The Future of Substack and Newsletter Publishing Platforms

May 16th, 2023
Co-authored with Lucas Romualdo
Link to a fully cited version of this paper

Product-Market Definition

Our market of interest is email-based newsletter publishing platforms, specifically English-language content for U.S.-based readers and content creators. These platforms distribute written content to a list of voluntary—sometimes paid—subscribers via email, and are distinctive as a bundle of services. The core of this bundle are email distribution, listserv management, newsletter construction and formatting tools, website hosting, and payment processing; differentiations of this product could include promotional services, analytics, or an advertiser marketplace. These services take advantage of specialization, outsourcing these settings to platforms with more experience doing so as opposed to the writers with less technical experience and interest.

There are other services that are on the border of this product-market. For example, blogs can publish on WordPress and Medium, which are primarily web hosting services. While these services also allow writers to collect paying subscribers and send emails, their offerings are not optimized for newsletter creation and distribution, which leads to a somewhat different creator base who depend more on search engine optimization and advertising revenue. Meanwhile, many larger news and publishing organizations, or content creators, are vertically integrated with respect to newsletter publishing: Morning Brew and Bloomberg, for example, offer email newsletters, and manage listservs, email distribution, and payments in-house.

Thesis

Newsletter platforms have disintermediated traditional media such as newspapers, allowing content creators to connect directly to their readers through email delivery and readers to create an a la carte selection of writing to read as opposed to a bundled newspaper publication. In recent years, newsletter platforms have grown in popularity, attracting well-regarded writers from newspapers, fueled by an expanding market of readers who are moving away from traditional written media. Platforms’ recommendation features have also fueled newsletter subscription growth. Currently, in the industry’s early stage, readers still add newsletters as opposed to replacing them. To writers, recommendations feature on platforms like Substack creates complementarity with other newsletters: more newsletters on the platform increases the number of newsletters recommending yours; combined with a growing platform readership, this will grow your subscribers.

However, the number of writers will grow over time: the rise of AI-assisted writing lowers the cost of producing an argument and synthesizing information, and will lower the barrier to entry for newsletter writers, while existing writers on other mediums will migrate to newsletters. As the number of newsletters increases, readers will near their reading capacity and the newsletter market will begin to saturate. To readers, newsletters will become increasingly substitutable: most new subscriptions will have to replace an old one, especially if those subscriptions are paid. 

This substitutability will reduce the upside of the recommendation feature to large creators, as benefits of subscriber growth through Substack’s network will not cover Substack’s higher price compared to other platforms. Because all large platforms have exportable email lists and creator-owned Stripe accounts for paid subscriptions, which allow hassle-free platform migration, large creators will migrate off of Substack. Meanwhile, Substack will use its network to service smaller, newer creators seeking a platform to act as an incubator and help them to ‘break out’ and ‘graduate’ from Substack once their subscriber base becomes stable. Thus, the market for newsletter platforms for mature creators will be monopolistically competitive, earning zero long-run economic profits, while Substack will earn positive profits as an incubator.

Industry Overview

Newsletters disintermediate writers from publishers and allow readers to create their own bundles of written content. Throughout the 20th century and early 21st century, most of the written content was published by newspapers, which distributed writers’ content to readers as a bundle. Newspapers are vertically integrated with ‘upstream’ content creators, where readers are ‘downstream’. Directly connecting to readers was expensive as delivery could not take advantage of the ease of email and online subscription, and bundling content meant that readers could pay for one newspaper and get all of the written content they wanted to read; however, readers with specific preferences had to pay for the whole bundle. From an operational standpoint, newspapers have historically earned revenue through subscriptions and advertising. 

The email newsletter publishing platform industry evolved from blog publishing platforms in the early 2010s. At the time, most writers were either employed on salary by larger publications, and thus had limited economic upside in their work, or were independently publishing through blogs. However, the success of these blogs was limited by their less sticky readers, and relied on relationships with advertisers for revenue. Newsletters allowed writers to bypass these issues by sending written content directly to reader email inboxes instead of having readers actively seek out a blog’s website, and using subscriptions for revenue to increase reader stickiness—only readers who valued the writing highly would pay for it. Writing newsletters therefore became an attractive option for writers to own their upside because of distribution permitted by newsletter publishing platforms.

Newsletter publishing platforms thus emerged as a way to reduce costs for writers, the platforms’ consumers, providing a bundle of necessary services for writers: email publishing, listserv management, payment processing, website hosting, and newsletter layout. Platforms also include various additional features as a means of differentiation, such as advertising support, analytics, and, recommendations and network elements. In particular, recommendations features, which allow the platform or newsletters to recommend other newsletters to their subscribers, have become increasingly important in the industry and are the defining feature of its leading player, Substack. 

