The publishing business has experienced major changes over the past few decades. When paper newspapers were the primary vehicle for distributing news, the publisher held the power. Because there was a large upfront investment (printing press and trucks) and a physical moat, once a publisher established their business, they were hard to dethrone.
To begin, we must look at how advertisers view this market. Before the internet, an advertiser that wanted to target a certain demographic could buy ads on TV, radio and billboards and expect that they were getting as much mind share as possible. There weren’t a ton of other places that sold ads.
When the internet was introduced, there was a new layer involved. Consumers were now spending more and more time online. With that, the offline ads that they purchased before were seen less and less. Advertisers needed to start buying ads online, where the consumers were spending their attention.
When advertisers looked online to buy ads, they were faced with several questions. Which websites did the target audience go to? Where was the target audience spending the most time? Where was the target audience most engaged? Consumers were visiting hundreds and thousands of different websites and the internet didn’t have a clear way for advertisers to reach that online audience in a practical manner.
“Infinite” ad inventory is an important concept created by the internet. The scarcity on the number of newspaper ads (paper space) or TV ads (commercial time) meant that inventory was limited, and the value could increase with volume of viewers. But because there is an unlimited number of web pages, there is an unlimited amount of ad inventory. Ads can be placed on any web page and web pages can be created ad infinitum. With no information about the consumer, web page ads alone are very low value for advertisers – there is nothing holding users’ attention and there is no supply constraint.
One step up from that, which several websites had, was content that was focused on a specific topic. This was now on par with newspapers and TV. An example could be a golf brand previously advertising in the sports section of newsletters and on the ESPN channel now also advertising on sports websites. However, since there were hundreds of sports websites (and hundreds more lifestyle websites that would be interesting to an ad buyer) it was hard to decide which ones were most valuable.
Another problem with this advertiser relationship with websites was logistics. Not only was it hard to decide which websites to buy ads on, but it was hard to coordinate. Each website often had their own ad sales team and there wasn’t enough time and resources as a buyer to deal with all these sales teams.
Ad networks, most notably Facebook and Google, provided solutions to advertisers’ buying problems. The main thing they provided was more data on consumers. For Facebook, this was who you were, who your friends were, and what you liked. For Google, this is what you were searching for. With this information alone, ad networks became much more valuable for ad buyers because they could target consumers better than anything before.
An additional benefit – which isn’t trivial – is they solved the logistical problem. Not only did ad buyers no longer have to deal with hundreds of sales teams, but they could buy ads online through programmatic advertising. As with many marketplaces, there are network effects to ad networks. With more traffic and more data, these ad networks became incrementally more valuable to ad buyers.
It’s important to note that small-scale ad networks were hard to build. Paid acquisition of traffic to an ad network site was an arbitrage between the CPM for paid traffic and the CPM of the ad inventory sold. This arbitrage was unsustainable. Growing organically was possible, but it took a lot of time and resources to get the millions of users that would sustain a business.Additionally, without user data (and a wide variety of users) CPMs were likely to be very low. Ad networks proved to be one of the strongest winner-take-all environments, and Facebook and Google emerged as the two big players. This wasn’t an easy thing to accomplish and is why Facebook and Google are such valuable companies.
Ad networks changed the relationship between publishers and users. In any ad-based business, the product is the user, not the content. Advertisers pay the ad networks for users’ attention. The difference between online and offline ad sales was that ad networks had data on users. Their business models hinged on two things: attention and data. Everything else, including the content, was in service of those goals.
For Facebook, this made the content on the feed commoditized. The content of the feed didn’t matter as much, as long as it kept users on the platform. Advertisers were happy because they were able to target based on factors that had little to do with the content, such as demographic and preference data that came from user profiles and user actions (liking pages, etc.). With effectively infinite user-generated and publisher content on Facebook, Facebook’s job was to create a feed that users stayed on as long as they could, serving a targeted ad every so often.
This was bad for content creators that relied heavily on the Facebook news feed. Remember that advertisers, the people who paid Facebook, were happy as long as 1) they had data to target users and 2) users stayed on the platform to see the ads. The type of content that users were viewing was little concern for the revenue source of Facebook, outside of their ability to extract data from it to increase targeting quality. Buzzfeed, Zynga and many others that tried to leverage the Facebook platform struggled because the quality of content is not what Facebook was optimizing.
