Welcome to Marketing BS, where I share a weekly article dismantling a little piece of the Marketing-Industrial Complex — and sometimes I offer simple ideas that actually work.
If you enjoy this article, I invite you to subscribe to Marketing BS — the weekly newsletters feature bonus content, including follow-ups from the previous week, commentary on topical marketing news, and information about unlisted career opportunities.
Thanks for reading and keep it simple,
Edward Nevraumont
Spotify’s big deal
On May 19, Spotify announced a three-year partnership with podcast host Joe Rogan. An article in the Wall Street Journal describes the key elements of the deal:
Joe Rogan is taking his podcast exclusively to Spotify in a licensing deal worth more than $100 million, according to a person familiar with the matter.
The licensing agreement is one the largest such deals in the rapidly growing podcast business.
The comedian, television host and mixed martial arts commentator has become one of the most influential and lucrative podcast hosts in recent years, especially with his popular “vodcast”—video podcast—format on YouTube.
Mr. Rogan, 52 years old, had so far withheld his podcast from Spotify, saying the streaming service doesn’t pay enough and he had been generating significant revenue on other services such as Alphabet Inc.’s YouTube. His full library, dating back 11 years, is to hit the service Sept. 1, and become exclusive to Spotify after that, before the end of the year. His video podcasts, which will also appear on Spotify, will no longer be available on YouTube.
Joe Rogan was a journeyman stand-up comedian for about 15 years, until he attracted mainstream attention for three different things: a starring role on the sitcom NewsRadio, color commentary for UFC (Ultimate Fighting Championship), and hosting the stunt-based reality TV show Fear Factor. Back in 2009, Rogan experimented with ways to livestream a comedy talk-show that he could also release as a downloadable podcast. As of June 14, The Joe Rogan Experience has produced 1491 episodes. About three times a week, Rogan sits down for long-form interviews with just about everyone: celebrities, politicians, scientists, business leaders, and more. In one particularly notable episode, he smoked marijuana with tech mogul Elon Musk.
The Joe Rogan Experience consistently ranks first or second on Apple’s list of top podcasts. In 2013, the show started posting videos on YouTube. Each episode attracts approximately two million views — the Elon Musk one has 35 million!
In terms of content and audience, many media analysts compare Rogan to shock jock Howard Stern. A further similarity: Stern signed a massive deal to move his popular radio program from free FM channels to Sirius Satellite Radio — a subscription-based platform. Likewise, Rogan fans will need to sign up with Spotify to continue listening to his podcasts (Spotify offers both free and paid membership options).
The Rogan deal is not Spotify’s first foray into podcasting. In fact, the platform has gone “all in” on podcasts, with a series of bold moves including acquisitions of The Ringer (sports website and podcasts), Gimlet (NPR-style professional podcast producer), Anchor (podcast production tool for creators), and Parcast (True Crime podcast production company).
Spotify’s master plan
What is Spotify trying to accomplish with the Joe Rogan partnership?
I can think of two (non-exclusive) possibilities:
Become the Netflix of audio
Become the Facebook of audio
Netflix and Spotify both stream media, but with two very different business models. The Netflix model is straightforward: the company hosts a platform of video content and they charge customers a monthly fee to access that content. Aside from a 30-day free trial, there are no (legitimate…) ways to enjoy Netflix for free. They have very high fixed costs — paying licence fees for content from other studios as well as producing their own original content — but very low variable costs. As Netflix grows their subscriber base, the cost to produce a given amount of content remains unchanged. Regardless of their margins today, they can “grow into profitability” just by acquiring more customers. And if Netflix’s audience continues to increase, they can afford to licence or create even more content, which makes their product better, which attracts more subscribers ad infinitum (or at least until they run out of humans who want to watch premium long-form video).
On the surface, Spotify’s model seems comparable to the Netflix one. People pay a monthly fee to access a massive library of music and podcasts (the free version offers limited options and lower audio quality). Due to the traditions of the music industry, though, Spotify pays their content producers a revenue share instead of a fixed fee. Plus, (almost) all of Spotify’s music content is also available on other streaming platforms.
Despite the superficial similarities, Spotify’s model is the OPPOSITE of Netflix — they have low fixed costs, high variable costs, and non-exclusive content. If Spotify grows their audience, their margins do NOT improve; they continue to send the same percentage of their revenue to the rights holders of their audio content. One other key difference: unlike Netflix, Spotify does not have a “flywheel” to help build momentum. As Netflix collects more revenue, they can create more exclusive content (like Ozark or Stranger Things) to lure more subscribers. Spotify does not have that same opportunity. In the music streaming industry, having the biggest collection doesn’t really matter — you can listen to (almost) all of the same songs on Apple Music, Amazon Music, Pandora, etc.