It is important to note that not all newsletter publishing services are independent, and there are substitutes for these services. For example, many large publications which already have payments and email distribution systems do not outsource newsletter distribution to these platforms instead using their own systems; the same is true for very large newsletters businesses such as Morning Brew. This offers one possible substitute for newsletter platforms: purchasing all of the individual services bundled by the platforms and distributing emails in-house. Meanwhile, as mentioned, blogs and writing for a publication are also alternatives to email newsletter publishing for newsletter writers; others, such as podcasts, are both substitutes and complements to newsletters (a writer could run a podcast show instead of writing, or provide both to their subscribers). While these substitutes are relevant to determining demand in our market of interest, they are still adjacent markets.

Ghost, the first independent newsletter publishing service, was founded in 2013 as an open-source nonprofit platform. Ghost currently has 19,735 writers on its platform, earning an average annual revenue of $310 per writer. It offers four versions of its publishing service—starter, creator, team, and business—which are priced at a fixed annual or monthly fee for writers based on their number of readers and have increasing capabilities with each tier. As Ghost’s price to writers increases with the number of subscribers, though at a decreasing rate, it is effectively a variable cost to writers; the different versions of Ghost are a form of price discrimination. Ghost also has no network effect: there is no specific way for subscribers to one Ghost newsletter to find other newsletters that use Ghost. Moreover, because Ghost is open-source, writers can self-host Ghost’s software on another service, such as Digital Ocean, instead of paying Ghost for its newsletter publishing. However, this activity would be considered outside of our product-market, as writers, in this case, are vertically integrating with their publishing service by self-hosting Ghost’s bundle of services.

Beehiiv, another newsletter platform, was founded in 2021 by former employees of Morning Brew, a business newsletter. The platform offers similar features to Ghost. Like Ghost, it charges writers based on a monthly fee and offers three product versions with various levels of customization and analytics tools, called Launch, Grow, and Scale. Two features differentiate Beehiiv from Ghost: first, having already cultivated relationships with advertisers, Beehiiv offers an advertising marketplace for writers, allow writers to more easily place ads in their newsletters and earn non-subscription-fee revenue. Second, Beehiiv has a network element, allowing writers to recommend other Beehiiv newsletters to subscribers. For example, Beehiiv offers a service, Boost, which acts as a paid referral process for writers: writers who recommend a newsletter are paid for every one of their subscribers that also signs up for the recommended newsletter. However, Beehiiv does not have a central website where readers can view all Beehiiv newsletters, making it harder for new Beehiiv writers to take advantage of the reader network. In terms of pricing, the Launch service is available to writers until they reach 2,500 subscribers and Grow until 10,000. Notably, Launch does not allow creators to charge users for subscribing or generate advertising revenue through Beehiiv’s ad network. Launch is free to writers, while Grow costs $42/month and Scale $84/month billed annually, regardless of number of subscribers, such that Beehiiv’s service is effectively a flat, fixed cost for writers.

Of the major newsletter platforms, Substack is both the largest and most distinct, founded in 2017. Unlike the other major platforms, Substack does not directly charge a fee to writers, instead taking 10% of writers’ subscription revenues. While Substack charges a fee only on paying subscribers, Ghost and Beehiiv charge creators for the total number of subscribers. On one hand, Beehiiv’s cost to creators is essentially flat with respect to number of subscribers, and Ghost charges higher prices with more subscribers at a decreasing rate. On the other, Substack’s is linear and thus more expensive than the other two, though this pricing concerns only the number of paying subscribers. In the chart describing profits by platform (Appendix A), one can see how, as the number of subscribers increases, a newsletter’s profitability is increasingly lower for a writer on Substack compared to the alternatives—a writer with 200,000 subscribers would earn $233,000 less on Substack than on Ghost every year. This is also highly sensitive to the percentage of subscribers paying for the newsletter: for newsletters with more paying subscribers, Substack must provide much more value through its unique network in order to justify the cost. As a result of these differences, Beehiiv’s pricing model most favors newsletters with more subscribers (paying or otherwise), while Substack’s is more favorable to writers with fewer paying subscribers, which would presumably imply fewer subscribers overall. Pricing thus introduces a key tension between large and small creators, based on the number of subscribers.

In addition to its unique pricing model, Substack also differentiates itself through a more robust network, through recommendations and its website. Recommendations can be thought of as a search device. Many Substack readers are introduced to the platform by signing up for a particular newsletter, and would not otherwise explore the newsletters on Substack. The recommendations feature, which asks users to subscribe to similar newsletters while they sign up for the first, facilitates this exploration to the wider Substack network. In this sense, creators are complements—subscription to one creator is associated with a subscription to another creator. Readers do not know what they want to read and the recommendations feature helps them figure this out. An important feature is that the default recommendation feature is writer-controlled: when a reader subscribes to a creator, the creator chooses what other newsletters to recommend. A reader will not get recommendations outside of what at least one of the creators they are subscribed to recommends.