Another important point is that the internet massively increased the TAM for all publishing. There are roughly 130 million households in America, but 5 billion people have smartphones and over 7.5 billion people on earth total. The purchasing power is certainly not equal, especially in developing countries, but the increase in reach means there is a larger market. It also means there is more room for niches to develop into sustainable business models.
In the internet age, both pieces of content and individual creators have more power. Individual pieces can go viral by themselves. Individual writers develop a rapport with readers by producing consistently high-quality content. This rapport is built on existing news outlets as well as independent blogs and social media.
As discussed, ad-supported business models are hard for publishers without a very high number of users. Since high-quality, low frequency content doesn’t always scale to tens of millions of users, niche content creators will increasingly rely on subscriptions for monetization. Besides ad-supported and subscription, some content will be created for the purpose of content marketing. Publishers may provide content for free in exchange for publicity, branding and credibility. Additionally, there will always be content created without monetization intent.
Almost all newspapers have created a digital product. However, few of them have been successful in switching their business model to digital. One outcome of unlimited distribution is only the best of the best survive. In national news,for example, The Washington Post and New York Times are dominating digital, while many other newspapers are struggling online.
Additionally, many digital publications are ad-driven. I’ve heard stories of journalists being pressured into writing several lower-quality pieces compared to fewer high-quality pieces. This is a result of the ad-driven business model and can occur even without bad actors in the system (not to say that there aren’t any).
For these reasons, journalists have incentives to look for other options. A business model such as subscription aligns the value that the writer provides with the end user. Not to mention that several of these newspapers have serious long-term questions. The Athletic CEO Alex Mather has mentioned on podcasts that it’s relatively easy to recruit journalists from legacy newspapers. The pitch is they can do the best work of their career because the business model is focused on providing value, not impressions or clicks.
People who don’t come from traditional journalistic backgrounds also produce content. The amount of people who do this will increase. Many of these people will consistently publish and eventually develop trust and loyalty with an audience. While 10k-50k email subscribers isn’t usually enough to command strong enough CPMs to sustain an ad-business, a writer only needs 1,500 paid subscribers at $50/year to have a $75k/year business. Platforms like Substack, Patreon and Memberful offer platforms to facilitate this.
Overall, building subscription businesses is a great way to incentive creators to make their best content. Paying for content is a signal that users trust and value content creators instead of relying on ads to pay for content creation.
One of the most interesting companies in this space is Substack. Substack is an email platform that makes it simple for writers to create, manage and distribute content. It’s free to create a free newsletter as a writer. If the writer wants to charge for their newsletter, they writers charge is between $5-$10 per month on average and usually have 70%-80% of their content exclusive to paid subscribers. Substack also offers features to include audio content and have readers comment on posts.
Substack was founded in 2017 and went through YC in winter 2018. They raised a $2M seed round in April of 2018 and raised a $15.3M Series A from Andreessen Horrowitz in July of 2019. The raise from Andreessen was a bit of a surprise. While they didn’t list a valuation, it did solidify expectations of growing quickly into a very large business.
Substack makes money by charging 10% of all subscriptions. If a creator charges$10/mo/sub, Substack gets $1/mo/sub. They don’t charge for accounts that are free, regardless of how many free subscribers.
The reason I find this company so interesting is it doesn’t seem like a venture-backable business at first glance. In order to get $100M in revenue, they would have to have $1.0B in gross subscription volume. When they went through YC, they had an average revenue / subscriber of $70. If that stays the same, Substack will need about 14.3M paid subscribers to break $100M in revenue.
Let’s benchmark this 14.3M paid subscribers looks like against some other paid publications:
Additionally, some user numbers from other kinds of media that compete for consumer attention:
All of these are different from Substack. But they are paid media subscription models that have worked for consumers in the past. They are useful in rough benchmarking.
Substack has a business model that creates unique unit economics. Substack isn’t trying to build the next The Economist or The Athletic. It’s not branded; the consumer doesn’t need to know what Substack is. Instead, the growth will come from consumers resonating with individual creators.