Podcasts change the equation.
With podcasts, Spotify could become the Netflix of audio. Instead of paying a revenue share for music, Spotify could negotiate fixed costs for the licensing of third-party podcasts. Moreover, podcasting is still a wide-open industry. Cracking into the music business to sign and promote new artists would be a very difficult venture, but Spotify could find success producing their own exclusive podcasts — available only to their subscribers. Spotify could invest in podcasts, improving the quantity, quality, and variety of their collection. Better podcast content could attract more customers, boosting the company’s revenues (and at higher margins due to the fixed cost structure of the podcast deals). Greater revenue allows for more reinvestment in content, creating a flywheel like Netflix.
One possible (and significant) hitch: this model relies on finding enough people who love podcasts enough to pay a monthly subscription. As I described in a previous newsletter, Tim Ferriss attempted to move his popular podcast to a subscription. Ferriss quickly reverted back to free podcasts when he realized two things: (1) people were not very willing to pay for podcasts, and (2) he could earn more money hawking products.
With all of these ideas in mind, Spotify might plan on becoming the “Netflix of audio,” but I don’t think that is the most likely strategy.
Instead, I believe Spotify is shooting to become the Facebook of podcasting — which would be a very important development for marketers.
Self-serving advertising
In the pre-internet era, many companies used a standard plan for launching a new product and boosting brand awareness:
Run enough television ads so that people knew about your product (create demand / mental availability / discovery).
Work with large retailers to ensure that your product was on the shelf (create supply / physical availability / convenience).
The two strategies worked in tandem. If you could show the retailers that you were driving demand (with the TV spots), they would be willing to not only place your products on the shelf, but also offer good prices. And with the availability on the retailers’ shelves, you could achieve a positive ROI on your demand-driving television spend.
Today, the mechanics of a launch are completely different.
Customers can directly visit your website, but availability is largely driven by Google and Amazon. As such, companies vie to appear at the top of both sites’ search results. Television advertising remains more powerful than many people give it credit, even though its impact has eroded over time. Now, Facebook is the most important paid discovery channel for many companies and brands.
Compared to traditional television, Facebook has two significant disadvantages:
Facebook ads are far less immersive. Most ads are rudimentary combinations of images and text. Videos ads might auto-play, but not with the accompanying audio.
Facebook ads are targeted. As such, companies cannot benefit from “second order effects.” For example, suppose Corona ran TV commercials proclaiming their product as the “beer of beach parties.” I might bring Corona to a beach party — not because I think that I was influenced by the ad — but because I believe that other people were influenced by the ad and will therefore appreciate that I brought a beach beer to a beach party. This phenomenon only occurs for “non-targeted” ads (like TV commercials) that you know everyone will have encountered.
Facebook also holds a number of advantages over both television and other online ads:
The ad unit is very good. The ad appears as PART of the “feed” experience, rather than something that sits OUTSIDE the content (like a banner ad or a television ad break).
The tracking is excellent. You know exactly who sees your ad, who clicks on it, and if they buy your product (or take another action).
The targeting is very good. Facebook’s lookalike audiences are far better than targeting based on a television show — or even based on demographics and interests.
The friction to advertise is very low. The Facebook ad platform is self-serve, offering companies the opportunity to get started for as little as one dollar.
Podcast advertising, in theory, could hold some of the advantages of both television and Facebook:
The ads could be immersive. Even without video, sound-based ads can be extremely impactful.
Podcasting could create strong ad units. If you are a regular podcast listener, you have likely heard ads that are part of the experience, rather than outside of the content. For example, conversational hosts (like Marc Maron and Dax Shepard) often improvise around the ads with personal anecdotes for several minutes. In contrast, journalistic shows (like The Daily and This American Life) use pre-recorded spots.
There are three significant problems with podcast advertising: tracking is terrible, targeting is almost non-existent, and friction is very high.
Tracking is mostly restricted to two metrics: “number of downloads” and “number of coupon codes activated” (e.g., “Go to podcastname.com/product to get 10% off”). Counting the number of downloads can easily overstate the actual number of listeners. If you subscribe to any podcasts, your device might automatically download new episodes; this counts you in the audience stats, even if you never play that episode. Plus, there are no ways to verify how many people listened to the entire podcast or how many people pressed the “skip 30 seconds” button to miss your ad entirely.
In terms of targeting, some podcast producers have accumulated customer surveys (of variable quality). You could try “interest targeting” based on the topic of the podcast, but that’s less helpful for mainstream podcasts that capture broad audiences. Some podcasts, like the ones on NPR, attempt to offer geographic targeting, but this strategy tends to be inaccurate, expensive, and only available for a small subset of the listening audience.