Substack’s recommendation feature is similar to Beehiiv’s, but is more powerful because Substack, as a first mover in network-based newsletter platforms, has invested more in acquiring creators and readers. Substack's recommendation feature drives 40% of subscriptions and 20% of paid subscriptions on its platform. This is aided by Substack saving user information in one account, which allows readers to more seamlessly subscribe to other newsletters once they have subscribed to one. On Ghost and Beehiiv, meanwhile, subscribers must re-enter their information every time they wish to subscribe to, and pay for, a newsletter, which adds an element of friction to the network. Thus, in addition to the bundled services provided by each of these platforms, writers gain the additional benefit of subscription growth through the recommendations feature, helping to justify the higher price of Substack to creators with larger newsletters.

To build a network of readers and writers, Substack began by attracting writers with large and established followings to the platform, providing them advances on revenue, with the purpose of drawing in readers. Substack readers would then become part of Substack’s network, making Substack attractive to writers who could rely on the network to drive subscriptions through recommendations or the exploration page. Substack’s network creates an entry barrier for other newsletter services using a network-based strategy. According to Stigler, a barrier to entry is a cost of producing (at some or every rate of output) that must be borne by firms seeking to enter an industry but is not borne by firms already in the industry. The entry barriers to the core services in this industry are somewhat low, but they have nonetheless deterred a flood of entrants into the market. For platforms less focused on cultivating a network, there are also entry barriers in the sunk costs of product development, namely incumbents’ sunk investments in creating numerous features for writers that potential entrants have not sunk, such as added customizability and analytics tools. For those platforms seeking to build a network through recommendations, entry barriers are higher: the size of both the reader and writer network can serve as an entry barrier, as incumbents with a large network have already sunk the costs of creating and maintaining its relationships to creators and subscribers, which potential entrants could not easily replicate without incurring significant fixed costs. AI-assisted code writing tools like Github Copilot lower those barriers by decreasing the cost of producing code. As a result of these entry barriers, the market for newsletter publishing services is an oligopoly, with product differentiation through network size and features offered to writers.

These entry barriers are exemplified by two big-tech names which briefly entered the newsletter platform space: Twitter and Facebook. Revue, a Dutch newsletter platform, was acquired by Twitter in January 2021. Revue was known to provide a more simplified service, with fewer customization options, a trimmer web hosting service, and fewer promotional and analytics services than Substack; this was reflected by Revue’s pricing, a 5% cut of subscription revenues compared to Substack’s 10%. Initially, Twitter lightly integrated Revue into its social media platform, allowing Twitter users to explore newsletters on Revue through a separate tab and to subscribe to Tweeters’ newsletters directly on Twitter, providing some of the promotional features Revue previously lacked. Twitter’s entry into this market, however, was unsuccessful: less than two years after the acquisition, Twitter shut down Revue. Facebook had a similar experience as a newsletter platform: in June 2021, it launched newsletter platform Bulletin. Mirroring Substack’s launch, Facebook recruited a group of independent writers on various topics to migrate to their platform; Facebook did not charge these writers for the use of their service. However, unlike the other platforms, Facebook also restricted the set of writers, requiring an invitation to write over Bulletin. Like Revue, Bulletin shut down in October 2022, ending big tech’s foray into newsletters. 

These two entry failures exemplify the presence of entry barriers in the newsletter publishing platform market. On Twitter’s part, many well-known writers are also prolific Tweeters, but Twitter was unable to successfully attract these writers, already on its platform, to its easily-accessible newsletter service. In the case of Tweeters who already had newsletters, migration was seamless, and Revue’s 5% cut of subscription revenue was lower than Substack’s 10% cut. In spite of these advantages, Revue’s simpler features indicate that sinking investment costs in features for writers creates a significant entry barrier. In the case of Tweeters who did not already have a newsletter, the same issue of features would apply, but the issue of network quality is also at hand—Substack could attract new writers with its existing readers through the recommendations feature. While Twitter was able to suggest newsletters to a writer’s Twitter followers, conversion was weak, suggesting that Twitter’s readers had different preferences than newsletter readers. Intuitively, written content of 140 characters should attract consumers with preferences that differ from consumers who read long-form newsletters. Thus, having a ‘network’ was not sufficient to bring writers in—it had to be a network of newsletter readers. The case of Facebook speaks to another key entry barrier: attracting writers. Facebook initially tried to build Bulletin by paying over 100 writers to produce content for the platform. The failure to execute this plan suggests that either Facebook’s user network had different preferences than a newsletter platform’s readers, as with Twitter, or that the writers were not able to bring readers in. This seems especially prescient because while Facebook attracted writers with experience and expertise, they did not attract writers who necessarily had followings, which is necessary to build the reader network. These two cases substantiate the presence of entry barriers of sunk product development costs and attracting the appropriate readers and writers to a platform’s network.

