Substack takes 10% of every paid subscription, every month, for as long as the subscriber stays subscribed. The fee never goes down. It is a percentage of money the writer earned, not money Substack earned for the writer.
The cut has acquired a name in the last year. Writers leaving Substack call it the Substack tax, Slashdot picked up the phrase, and the complaint behind it is older than the phrase. The math is just more visible than it used to be. A publication doing $100,000 a year on Substack hands the company $10,000 to keep doing the same job a flat-rate tool can do for under $200 a month.
This is the math, line by line, with the honest counter-argument for staying.
What the 10% actually is
The fee is on revenue, not profit. Substack's own pricing page is plain about it: 10% of all paid subscription income. Free posts cost nothing. Paid posts cost a flat 10% of every dollar a subscriber pays you, for as long as that subscriber renews.
Two things to notice. First, the fee compounds with the subscription itself. Every renewal pays the tax again. The relationship between the writer and the reader produces revenue for years; Substack collects on it for years. Second, the 10% is on top of payment processing. Stripe is a separate line item, not absorbed by the platform fee.
Once Stripe is in the mix, the combined effective rate lands somewhere between 13% and 16% of gross paid revenue. That is the number to plan around.
The math at 1k, 5k, and 25k paid subscribers
Three worked examples. Substack's 10% only, no Stripe yet, to keep the picture clean.
- 1,000 paid subs at $5/mo. Gross $5,000 per month. Substack takes $500 per month. Yearly toll: $6,000.
- 5,000 paid subs at $7/mo. Gross $35,000 per month. Substack takes $3,500 per month. Yearly toll: $42,000.
- 25,000 paid subs at $10/mo. Gross $250,000 per month. Substack takes $25,000 per month. Yearly toll: $300,000.
The third row is the one that breaks publications. $300,000 a year is a senior engineer. It is two full-time editors. It is half a year of payroll for a small team. Substack's argument is that the in-app network (Notes, recommendations, the reader app) makes up for the cut. At a six-figure publication, that argument has to do a lot of work.
The Stripe surcharge is real and it stacks
Stripe charges 2.9% plus $0.30 per card transaction. The recurring product Substack runs on (Stripe Billing) adds another 0.5% on Billing Starter or 0.8% on Billing Scale, per Stripe's billing pricing. Substack passes both through.
For a $10 monthly subscription, the full chain is:
- Substack 10%: $1.00
- Stripe card fee: $0.59 (2.9% + $0.30)
- Stripe billing fee: $0.05 (0.5%)
- Total fees: $1.64. Net to the writer: $8.36.
So 16.4% of every paid dollar goes to fees on Substack. The 10% headline is the largest cut, but it is not the only one. A $5 monthly subscription nets the writer closer to $4.20 after the full stack.
What flat-rate looks like instead
Most non-Substack newsletter tools charge a flat monthly fee tied to list size. The transaction fee, if any, is just Stripe's standard cost. There is no layered platform percentage on top of the revenue the writer brought in. Nashra is one of these. Our Newsletter and Publisher plans scale by total subscribers, not by paid revenue.
Run the comparison at the same three list sizes:
| Scenario | Substack per month | Nashra plan | Nashra per month | Approx. yearly delta |
|---|---|---|---|---|
| 1,000 paid × $5 | $500 + Stripe | Newsletter (3K subs) | $23 | ~$5,700/yr saved |
| 5,000 paid × $7 | $3,500 + Stripe | Publisher (6K subs) | $69 | ~$41,000/yr saved |
| 25,000 paid × $10 | $25,000 + Stripe | Publisher (25K subs) | $189 | ~$298,000/yr saved |
Both columns still pay Stripe's share. The difference is whether a second party takes a percentage on top.
The honest case for staying on Substack
A fair piece needs to argue the other side.
Substack's network is real. Notes is a working in-app feed that, for some writers, drives a meaningful share of new subscribers. The Substack Reader app gives existing subscribers a calm reading surface that does not compete with the inbox. The recommendation engine has produced several public examples of writers crossing tens of thousands of paid subscribers faster than they would have on their own.
If a publication is genuinely growing on Substack's distribution, not just publishing there, the tax is buying something. The question is whether the growth from the network is worth more than the cut. For most publications above a few thousand paid subscribers, the answer slides toward no. The network helps most when a writer is small. As the audience compounds, more of the growth comes from the writer's own work, and the 10% is paying for distribution that has already happened.
When the tax stops being worth it
A rough rule. Look at the last 90 days of paid subscriber growth. Of new paid subs, how many came from Substack-internal sources (Notes, recommendations, the explore feed) versus the writer's own newsletter, social, and referrals? If the Substack-internal share is under roughly 30%, the network is no longer earning the tax. The other 70% is the writer's own work, and Substack is taxing it.
At that point the math is the only argument left, and the table above is the math.
The deeper objection to Substack-style models is not the percentage. It is the architecture. The subscribers live on Substack's list, in Substack's app, behind Substack's discovery rules. The relationship is mediated by a third party. When that party changes terms, raises fees, or shifts the algorithm, the writer absorbs the change. Notes' rise is a recent example of that. The Substack-tax conversation is another.
Closing
A subscriber converts roughly 10× better than a follower. The math in our piece on subscribers vs followers is line by line; the short version is that the gap exists because the relationship is direct. The list is the writer's. The send is the writer's. The pricing is the writer's. If a paid newsletter is built on a percentage of someone else's platform and someone else's app, the algorithm has been re-introduced into the channel that was supposed to be the antidote to it.
Nashra is the publishing OS for the flat-rate version of this. One newsletter tool, one subscriber list as the spine, one price that does not scale with the success the writer paid for in advance. The side-by-side detail is on our Substack comparison.