Likewise, buying podcast ads is a difficult process. You generally need to negotiate directly with a podcast producer, which is expensive for both sides. If your podcast is — like most — modest in size, then building an advertising sales force is an impossibility. If you are a media buyer looking to optimize your time, then you narrow your focus to the podcasts with the largest audiences. As a result, ads are usually bought in large “packages” that run over extended periods of time across multiple podcasts. (You’ve likely heard repetitive ads from a few major players: Squarespace, ZipRecruiter, Indeed, etc.).
Due to issues with tracking, targeting, and friction, podcasts vastly “under-monetize.” A Statista report estimated the total podcast listening time across the US in 2019 at 12 billion hours. For the same period of time, Statista estimated total adverting revenue as $420 million. Based on that information, we can calculate an average revenue/hour of ~3.5 cents. Television viewership among American adults is astonishing — as high as 386 billion hours in 2019. Total television advertising revenue for 2019 was pegged at $76 billion. That gives us revenue/hour of ~19.7 cents for television — almost 6x more than podcasts.
Some companies are seizing on the opportunity to improve the efficiency of podcast advertising. You can imagine someone starting a marketplace where podcasts can put up inventory and media buyers can buy spots; the self-serve process could use an auction format (like Facebook ads). If successful, platforms like that could deal with one of the gaps in the podcast market, but there is still a need to fix terrible tracking and non-existent targeting.
Enter Spotify
Spotify’s podcast player works differently than others on the market. Apple, for instance, allows anyone (with an iTunes account) to download podcasts for free; additionally, the Podcasts app on a Mac or iPhone will automatically download new episodes from a person’s list of subscribed podcasts. In contrast, Spotify has emphasized the streaming of podcasts. There is an option to download podcasts, but that feature is limited to their customers who pay a monthly fee for Spotify Premium.
Spotify’s preference for streaming, coupled with the info they collect from registered users, could help podcasts close their advertising gap with Facebook.
Let’s review what Spotify knows about a listener:
They know who the listener is. Now marketers can target based on demographics.
They know when the listener played the podcast. This allows targeting by time of day. Plus, Spotify could track if the ad was heard (or skipped). Now marketers can pay based on a more accurate number of ads heard, rather than total downloads.
They know where the listener is. Spotify can track the user’s location, as well as their billing address, and other places where they spend their time. These granular details offer even better geo-targeting than Facebook
They know what the listener has listened to in the past. Marketers can target interests with greater precision. You might be listening to the Joe Rogan Experience today, but if I know you were listening to the MarTech podcast yesterday, then I can target you with marketing technology ads on Rogan’s show.
What key element is still missing? The ability to build predictive lookalike audiences. When Facebook launched lookalikes, it was a poor substitute for micro-targeting via interests. But when Facebook was able to incorporate purchase data into their algorithms, lookalikes became vastly better than any other targeting method. For the same reason, some companies are having great success with Amazon’s ad platform (Amazon holds a fair amount of proprietary purchase behavior they can use to make lookalikes). Unfortunately for Spotify, they do not possess any proprietary purchase data, which will limit its ability to algorithmically target.
And there is a remaining challenge: scale.
As I wrote about in a previous newsletter, there is a fixed cost for a marketer to use a new channel. If the total potential size of a channel is not that big (e.g., Reddit vs Facebook), it is just not worth the marketer’s time to expend the effort needed to optimize the channel and make it effective. This concept is especially true in situations where targeting matters. For instance, suppose a podcast boasted having one million listeners. That sounds great, but if the audience was spread equally across the US, it would only translate to 700 “women based in Denver” — not nearly a big enough audience if I wanted to advertise for something like a women’s clothing store in Denver. When trying to target based on regions or interests, scale matters a LOT.
You can understand Spotify’s interest in Joe Rogan. His podcast is downloaded ~190 million times per month. His audience is enthusiastic and loyal. I expect many of his fans will happily go through the trouble of downloading a new podcast app.
As an avid podcast listener, there are many personal reasons why I would not like to see podcasts move into Spotify’s “walled garden.” As a marketer, though, I see great potential in Spotify’s future. They might finally bring podcast advertising into the light and perhaps even provide an alternative to television (especially as that channel continues to decline in relevance).
Keep it simple and stay safe,
Edward
If you enjoyed this article, I invite you to subscribe to Marketing BS — the weekly newsletters feature bonus content, including follow-ups from the previous week, commentary on topical marketing news, and information about unlisted career opportunities.
Edward Nevraumont is a Senior Advisor with Warburg Pincus. The former CMO of General Assembly and A Place for Mom, Edward previously worked at Expedia and McKinsey & Company. For more information, including details about his latest book, check out Marketing BS.
Comments