Networks & Two-Sided Markets

Substack’s main differentiation from other newsletter publishing services is its network. Services like Ghost and Beehiiv offer publishing services as production inputs for writers, who use the platforms to distribute their content to end users, or readers. Beehiiv has some network elements, such as a recommendation feature when you sign up for a newsletter, but are largely distribution services that connect individual readers to individual creators, rather than connecting its set of creators to its set of readers. In this respect, these publishing services are essentially an upstream input for creators.

On the other hand, Substack operates as a platform that connects its creators to its readers. On Substack’s website, readers can ‘shop’ for different newsletters based on their interests, while its recommendations feature creates complementary effects between creators through sharing readers. Unlike other publishing services, where writers acquire readers individually and use that revenue to pay the publishing service, many writers on Substack rely on Substack’s reader network to grow. Similarly, some Substack readers are on the platform to find new writers. Thus, Substack’s platform has elements of a two-sided market. Substack’s pricing model charges only writers, with readers’ access to Substack ‘subsidized’ by the take rate from writers, who in turn fund this through subscription fees from readers. As a result of their network-based model, Substack is reliant on cultivating a reader base; a reader base is what increases writers’ demand for Substack’s network. On the other side of the market, Substack’s ability to generate profits depends on their ability to attract writers with paying subscribers, meaning that Substack is incentivized to promote the entry of writers whose content can fill niches and attract sufficiently price-inelastic readers who are willing to pay for a subscription to those newsletters. Therefore, Substack not only depends on the size of its network to generate revenues, but also its strength: the more the platform can induce readers to become paying subscribers, the higher Substack’s revenue. Available data would support this assertion: the recommendations feature, or network aspect of Substack, drives 40% of subscriptions on the platform and 20% of paid subscriptions. Readers are heavily influenced by Substack’s recommendations and writers depend on Substack’s network for a sizable portion of their revenues, demonstrating that network size is essential for Substack’s success as a publishing service and platform.

In a network business model, platforms often offer introductory pricing, allowing the platform to grow its reader and creator network, or its installed base. Once the network reaches sufficient size, the base becomes sufficiently locked in—users do not leave, as other platforms have less content, and creators do not migrate to another platform for fear of slower subscriber growth. Thus, the platform is now able to extract more surplus from its users and creators, such as through higher prices for access to the network. Beehiiv has applied elements of introductory pricing, with prices much lower than its better-established, extremely-similar competitor Ghost, that are also flat relative to the number of subscribers, giving an extra cost boost for larger creators. Substack is building its network through introductory pricing—the service grew a lot, for example, by signing well-known writers with followings who could induce paying subscribers. Substack has offered sweetheart deals to these high-profile creators, giving them an advance on subscription revenues. This would help Substack build a large network of creators who will draw in a larger network of users; in essence, building an installed base. Creators themselves often offer ‘introductory pricing’ to their subscribers by allowing them to subscribe for free, with the hope of converting them to paying subscribers later on to access gated content. Because Substack does not charge writers for free subscribers and only takes revenue from the paying subscribers, this is a form of introductory pricing by Substack to writers (compared to Ghost and Beehiiv which charge for free subscribers). When subscribers convert from free to paying, Substack then collects more of that marginal revenue than Ghost or Beehiiv, increasing prices to creators once their readers become willing to pay. Allowing the ‘gifting’ of one-month subscriptions of certain newsletters by current paying subscribers is also a form of introductory pricing on Substack by both the platform and creators to readers, which is done in the hope of gaining a paid subscriber. Beehiiv is pursuing a different form of introductory pricing. Its “core” newsletter platform service costs much less than Ghost and has a constant price until 200,000 subscribers, which for a large range of subscriber values is less than the cost of hosting.

