Time for a new adventure
I’ve transitioned from being a full-time employee at Stripe into being an advisor. (As always, comments here are in my own capacity and I use no information that is not otherwise public.)
I am going to take a few months off to rest, reflect, and plot before my next professional adventure. One thing I've already decided: I have very much liked writing the past ~30 issues of Bits about Money. Many readers have told me they very much enjoy it. And so, I’m carving out time to do 20-26 issues this year.
If you find BAM routinely professionally useful, good news: you can pay for it, and watch it get even more useful. You can read more about memberships or use the following for the most popular ones:
(These include a discount of 20% for early adopters, through January 25th.)
I will not put a “paywall” in front of BAM. I broadly think paywalls are a terrible user experience and, at the margin, as a writer I would prefer increasing spread of ideas over maximizing for revenue.
It would be traditional at this point in infoproduct marketing to hit you with a very long sales page followed by the buy button and a P.S. Those tactics are used because they are brutally effective. But it would be far truer to form to instead dive into the fascinating financial infrastructure behind email newsletters.
What. You can’t not have seen that coming.
The long history of paid niche publications
Many Internet natives might not know that we’ve had paid newsletters for almost a century. The most vibrant and commercially viable subsector of them was boutique financial commentary. You’d get hot stock tips delivered by the postman and pay for them via check or, later, credit card transactions done on the card-not-present model.
One publication well-known enough by cognoscenti to get namechecked in The Big Short (movie edition) is Grant’s Interest Rate Observer. It has cranked out two issues a month since 1983. Grant’s is both for historical and brand positioning reasons heavily tied up with the personal brand of Jim Grant, and is stubbornly independent.
This is notable because there are two natural sizes for publishing businesses: boutique and gigantic. Another well-known financial newsletter business, MarketWise, runs 180 separate newsletters, many with their own figurehead beloved by their audience. (Byrne Hobart has an excellent writeup on their business.) This lets them cross-sell aggressively, but also lets them bolt on new products with very little marginal increase in organizational complexity. They have the economies of scale with regards to publications that other businesses which sell IP have with regards to bits.
Newsletters in the Internet age
One of the earliest successful Internet-native businesses was Motley Fool, which has been around since before the AOL days. They are, loosely speaking, an education/content/application empire which primarily monetizes by converting free users into paid email newsletter subscribers. They collect these free users with a PR strategy, great use of channels (and adaptations over the years as the winning channels have changed), voice, style, ability to SEO for every possible keyword relevant to a retail investor, etc.
Motley Fool has something which very few independent publishers, and surprisingly few publishers of any size, have: a web development team. They can amortize the cost of those developers, and the other overheads inherent in their business, over each additional cohort of users they recruit.
And back when Motley Fool was the young hip upstart, their web developers implemented something which was ridiculously difficult in 1997: you could buy a subscription from them, with a credit card, over the Internet. Generally speaking, at the time this took about six months of implementation work and $250k+ in hard costs, like licenses for various software you’d need to run it.
We’ll return to online credit card processing in a moment, but note the innovation here is about payment from the customer to the business. The business still has to pay a writer, in a fairly traditional way. Motley Fool uses some combination of full-time employees and freelancers, like many other publishing businesses.
I have no inside knowledge of their compensation structure, but I’d hazard a guess that their writers are like most employees and freelancers: they have relatively fixed upside from doing their job and are in some way accountable for results, and perhaps even shown graphs of business impact, but they are not tied to their results in the way that e.g. sales representatives are.
A startup employee might say “They don’t have equity in their work.”
The difficulty of many-to-many payments in publishing
Authors are historically compensated differently than writers. To oversimplify a bit, in the interest of not losing the narrative thread, authors do have equity in their work.
They sell almost all of it for a very small amount of money (the advance) to a publisher. The publisher arranges for a very small amount of value-added services (editing, cover design, etc) and then places the book into wholesale and retail channels. As those channels generate sales, the publisher gets paid money, which they allocate a small portion of to the author (in the form of royalties). The royalties first “earn out” (pay back) the advance and then start getting distributed to the author on roughly a quarterly basis.
As someone who loves the craft of writing words, I hate what publishers have done to my people over the centuries. As someone who can appreciate industrial scale accountancy, it is absolutely amazing to me that this entire system was made to work before spreadsheets, nay before electronic recordkeeping, nay before electricity.
If you talk with authors, they will have many things to say (few of them positive) about the experience of being one of thousands of contractual participants in the ledgers of their publishing house. But many of them have (sometimes surprisingly) good things to say about their publisher as a platform. They have relationships with the best editors. Their institutional imprimatur was career-accelerating outside the two covers of the book. Work with a publisher and you might someday get a polite call from Japan from someone who wants the translation rights so that an extremely skilled professional can laboriously translate your book and put it in ten thousand bookstores.
Platforms can be wonderful businesses, for all parties involved. And one of their value adds is frequently “We’ll take care of the money movement for you.” There exists a complex web of commercial relationships which connects your reader to their friendly neighborhood bookseller to a distributor to the publisher. You, as an individual author, are probably disinclined to manage money movement, and you would certainly not have the capacity to serially contract with bookstores to get your inventory onto shelves.