Because consumers subscribe to individual newsletters, Substack has different unit economics than other publications like The Athletic. In a branded publication, operating leverage is gained both from providing a software infrastructure layer and also having salaried writers. As The Athletic adds more subscribers, the company captures more incremental value because a higher portion of their expenses are fixed.
In venture-backed businesses, operating leverage is extremely important. Increasing marketing spend can often increase users, but if increasing users doesn’t convert to cash flow, it’s probably not worth it to spend the marketing dollars. Ultimately, operating leverage is a measure of how much of your costs are fixed vs variable. Substack has lower operating leverage than The Athletic because as subscribers grow, they have higher variable costs tied to subscriber growth.
That said, more value flows to individual creators on Substack. This mechanism will ultimately fuel their growth. Substack is less profitable for each new subscriber, but the creator can make a lot more. This makes the platform more attractive to creators who have a strong personal following.
Email is a unique medium. The choice to build Substack on email is not insignificant. The email inbox is a very personal space. Subscribing to something and not churning is a strong sign that content is valuable to the reader. It’s a much more powerful signal than a follow on social media, a view on an individual post, or a listen to a podcast. I don’t have data on this, but my guess is the conversion of free to paid email subscribers rivals any other medium. Also, while podcasting is rising in popularity, it’s still hard to share over text.
Substack doesn’t charge writers for creating free newsletters. This enables writer growth and additional reach through hobbyists or people just getting started. For creators with free subscriptions, it gives them a way to have an easy-to-use backend for no cost. By setting a minimum of $5/month on the paid side, it eliminates the writer tendency to price too low. In this case, if a creator can get 1,000 subscribers at $70/subscriber/year, they can make around $63k a year in net revenue, right around the median salary in the US. This is great for the consumer as niches are often disproportionately valuable to a small number of people. It’s also great for the creator because they can sustainably do something they love in a system that has aligned incentives.
One of the reasons why an investor might pay for rapid growth is network effects. It’s an overused term these days, but it describes when the value of each additional user increases the value proposition for other users.
Email subscriptions don’t have network effects by themselves. Right now, there isn’t a large switching cost for a creator to leave Substack – the creator owns their email list and content. The interesting part, however, comes in when the platform provides enough value that a writer would lose subscribers by switching platforms and when subscribers would lose value by switching newsletters.
Currently, Substack has a feature called Community that allows subscribers to comment on posts. This is valuable for both sides: if subscribers are willing to pay for a writer’s email, they likely have a lot in common with other members of the subscription list. And the writer creates even more value for readers, making the subscription stickier and more valuable.
If Substack can get a enough writers and subscribers on their platform, it starts to look like a high-quality, ad-free version of Facebook Groups or Reddit. The real power comes when a platform combines these communities. While some of us are a part of one-off communities now, if they were all integrated in the same interface, it could become our default online third space. With enough topics and a high enough quality of community, Substack could create a habit to come back over and over again.
While the exact product decisions will depend on how the platform evolves, it’s not hard to imagine a few features to increase value between communities:
Network value is derived on the writer side in a few ways. First, it can lower customer acquisition costs by being apart of the network. The network also decreases churn by having other reasons the subscribers use the platform.
Network value is derived on the user side by aggregating niche topic interests and connecting readers with each other across platforms. On Twitter, for example, it’s difficult to display the breadth and depth of interests to someone that comes across your profile. By having a profile that shows which communities you pay for and how active you are in them, there is a higher chance of creating connections with other readers.
At first it seems ridiculous that people would not only switch attention from modern social networks like Facebook and Twitter, let alone pay for it. But the catalyst could be when the highest quality writers and dedicated niche communities (that can’t exist on the ad-supported internet or choose to write in private) are only available on Substack, it becomes much more compelling.
It’s these network effects that prompted Andrew Chen to pull together Marc Andreessen and other a16z content professionals to meet with Substack to get a term sheet signed in less than a week. There certainly will be challenges, but the potential to build a billion-dollar company with a sustainable moat is real.
There are a number of risks with the investment:
Despite these risks, it makes a ton of sense from a16z's perspective and I would’ve invested as well. These risks are all risks that they want to bet on, and the Substack team is talented, motivated and prepared for the challenge.
Join my newsletter! Each week I breakdown interesting finance and investing topics. I put in hours of research so that you can spend minutes learning. Unsubscribe at any time.