The final step of a network, locking in subscribers, is also increasingly becoming a feature of these newsletter platforms. Substack, for example, has placed greater emphasis on users reading newsletters through its mobile app, instead of email. Technically, this would fall outside our product market of interest, because users’ incentives to check apps are different than receiving emails, but it has vital implications for Substack’s network strategy. Rather than having users check their email for newsletters, where an email from Substack is equivalent to one from Beehiiv or Ghost, Substack is pushing for readers to access content through Substack’s app, making those newsletter readers more exclusive to Substack and effectively ‘locking’ them into the network. This approach is debatable, as readers might not check Substack’s app as often as their email, but there is also the benefit for a creator of using Substack’s popularity as a platform to build their own individual readership. However, newsletter readers are likely to check their email regardless, meaning that the switching cost of reading through an app is likely higher than continuing to use email, and this attempt to lock in readers may be unsuccessful. When Substack first released its mobile app, by default it paused email notifications for subscribers who installed the app “to prevent duplicate notifications.” However, the attempt to lock in readers and writers to Substack’s platform encountered criticism and the change was quickly reversed, thereby keeping email notifications intact.
Another way in which Substack is trying to close its platform is the introduction of Substack Notes, a platform for Substack writers to publish short-form, public content; essentially, an in-house Twitter. This is especially notable because Substack, and all of the major newsletter platforms, are heavily integrated with Twitter, allowing writers to seamlessly promote content through their Twitter accounts. A recent spat with Twitter meant that Substack newsletters could no longer be easily promoted on the platform; instead, Substack Notes allows writers to engage with their readers, and the wider public, in a similar format as on Twitter, but within the Substack platform. Other newsletter platforms are still integrated with Twitter; moving short-form content to Substack would limit readers’ time to engage with the newsletter writers not on Substack, who promote their content through Twitter. Thus, readers and writers will be further locked into Substack’s network. However, as with the email example, it is unclear that Substack’s effort would be successful: Twitter currently has far more readers than Substack, and far more prospective subscribers to newsletters than those who are actively searching Substack Notes. Substack’s weakened integration with Twitter may favor other newsletter platforms instead of locking in their base. 

However, Substack is fundamentally unable to lock creators in. Substack, as well as all other major newsletter publishing services, allows writers to easily migrate their content and email lists (as all payments in this industry are handled by creator-owned Stripe accounts, those can be transferred easily as well). Thus, Substack has no way to effectively lock creators in. If a creator determined that the benefits of being attached to Substack’s network were less than Substack’s cost premium over alternatives, they could migrate their subscribers to Ghost or Beehiiv, earning equal revenues at a lower fee. Readers, meanwhile, would still receive the exact same newsletter. Thus, readers are not harmed, and creators lower their costs while losing the benefits of the recommendation feature to gain new users. This would indicate that Substack has no route to extracting more surplus from current creators, as any increased price for users or creators would have to be offset by the growth benefits of its recommendation feature.

Reader Demand Saturation

Substack’s network dilemma leads us to consider how Substack could raise prices without creator flight. The premise of our thesis is that over time, reader demand for newsletters will saturate; alternatively, readers have a limited capacity to read, and hence, writers will increasingly compete with one another for readers’ attention.

To understand the drivers of saturation, we can consider the market for email newsletters, specifically Substack writers. Entry barriers in content creation are fairly weak—literacy in the U.S. is high and fixed costs are low. Moreover, in an unsaturated market, readers are still willing to add newsletters that suit their preferences, meaning that the readers ‘acquired’ by existing newsletters does not preclude an entrant from ‘acquiring’ those same readers. The rise of AI-assisted writing, such as with ChatGPT, will also enable many more writers’ entry and for each writer to produce more newsletters. Individual writers will therefore be able to produce much larger quantities of writing at a lower cost of production, thereby ‘flooding the market’ of newsletters without falling into unprofitability. Given these low entry barriers, and obvious differentiation, the newsletter market would be monopolistically competitive, meaning that short-run profits can be earned through product differentiation, but copying that differentiation would prevent long-run profits. However, this comes with a caveat: part of a newsletter’s differentiation is its author’s name and reputation, which cannot be easily replicated or replaced by an entrant, providing some avenue for existing newsletters to build long-run profits. 

However, this differentiation is only successful insofar as there is demand for such variety, or heterogeneous consumer preferences. To a certain extent, readers will have exogenous preferences for what they read based on their personal interests; these preferences may also exogenously change with varying degrees of frequency, depending on that reader’s strength of preferences. Horizontally, newsletters can differentiate in both topic and style, and, as a result, will attract different sets of readers. From this perspective, heterogeneity among readers should not saturate a market: as long as a new creator can find a new topic or style niche, which can vary significantly, they can earn short-run profits and will thus enter the market. However, there are also some limits to this heterogeneity: while writing style and topics can vary almost infinitely, the newsletter format more broadly is decidedly longer-form. For example, a Beehiiv newsletter is not a close substitute for its author’s Twitter feed because the depth of content and frequency with which it is delivered differs, which affects the informational value to the reader.

There is also an aspect of vertical heterogeneity, concerning willingness to pay for newsletter content. Americans’ wealth distribution will imply reasonably broad heterogeneity; this willingness to pay will also restrict quantity, as Substack requires writers to charge readers at least $5 per month for a paid subscription. Thus, readers’ willingness to pay for newsletters overall will also be limited by their willingness to pay for each individual newsletter: even if someone’s willingness to pay for newsletters overall is a total of $100 per month, if they do not value any individual newsletter over $3 per month, they will spend a total of $0 instead of $100. This may somewhat limit their vertical heterogeneity.