Internet businesses took platforms to the next level, by amping up the value to both sides of the creator economy (writers and readers, in the case of publishing) while frequently simplifying the transactional chain in the middle. This eliminated a lot of economic friction. That redivision of the pie has had some interesting implications, not just for how (mechanically) authors get paid, but for what business models they work under, and for how certain sectors of the economy expect to consume information.
We’ll return to the complexities of payments in a moment, but for now let’s get back to newsletters.
I have for many years remarked that the single biggest thing Substack accomplished wasn’t hooking a WYSIWYG editor to email sending infrastructure, implementing an optional paywall, or convincing authors to use it.
The single biggest thing Substack accomplished was legitimizing a Schelling point: interesting intellectuals can charge money for doing their thing and people who interested in the life of the mind should pay for their output. The Schelling point, the place at which these two groups would choose to meet without any coordination mechanism, was the writer’s Substack. (Alright, strictly speaking, the point they usually met was sometime before, and the point they met for the purpose of the instant transaction was when someone tweeted a link to their new Substack. Substack's existence is one of several billion data points that Twitter has not shipped the products that its users really want from it.)
Paying for niche newsletters was not an entirely novel proposition. Newsletters have been a huge business for a century. Email newsletters are probably older than the World Wide Web. There were many successfully commercialized email newsletters prior to Substack.
But Substack vigorously pushed the model to a point where… it became remarkably unremarkable, in certain parts of the Internet. Policy, tech, and financial circles discovered that while they liked their respective sections of Twitter they loved their respective Substacks even more. Many people began to routinely expense Substacks, just like people in almost every industry have long routinely expensed trade press and conferences. It was simultaneously revolutionary, incremental, and quickly faded into the landscape of intellectual life.
So three cheers for Substack. I am a frequent reader, though I run Bits about Money on a different platform (Ghost, for reasons discussed below).
In some corners of the tech industry, there is a sharp distinction made between marketplaces (where there is a good or service being exchanged and the software just facilitates the transaction) and platforms (where the experience of the software is the thing being sold). eBay is definitely a marketplace. WordPress is definitely a platform. Substack is a platform with some marketplace-adjacent elements.
Substack is a platform with a marketplace's monetization model. They connect supply and demand in an environment they mostly control. They are paid when supply and demand match. Part of the value of the business is network effects, which you have read much about before. Part is about building a wonderful product experience. Part is about methodologically bulldozing barriers to people deciding to run a newsletter, like their experiments with providing writers with healthcare coverage or legal defense.
And a key part of many platform businesses is payments.
Platforms want to have payments but don’t want to build or run payments
Platform businesses and marketplaces are brutally difficult to operate, chiefly because they combine all the standard startup challenges with “... and now do it all over again or all your successes are for naught.” This is because if you win just demand or just supply you don’t end up with half of a winning business, you end up with no business worth speaking of.
And so these businesses look at payments and see something they both want urgently to be solved and would like to avoid working on as much as possible. It is crunchy, involved, heavily regulated, and deep. The team likely has limited pre-existing advantage in e.g. banking relationships, understanding of regulations, etc. They likely consider payments as a very important detail but not a core source of advantage for the business; executing excellently on it won’t solve demand or supply by itself.
Payments also get fractally complicated when moving internationally in a way which many parts of running platforms do not. You don’t need to double the size of the engineering team the first time someone sends an email across the US/Canadian border. But the amount of complexity you’ll bite off for trying to pay a single author on the other side of that border from HQ is considerable.
And so platforms often end up looking to buy a solution rather than building one. I’m pretty partial to Stripe Connect, for all the obvious reasons, and it is handily the market leader here.
Connect is an extremely important product for Stripe, in a way which is perhaps less legible than the importance of Payments. Stripe is far better known for Payments proper. Everyone understands how Payments works from a customer perspective and can successfully intuit that the main thing a business gets out of it is “... and then the money shows up.” Even among long-time happy users of Stripe products, most businesses won’t have run platforms or marketplaces.
But Connect is pretty galactically important in the Stripe universe, and powers many of the businesses that are now considered synonymous with the Internet economy. (Newsletters are, FWIW, pretty far down that list.)
Why is this so hard? Well, credit cards (and other payment rails) are legacy infrastructure. The core motivating use case assumed that, in each transaction, you’d have one buyer and one seller. They did not successfully anticipate multiparty transactions. They did not successfully anticipate brokered transactions where a buyer and seller were brought together temporarily by a mutually trusted party.
These models have existed before. Publishers predate electricity. They solve the incredibly complex undertaking of building payment rails between every possible reader and every possible author by... punting.
That is strictly too hard to solve. Instead, put a few layers of trusted intermediaries between them. Have each act as if it is just doing 1:1 transactions serially. Then tie it all together with a hugely impressive legal and accounting apparatus created mostly by bilateral agreements with extremely limited visibility into the overall system. Victory!
