The credit union of publishing platforms

The credit union branch nearest my apartment has a sign above the tellers that reads, in small type, that the members are the owners. I moved my accounts there six years ago, mostly on a whim and partly because the big bank had spent that year inventing fees. The experience has been exactly what you'd predict. The teller knows my name. The mortgage rate was fine. There is no app redesign, no growth team, no envelope in the mail trying to sell me a credit card with a padded leather texture on it. There is also no single thing about the credit union that is better than the bank in any way a marketer could put on a billboard, and I have no intention of leaving it, which took me years to understand and is the entire point.

The scene is the argument. Publishing platforms face the same fork as banks: who owns the pipe, and who profits when it grows. Ghost is the credit union of this market, a nonprofit foundation running an open source publishing platform, funded by the subscriptions of the people who publish on it, and everything about the product, good and bad, falls out of that structure the way the fee schedule falls out of a bank's.

The zero fee is real, and the rent is where it lives

Ghost(Pro), the managed service, starts at $18 a month billed yearly for a solo blog or newsletter with up to 1,000 members, runs $29 for the Publisher plan that unlocks paid subscriptions and custom themes, and runs $199 for a Business plan with ten thousand members. There is no fee on your subscription revenue, ever: the platform takes zero percent, Stripe's processing cut aside, and email sends are unlimited on every plan. Self-hosting the software itself costs nothing but competence. Their own comparison table puts the honest sentence on the page: when you self-host and the servers catch fire, you're the one who loses sleep. That's the deal in one line, and I trust a pricing page more when it admits its cost in the second person.

Run the arithmetic I've been carrying across these reviews. A writer with a thousand paying subscribers at five dollars a month hands Substack five hundred dollars every month and Gumroad ten percent plus fifty cents per sale, while the Ghost bill stays at twenty-nine dollars whether she has ten subscribers or a thousand. At zero revenue, the comparison inverts and Substack is free while Ghost is rent. Ownership has a fixed cost and dependence has a percentage cost, and your income is the variable that decides which one is a bargain. That's not a preference. It's the shape of the two curves.

A nonprofit owns the pipe

Consider the incentive, because this is the one company in the series where the audit comes back boring. Ghost is built by the Ghost Foundation, the code is open source on GitHub, the money comes from hosting subscriptions rather than venture rounds, and the founder, John O'Nolan, has spent a decade writing public letters about the business model rather than growth updates about it. The product decisions fall out of the balance sheet. No ads. No engagement feed engineered to hold you. Exports supported from day one. Their feature table openly mocks platforms that stamp their logo over your publication to capture your audience, which is a real criticism of real competitors and also, obviously, marketing, and both things can be true.

The tension worth watching is that Ghost is building discovery anyway. The current platform ships an Explore directory, recommendations between publications, and distribution into the open social web through ActivityPub, the protocol behind Mastodon and Bluesky. Every pipe in this series has wanted to become a store, from Gumroad's marketplace to Substack's network, and Ghost's directory is a pipe making the same wish. The difference is who the wish serves. A venture-funded company monetizes discovery because a board demands it. A nonprofit monetizes discovery to keep the lights on, which caps the ambition at survival rather than expansion. (The case they make for open subscription standards is laid out here.) I've been waiting three years for the structure to betray the product. It hasn't yet, and the wait is the review.

The case for the network you don't own

Now the strongest counterargument, argued properly, because it's genuinely strong. Ownership without distribution is a filing cabinet. Ghost's discovery layer is a directory; Substack's is a feed that has turned strangers into paying subscribers by the millions, and Beehiiv's is a machine built by people who ran one of the best email businesses of the decade. A writer with no audience gets nothing from a clean incentive structure. She gets subscribers from network effects, and network effects live on platforms with money and appetite. The practical complaints back this up: Ghost's growth features are modest, the member cap means a $18 plan that succeeds upgrades to a $29 plan and then to $199, and the analytical depth only arrives on the higher tiers. There's also the honest count of what "yours" means. The website, the list, the archive, the checkout: all of it moves if you decide to move it. The readers who found you through a network will not, because they were never yours to move.

I hold both of these fully, which is why the verdict will sound like a coin that won't land, and I've decided that's the correct answer rather than a failure of nerve.

The variable is where your readers come from

It depends, and the dependency is the bottleneck. If attention is your problem, the networks are worth their percentages, and Ghost is a beautiful tool for a writer who has already been found. If the problem is keeping what you build, then the nonprofit owns the pipe and so, functionally, do you, and the rent is the cheapest insurance in publishing. Watch one variable over the next few years: whether the Explore directory starts selling placement, ranking, or promotion, the way every marketplace before it learned to. The credit union is a credit union right up until the day it opens a casino. After that, it's a bank with a friendly sign.