From a time perspective, there is also a very clear upper bound on newsletter readership. In terms of reading broadly, Americans read less every year, reading for pleasure less than 20 minutes a day in 2021, down from 23 minutes per day reading in 2004. This has not been supplanted by reading news: estimated newspaper circulation is at its lowest point since at least 1940, with 13 million papers lost from 2014 to 2020. Online news readers increased by only five million in that same period, suggesting that some eight million news readers were lost in those six years and not to reading books or magazines. Moreover, the average minutes per news website visit is down 30% since 2014, suggesting that Americans are also willing to spend less time reading news. Newsletters are unlikely to have filled this gap, with Substack having only been founded halfway through this period and, in 2020, having only 250,000 paying subscribers, implying a total readership between two and six million, given that typically, 5-10% of Substack readers pay. There is therefore an undeniable decline in reading in America, which not only suggests that Americans have a binding reading capacity, but also a decreasing one. Thus, readers’ heterogeneity in willingness to pay for newsletters is limited by both Substack’s pricing scheme and their time constraint. Some of this vertical heterogeneity is expressed by how writers will offer both free and paid content — in effect, price discrimination, based on readers’ different willingness to pay.

Horizontal heterogeneity means that any prospective newsletter entrant with an appealing differentiation can earn short-run profits and will thus enter. To do this, new newsletters must seek out a niche to appeal to that heterogeneity and not enter into a niche already dominated by other newsletters; it is further important to note that more general topics areas appeal to a wider range of readers, and thus are ‘taken’ by earlier newsletters, which are likely to be larger by having a broader focus that appeals to more readers, and more time to acquire subscribers. Hence, new newsletter content will be increasingly narrowly-focused with respect to topic or style in order to achieve that niche differentiation and profitability. Meanwhile, this increasing number of newsletters will compete with a readership with bounded vertical heterogeneity, where each additional newsletters increases a reader’s likelihood of removing another. Thus, readers will only want to add newsletters that appeal to their preferences more so than their existing newsletter portfolio, favoring the niche over the generic, or the new over the established. 

This trend will be especially tied to recommendations features. Recommendations attempt to offer newsletters based on readers’ expressed preferences, and is therefore most effective when it offers a newsletter that a reader is more likely to subscribe to. Suggesting a larger newsletter with broader content is unlikely to ‘beat’ the existing newsletters in a reader’s portfolio, especially if that newsletter is older and the reader is likely to have seen it. However, suggesting a newer newsletter with a more focused topic, if it appeals narrowly to a reader’s specific interests, is more likely to prevail in replacing one of their existing newsletters, and is more likely to draw a reader’s excitement, which may also entice them into paying. The recommendations feature, therefore, will be more likely to promote niche content over more general, established newsletters. Currently, the smaller number of newsletters means that readers have less of an idea of what content is available and how much content they have capacity for. Thus, they are more willing to add, rather than replace, newsletters, so general newsletters with large readerships are not threatened by recommendations features promoting newer content, and fewer new newsletters exist to be recommended at all. However, the maturation of the newsletter industry will give readers more information about how much content they can handle in the long run, meaning that recommendations will increasingly focus on newer, niche newsletters targeted at beating a reader’s existing newsletter portfolio, further favoring newer newsletters over larger, established ones.

Increased Substitutability & Flight

The aforementioned dynamic speaks to a growing tension between large and small creators, defined by number of subscribers. Currently, readership on Substack is growing; the recommendation feature drives much of this growth. Because other writers choose the newsletters to promote to their subscribers, creators are complements to other creators. The more other writers on Substack promote a writer’s newsletter, the more subscriptions and revenues the writer earns. In this non-saturated market, readers also view newsletters as complements, rather than substitutes, as they are willing to add more newsletters without replacing them. With the newsletter market still in a growth stage, subscribers have not reached their reading capacity for newsletters, meaning the market is not saturated. Thus, readers are unlikely to substitute one newsletter for another. Furthermore, the newsletter market has a lot of differentiated content (recipes, political commentary, short stories, etc.) and, currently, not many newsletters, meaning that readers do not view most newsletters as close substitutes, increasing reader willingness to add other newsletters they view as unique.

However, because the newsletter market is monopolistically competitive, more and more newsletters will enter with more and more niche differentiations; then, readers will increasingly meet their capacities and become less willing to add new newsletters, unless an option is more appealing than the portfolio of newsletters they have already settled with. And while the number of free newsletters a reader is subscribed to is not as binding a ceiling as paid newsletters because readers can skim or ignore some newsletters, the money spent on paid newsletters creates a binding constraint, making paid newsletter subscriptions more substitutes than complements to readers. The recommendations feature becomes relevant here: given the reasoning discussed above, recommendations will promote newsletters to readers that are increasingly niche and newer, thereby promoting smaller creators more than large ones.

