the video market, which is $1 trillion. People spend about the same amount of time consuming audio as video content. “Are our eyes,” he asks, “really worth 10 times more than our ears?”

From 2014 to 2020, Spotify bought 15 companies, companies that build everything from data analytics to music and audio production tools to audio ad tools to licensing platforms, and podcasting networks. These companies included the Echo Nest (2014), Seed Scientific (2015), CrowdAlbum (2016), Sonalytic (2017), MightyTV (2017), Mediachain (2017), Niland (2017), SoundTrap (2017), Loudr (2018), Gimlet (2019), Anchor (2019), SoundBetter (2019), Parkast (2019), and now The Ringer (2020).

Spotify is directly mimicking Google and Facebook, and attempting to roll up power over digital audio markets the way Google and Facebook did over the internet. It has already done so in music. Here’s Rolling Stone, reporting on Spotify’s exploitation of its public utility platform of music distribution to organizes payola-style extortion against artists.

In Spotify’s pitch deck, which was sent to a distribution company, the platform advises artists or labels to spend at least $5,000 on these campaigns. If an artist pays that amount at 55 cents a click, the campaign should bring in more than 9,000 potential listeners over a seven-day period. 

The streaming service’s announcement immediately drew comparisons to pay-for-play at radio or, maybe more accurately, the retail variation, where record shops asked labels to pay for premium positioning. “Stores would say, ‘we want you to pay us for this space,'” explains George Howard, a professor of music business and management at Berklee College of Music. “Spotify is saying, ‘we want you to pay us to display your records.’ All this does is continue what payola always has done — the major labels, which have the most money and the most frequent releases, get the most play, consolidating the amount of art that is put out there.” 

Payola of this sort tends to work against small players, and cartelize industries into giants negotiating with one another. Will that happen in podcasting? Spotify is likely seeking to make it so, especially since it went public and has to answer to investors who want a story about the company’s strategy, which is a coded way to talk about acquiring market power. Last year, Spotify bought several podcasting companies, including podcasting network Gimlet Media Inc. (“Gimlet”), as well as Anchor, which made podcast creation and ad tools. Ek made it clear during this acquisition that the goal of Spotify is to dominate podcasting. Here’s what he wrote in his blog post explaining the acquisitions.

No other company has Spotify’s scale and audience around the globe with the 78 markets where we do business. No other audio company has the two-sided marketplace that we have built at Spotify — a marketplace that benefits artists and creators along with consumers. Nobody else has both audio advertising and subscription revenue model at scale globally. Nobody else in music has the engineering capabilities and the expertise in audio that we have at Spotify.  And with the addition of Gimlet and Anchor, Spotify will now become the leading global podcast publisher with more shows than any other company. 

Podcasting, he argues, can become a business based on scale and vertical integration. “As we expand deeper into audio, especially with original content,” he wrote, “we will scale our entire business, creating leverage in the model through subscriptions and ads.” The business can “become significantly larger when you add Internet-level monetization to it.” Ek isn’t really being clear about what he means by scale. Podcasting, with its relatively decentralized structure, is already a scale business. RSS is an open standard and tens of millions of people listen to podcasts.

What Ek really means is that he’s trying to privatize the open standards, basically take over the already large public commons. To the extent there is scaling going on, it’s just doing sales work to move ad budgets from big companies in other audio or video channels to Spotify. This is scaling like Google and Facebook did scale, they took over an already massive web organized by open standards and democratic participation, and pretended they built it so they could control it.

Ek also added typical language for aspiring dominant players, which is pablum on how there’s “fierce competition” in podcasting and how his company believes in a level playing field. Such language is almost always inserted into merger acquisition press releases by antitrust lawyers, and is a clear tell, like a kid with crumbs on his face preemptively telling his parents “I didn’t eat any cookies before dinner.”

Commodifying Independent Podcasts

Recently, Spotify announced a new advertising service, Spotify Podcast Ads with what’s called ‘streaming ad insertion.’ Here’s what it means.

The technology makes key data — like actual ad impressions, frequency, reach, plus anonymized age, gender and device type — available to podcasters and advertisers for the first time.

In previous years, podcasts have been delivered by way of downloads from RSS feeds, which would make this sort of data collection difficult if not impossible. The shift to streaming changes that, as Spotify can tap into its suite of planning, reporting and measurement capabilities, as it does for streaming music.

In other words, Spotify is doing what Google and Facebook did. First, it is privatizing the distribution standard for podcasts by converting more people to streaming their podcasts through Spotify services instead of the open RSS standard. That means it is becoming a gatekeeper to consumers. Second, it is vertically integrating into advertising, and developing the technology to grab data that podcasts generate through their trusted relationship with listeners. It then uses that for targeting.