And so when one wonders why newsletters “came out of nowhere” just a few years ago, I think part of it is Substack providing a catalyst for a change in user behavior, mostly on the demand side. (As Kevin Kwok has observed, after you've recruited sufficient people willing to pay money, professional supply will quickly come to dominate most of your market, drawn primarily by economic incentives.)
And I think another part of it is Connect enabling Substack to exist, without forcing Substack to be the counterparty to every subscription. That's a seemingly under-the-hood change with very large implications. If Substack was "selling" the contents of every Substack, it would be responsible for them, and a lot of risk mitigation practices that would entail would make it a much less open platform than it actually is. If authors were paid by Substack and not by users, they'd perpetually have to worry whether they were too tightly coupled to Substack. If it went away, they might lose a week of revenue (not the worst thing in world?) or lose the ability to charge the saved credit cards of all existing subscribers (extremely painful for a subscription business).
A deep dive into newsletter fund flows
So suppose you were hypothetically to subscribe to BAM. (Thank you for your support.) What would happen to your money?
BAM’s technology stack includes an OSS project called Ghost. It offers features comparable to Substack or ConvertKit, with the additional wrinkle that you could theoretically run it yourself versus relying on the service. (There exist many other services for sending newsletters; email is a gigantic business.)
Some people have asked me why I don’t use Substack to run a Substack. (I am extremely impressed that, as a relatively small startup, they took an activity that was a century old and branded it so thoroughly that “a Substack” is now thoroughly synonymous with “paid newsletter” in both directions.)
I’m an inveterate tinkerer and like the notion of being able to change the behavior of the site if the out-of-the-box version doesn’t work for me. And, in an unplanned bit of irony, after I left Stripe’s employ I went almost immediately into submitting free fixes for other peoples’ Stripe integration bugs.
If you sign up for a Bits about Money subscription, from the perspective of the technical and legal machinery for credit cards, you’re dealing with me. (Technically, my LLC.) Ghost is not a direct party to that transaction; they have simply written code to facilitate it.
Connect has a mode where a platform arranging a transaction could, without becoming an intermediary in the transaction, arrange to charge a platform fee for it. This solves for the main business model of many platforms and marketplaces businesses. Many software businesses like to index the price of their offering to customer success with it, to capture some of the vast customer surplus created by software while keeping it available to businesses of many sizes. Receiving a success fee for transactions is a very direct way to tailor pricing fairly exactly to customer success.
Interestingly, this is not the only way to gain economics from bolting payments onto a non-payments business. Financial services is frequently a game of splitting a margin pie to buy things that the business wants. One thing a financial service provider wants to do is grow. In lieu of spending margin on buying billboards in airports, a financial service provider could hypothetically incentivize business partners to recruit new users by paying them out of the margin gained on transactions from the users they recruited.
This allows for specialization between platforms and their financial partners. Platforms can focus on convincing more end users to buy The Thing, recruiting new suppliers, training them to be maximally successful, and acting as their support system for day-to-day questions, everything from “How do I log in again?” (you’d be shocked how often this is asked) to “I’m lacking the sense of community as an independent that I used to get from my newsroom. Do you have anywhere to mingle with other people who have same issue?”
Financial service providers can answer more focused questions like “When does my money arrive?" (you’d be shocked how often this is asked) and “Someone bought my thing from Germany and is complaining that my receipt is not compliant with their tax obligations aaaaaa what do I do now?”
You apologize to the user then point them at a dashboard where they can enter any arbitrary text they want to get printed on your invoices, and offer to reissue the invoice with that text on it. This gets them what they want without forcing you to become an expert in Fun New Issues In German Import VAT Treatment.
Every society shapes and is shaped by its infrastructure layers
Fairly pedestrian improvements in reconfiguring financial infrastructure, in changes in the flows of bits about money, unlock larger social changes.
One of the most interesting knock-on effect in the newsletter boom to me has been how it has changed negotiations between highly effective writers and the institutions that employ them. For a very long time, institutions had almost all of the market power, and chose to compensate their staff mostly in prestige rather than in money. The existence of Substack, and particularly the noted success of respected writers on it and the publicity around the leaderboards early, gave many writers a BATNA to the (fairly rigid and miserly) salary schedules at their institutions. You can credibly say "The newspaper business is contracting. You're lucky to have any job at all. $85k is the best we can do." right until your writers start saying "Hmm, funny, three of my friends with smaller followings make more than that on Substack; maybe you are not the person I need to have this conversation with."
I think that is salutary, both for writers and for the institutions (and societies) that depend on them. But it's also certainly not what you'd have predicted as an obvious second-order effect of improved multiparty payments infrastructure. That is part of what makes infrastructure such an endlessly fascinating topic.
P.S. If you would like to support continued deep dives into financial infrastructure, most far less navel-gazing than this one, please purchase a subscription to Bits about Money.
Want more essays in your inbox?
I write about the intersection of tech and finance, approximately weekly. It's free.