The issue of reputation also speaks to this tension: large creators who migrated their followings to Substack lent the platform credibility, which new and small creators become able to take advantage of with no additional benefit to the large creators. Substack’s pricing model also creates a divide between large and small creators. For small creators, a take-rate-based payment for Substack’s services is favorable: rather than paying for newsletter services upfront, and hoping that subscriber revenues will exceed that fixed cost, writers can generate revenue first and pay a portion of that to Substack later. This method is also likely cheaper for those creators who have a small subscriber base: an upfront fee usually exceeds expected revenues, preventing writer entry. Large creators, on the other hand, would pay less under Ghost and Beehiiv’s fixed fee model than the 10% take rate. Because Substack’s higher price to creators is justified by its network through its recommendations feature, large creators will therefore get less value out of Substack’s unique feature than they pay for, while the opposite is true for the smaller and newer creators. Thus, large newsletters will have an incentive to move to a cheaper platform without a strong recommendations feature in order to maximize their value.

Substack’s Future

As all large platforms allow hassle-free platform migration through exportable email lists and creator-owned Stripe accounts for paid subscriptions, large creators will migrate off of Substack. 

Locking in large creators and extracting surplus from them could be a profitable approach for Substack. As discussed, Substack is already attempting to lock creators into its network with the introduction of Notes and mobile app, albeit with limited success. Substack could seek to build on this lock-in by offering Substack as a subscription to readers on the app, and distributing that extracted surplus to writers. However, this would essentially subject writers to the same model under which they were employed by newspapers and magazines, making this approach unlikely.

Moreover, Substack has made no attempts to block the ease of migration for creators, thereby limiting the degree to which those creators are locked into the platform. Any attempt to foreclose on creator exit by blocking migration would likely result in legal action related to Substack’s Terms of Service, which state that creators own the list of their readers’ emails; payment lists in Stripe are not managed by Substack and Substack is thus unable to foreclose on payment services either. Substack’s failed attempt to remove email newsletters for app users, met with writer backlash, suggests that similar attempts at foreclosure would fail. Moreover, forcing large creators to remain on a platform that charges a premium for a harmful differentiation, will discourage prospective writers from entry, realizing that they are doomed to have diminished long-run profits, reducing Substack’s future network size and thus its value to readers and new writers looking to Substack’s network to grow their readership.

Another course of action for Substack, which we argue is more realistic, is that Substack will allow the exit of its largest creators and instead use its network to service smaller, newer, more niche creators seeking a platform to act as an incubator and help them to ‘break out’ and ‘graduate’ from Substack once their subscriber base becomes stable. Readers of large newsletters can remain subscribed to that content, but their Substack accounts will remain and offer them a way to search for new content through recommendations. In this scenario, Substack has sunk the most investments of any platform into building a network; these sunk costs create an entry barrier that would prevent other platforms from attempting to compete with Substack as a network-based newsletter platform. Moreover, concentration on one platform may be welfare-enhancing: readers gain the most utility from a platform with the most writers and vice versa, such that one platform with all readers and writers delivers more total utility than two platforms segregating the reader and writer bases. Already, we can see that Beehiiv’s attempt to build a network has been met with limited success, though it is still a relatively new feature. This could be attributed to Beehiiv’s need to spend on significant network investment costs: Substack has already sunk the cost of attracting readers to its writer network and bringing writers to its platform, while Beehiiv would need to spend significantly to woo writers away from Substack to then build a reader base. This entry failure of a firm which already has some cost advantages over other potential entrants, as Beehiiv has newsletter publishing technology and expertise, suggests that the network-based newsletter platform market will be either a very small oligopoly (by number of firms) or a Substack monopoly.

In a monopoly, Substack would clearly earn positive long-run profits, incentivizing it to keep this ‘incubator’ business model. In the oligopoly case, Substack has a key form of product differentiation that will enable it to earn positive long-run profits in this sub-industry with entry barriers: network size. Substack already has 20 million readers on its platform whose emails will remain even when large creators leave. Thus, new writers will be attracted to Substack over other network-based newsletter platforms for Substack’s larger base of readers, or potential subscribers. Even if Substack does not gain a monopoly in network-based newsletter platforms, its large and sticky network will provide it a form of product differentiation to creators that would allow it to differentiate and escape Bertrand’s paradox in an oligopoly. Network size would presumably always be favored by new writers, such that Substack’s differentiation would remain popular and it could still earn positive long-run profits with an incubator business model.