Spotify seems to be trying to build out control of advertising and distribution of podcasting. With gatekeeping power over listeners through its streaming service, and gatekeeping power over ad revenue through its advertising network, Spotify will eventually be able to force podcasters to live in its ecosystem. It won’t be impossible to get listeners without Spotify, just as newspapers can technically get direct traffic instead of traffic from Google and Facebook, but it’ll be very hard.

It’s not 100% clear how well this will work. Podcast ads are, like radio ads, often based on the trust of the host of the show. Hosts tend to read out ads, so they are more like paid sponsorships relying on the trust and voice of the host than traditional advertisements. But if the Spotify ad insertion tech is sophisticated and effective enough, they may be able to alter this dynamic, making any particular ad slot on any podcast less valuable.

No advertiser will care if you’re a listener of Joe Rogan or Bill Simmons, only that you are a 34 year old male with a certain income reachable in thirty forty different audio slots, which can then all go in an auction. Or even if they do care, competitive ad networks who offer the service you want will probably die. Then, just as the New York Times content becomes far less important online because Google can just find you that New York Times reader through another publisher outlet or Google’s own properties, the actual podcast becomes commodified, because all that matters is the listener data combined with the ad slots, not the show against which those ad slots are sold. This is another complicated way of saying the people who do the work of making and distributing a show don’t get the benefit from the work they do.

What happens in this scenario is that a few giants, not just Spotify, ultimately become dominant vertically integrated podcast, distribution and advertising platforms. These will then cut deals with each other, and cartelize the industry into a land of giants, much as Disney is doing in Hollywood. (In fact, I wouldn’t be surprised to see these two spaces merge.)

The Clayton Act and Incipiency Standard

None of this is inevitable. And in fact, there’s one significant difference between today and the mid-2000s, and it’s not a small one. Today, we understand that monopolies are dangerous. This means we can activate a whole series of laws designed to protect us from them.

Fortunately, there is a law to address the merger spree Spotify is engaging in, and it’s called the Clayton Act. Passed in 1913 and strengthened in 1950, the idea is to stop monopolies through mergers, even when such monopolies are in their incipiency. A lot of states have similar laws against mergers. Similarly, states and Congress and foreign regulators are debating privacy and data laws, and what is this Spotify problem if not a privacy and data problem? Podcasters, if they chose to organize around this market structure problem, could have substantial political power.

Today’s flourishing podcasting market is evidence that we can have a diverse and financially viable media market. We just have to stop gatekeeping in our markets for speech, prevent vertical integration, end the inappropriate use of other people’s data to let a middleman like Spotify inappropriately monetize art it didn’t create, and retain open standards like RSS for podcast distribution instead of privatizing public utility functions.

Oh, and it’s probably a good idea to tell the Federal Trade Commission to block the Ringer merger. They can do that, you know.

Thanks for reading. And if you liked this essay, you can sign up here for more issues of BIG, a newsletter on how to restore fair commerce, innovation and democracy. If you want to really understand the secret history of monopoly power, read my book, Goliath: The 100-Year War Between Monopoly Power and Democracy.


Matt Stoller

P.S. Here’s an email on CVS’s wonderful ‘integration’ of its different lines of business.

I have an additional tactic CVS/Healthcare companies recently employed to get even more money out of those of us unfortunate enough to be sick. I take a specialty pharmacy drug for an auto-immune disease. There is no cure but this medicine greatly improves my quality of life and allows me to function somewhat normally and be gainfully employed rather than on benefits. It is a very expensive drug, taken monthly so I factor that into my decision when picking a healthcare plan on the marketplace. Fortunately, the drug manufacturer offers a rebate program that can reduce the cost of that drug from $5000+ refill to $5. For years, I was able to put that co-pay assistance card on file with the pharmacy and they would bill that directly. My co-pay would be hit for the year and I could rest a little easier knowing a hospital visit would be covered. 

A few years ago, the insurance company stopped applying co-pay assistance payments toward my deductible. When I called to ask how they would even know that I had paid for a prescription that was fulfilled by the pharmacy using a co-pay assistance card, which they previously should have had no knowledge even existed, they responded that the pharmacy managed all their prescription drug benefits now so the information was "in-house" so to speak. Want to take a guess who the companies were? CVS Caremark and Blue Cross Blue Shield. For the record, on the South Carolina marketplace that year, BCBS was the only healthcare company offering plans through healthcare.gov. This is not a shot at the ACA, which helps me greatly by ensuring I cannot be denied coverage for my preexisting condition, but the idea of calling something a "marketplace" when there is only one company offering plans is ludicrous.

I am fortunate enough to have good credit and a way to absorb those ridiculous drug costs every year but I am sure millions of Americans cannot just put $5000 on a credit card and pay it off as over the course of the year. On a side note, I often wonder how much of the "drug cost" I pay goes toward the expense of commercials for the drug I see every Sunday during NFL games? 

Thanks for your time and I really enjoyed your book. I tell everyone I know to read it so they can learn about the bi-partisan efforts to erode small business in America.