The newsletter publishing services that large creators ‘graduate’ into will be a dissimilar market. Without the network element, there are no clear entry barriers: the sunk costs of current newsletter services like Ghost and Beehiiv are small, as these platforms bundle services that are mostly outsourced, such that they do not incur high fixed costs themselves. Thus, without significant entry barriers, the market for non-network newsletter services will be competitive. However, there is obviously differentiation in this market: for example, Revue offered a cheaper service with less customization features, while Ghost offers a more expensive service for more customization that especially appeals to technology-savvy writers. Hence, the market for newsletter services targeted at mature creators would be monopolistically competitive, with new entrants seeking short-run profits wrought by differentiation, yielding no long-term economic profits. This would create a bifurcated market for newsletters and their publishing platforms: one side catering to up-and-coming newsletters with network-based services and higher prices, while the other would cater to more mature creators and offer less of a network in exchange for lower prices.

Social Welfare

The bifurcation of the newsletter publishing platform market and the removal of gatekeepers in the form of newspapers and book publishers means that the best writers of a generation will have a higher chance of succeeding as writers than in the past. While writers would previously have had to work through a newspaper’s bureaucracy or another notable career to succeed as an independent writer, talented creators can now seek recognition directly through newsletter publishing platforms, lowering these reputation-based barriers to entry for writers. As a result, readers will benefit from increased entry of talented writers, improving the quality of the content they consume.

Newsletter publishing platforms make consumers better off because consumers choose a bundle that they enjoy the most, an a-la-carte bundle. An a-la-carte selection provides creators with an accurate signal of the market value of their newsletter, as paid subscriptions indicate market interest. Under newspapers, all the content was bundled, and newspapers often subsidized less popular content, and content supply did not accurately reflect reader demand.

Consumers now pay for a lot of subscriptions for all areas of life. In the past, most consumers had a limit to how much they were willing to pay for a newspaper, perhaps for one local newspaper and one or two national newspapers, and all of their news was bundled. Now, newsletter platforms promote subscriptions for each topic a reader demands. A similar bundle of newspaper content would cost readers much more on a newsletter service, as high-quality content is paywalled individually. Hence, there is a possibility that newsletters will end up extracting more surplus out of consumers than newspapers. That tension could result in a re-bundling of newspaper-like content on newsletters, like Bari Weiss’ The Free Press. This would limit the extent to which creators can extract surplus from readers: subscribing to a high-quality bundle is often a better deal than subscribing to individual newsletters. New business models, either managed through newsletter platforms or individual newsletters, might emerge as a result. Overall, while consumer welfare is improved in the short term, the medium-term impact of newsletter publishing platforms on consumer welfare is less clear, as increased attempts by creators to extract surplus from consumers in a saturated market could lead to a degradation of the quality of content available to consumers for a price similar to a bundled newspaper.

Conclusion

This paper highlights the rise of newsletter publishing platforms which directly connect writers to their readers through email, and the importance of recommendation features to creators. Currently, the creators compete in a positive-sum environment, as with growing reader demand for newsletters, readers treat newsletters as complements. However, as the newsletter market saturates and readers reach their reading capacity, readers will increasingly treat newsletters as substitutes. To increase reader reach, recommendations features will increasingly promote newer, more niche newsletters, which will lower the value of recommendations to large creators. These creators will then leave the expensive network-based Substack for cheaper alternatives like Beehiiv or Ghost, enabled by exportability of mailing lists and creator-owned payment accounts unique to the newsletter publishing industry. Thereafter, Substack will act as an oligopoly or monopoly incubator for up-and-coming newsletters due to its large network, while newsletter platforms without a large network will cater to established newsletters in a monopolistically competitive market.

There are some limitations to our thesis. It is not a given that the newsletter market will saturate: prospective writers, seeing the declining reading trends in the U.S., may be discouraged from writing at all, preventing higher growth on the writer side relative to reader growth. Substack may also attempt to lock in their writer base, rather than letting them migrate off of the platform, although this is unlikely based on past attempts. Lastly, from a welfare standpoint, there may be a possibility of coordination between platforms or between newsletter writers and platforms to raise prices.

A fundamental question that underlies this paper is how newsletters and written media in general will fare in the age of information abundance and shorter attention spans. Writer entry into the newsletter market will cause an abundance of written information that will undermine readers’ ability to search for newsletters without significant assistance, such as through recommendation features which would increase Substack’s power as a platform. Consumer apps are moving towards catering to shorter attention spans, exemplified by TikTok’s 150 million American users. Newsletters’ need to compete with other media for consumer attention may similarly push writers towards shorter-form, attention-grabbing content, perhaps aided by AI-generated summaries. This decline in quality and length will decrease readers’ willingness to pay for written news. Moreover, given an increasing supply of newsletters and a decreasing demand for reading overall, newsletters are operating in a declining market, limiting the long-term growth and viability of newsletters and their platforms.