TL;DR. The paid community is now a standard recommendation for anyone with an audience, and it is recommended almost entirely without numbers. There is a reason for that: no pure-play community platform files audited financials anywhere in the world, so there is nothing to cite. We worked around the gap by building our own dataset, compiling 600 communities from a public platform directory along with their listed price and member count. The median paid community charges 28 dollars a month and has 250 members. The distribution is severely top-heavy: the top 10 percent of paid communities account for 61.2 percent of all implied revenue in the sample. We then took every major platform’s published fee schedule and priced it against that distribution. Percentage-only platforms such as Patreon and Substack cost a flat 10 percent at every scale. Fixed-fee platforms cost far more below about 1,000 dollars a month of revenue and far less above it. And the premium tiers that platforms actively upsell only break even above 10,000 to 11,000 dollars a month, a threshold that 58.1 percent of the communities in our sample fail to clear even when you credit them with the most generous revenue figure their listed price and membership could possibly produce.
This is a companion to our comparison of recurring revenue models for solo operators, which covers when a membership is the right structure at all. This piece assumes you have decided it might be, and asks what it actually costs and what it plausibly earns.
Why there are no numbers, and what that tells you
Start with the disclosure problem, because it shapes everything downstream.
We searched the SEC’s EDGAR system for company records for Substack, Patreon, Kajabi, Skool, Teachable and Circle. None of them has one. They are all private, and none publishes audited financials. The public comparators that once existed have been leaving: Squarespace filed to delist in October 2024, Udemy was acquired and filed its delisting notice in May 2026, and Eventbrite was acquired in March 2026.
So when an article tells you the average paid community earns some specific amount, ask where that came from. In our experience it came from a vendor.
The vocabulary problem is just as instructive. “Take rate” is the metric everyone quotes for platform businesses. We checked the actual filings. In Etsy’s most recent quarterly report the term is defined explicitly and disclosed: revenue take rate of 25.9 percent, with the definition given as revenue divided by gross merchandise sales. In Shopify’s quarterly report for the same period, the phrase “take rate” does not appear at all. Shopify discloses revenue of 3,583 million dollars and gross merchandise volume of 115,567 million dollars, and anyone who divides those to get 3.10 percent is performing a computation, not quoting a disclosure.
That distinction matters more than it sounds. Most “platform X takes Y percent” claims in circulation are third-party arithmetic being presented as company reporting, and the arithmetic is often wrong. Thinkific, one of the few listed companies in adjacent territory, states a take rate of 4.4 percent in its investor presentation for the twelve months to 30 June 2026. The same slide reports commerce revenue of 13.8 million dollars and payment volume of 291 million dollars, which is 4.74 percent, and describes that payment volume as 64 percent of gross merchandise volume, which puts the rate against the wider base at roughly 3.0 percent. Neither denominator on the slide produces the 4.4 percent the slide states. We are not accusing anyone of anything; we are pointing out that a number in a deck is not a number in a filing.
Building the dataset
Because there is nothing to cite, we compiled our own.
On 3 September 2026 we collected the public discovery directory of Skool, one of the largest paid-community platforms, capturing 600 distinct communities with the price, billing interval and member count each one displays publicly. We then re-checked a random sample against the communities’ own live pages to confirm the capture was accurate. Member counts matched exactly, and the sampled price matched exactly.
Two limitations have to travel with every number below, and we will repeat them where they bite.
The directory is ranked and curated rather than randomly sampled, so it is biased toward communities that are doing well. Every level reported here should be read as optimistic relative to the true population.
And member count is not subscriber count. A community’s displayed membership includes free members, legacy members and people who joined and stopped paying. Multiplying listed price by total members therefore produces a hard ceiling on revenue, not an estimate of it. We call that figure implied revenue throughout, and it is always too high, sometimes by a very large factor.
Table 1. CEOtudent original dataset: what 600 paid communities actually look like. Skool public discovery directory, captured 3 September 2026, n = 600 communities.
| Measure | Value |
|---|---|
| Communities captured | 600 |
| Paid, billed monthly | 473 (78.8 percent) |
| Free | 98 (16.3 percent) |
| Annual or one-time pricing | 29 (4.8 percent) |
| Median monthly price, paid communities | 28 dollars |
| Mean monthly price | 43.93 dollars |
| Price range | 1 dollar to 1,000 dollars |
| 25th to 75th percentile price | 12 dollars to 49 dollars |
| 90th percentile price | 97 dollars |
| Median members, all communities | 257 |
| Median members, paid communities | 250 |
| Median members, free communities | 339 |
| 90th percentile members | 1,822 |
| Largest community in sample | 164,688 members |
The price distribution is tighter than the discourse suggests. Two-thirds of paid communities charge under 50 dollars a month, and the modal band is 25 to 49 dollars.
Table 2. Price distribution, paid monthly communities (n = 473). Derived from the dataset above.
| Monthly price | Communities | Share |
|---|---|---|
| Under 10 dollars | 102 | 21.6 percent |
| 10 to 24 dollars | 106 | 22.4 percent |
| 25 to 49 dollars | 163 | 34.5 percent |
| 50 to 99 dollars | 74 | 15.6 percent |
| 100 to 199 dollars | 20 | 4.2 percent |
| 200 dollars and above | 8 | 1.7 percent |
Note what is not here. The 500-dollar-a-month mastermind that dominates the marketing of this category is 1.7 percent of the market, and even that band tops out at eight communities in six hundred.
The concentration problem
Median figures conceal the thing that most determines whether this works for you.
Multiplying each paid community’s listed price by its member count and ranking the results, the top 10 percent of paid communities account for 61.2 percent of all implied revenue in the sample. The median implied revenue is 7,144 dollars a month; the mean is 24,711 dollars, three and a half times higher. When a mean runs that far ahead of a median, you are looking at a market where a small number of operators earn most of the money.
That shape is not unique to communities, and we have found it in every creator-economy dataset we have examined, including in our analysis of how long it takes to replace a salary with independent income. But it has a specific consequence here, which we will come to when we price the platforms.
One finding did surprise us. The correlation between price and member count, measured on logarithms, is only weakly negative at about -0.17. The intuition that charging more necessarily means a much smaller community is not strongly supported in this sample. Price and size are close to independent across most of the range.
What the platforms actually charge
Now the part that is fully verifiable, because it is published. We collected the current fee schedules on 3 September 2026.
Table 3. Published platform pricing, September 2026. Platform fee and platform take rate as listed. Card processing is excluded from the comparison because every operator pays it regardless of platform; for reference, the standard rate is 2.9 percent plus 30 cents per domestic transaction.
| Platform and tier | Monthly fee | Platform take rate |
|---|---|---|
| Skool Hobby | 9 dollars | 10 percent |
| Skool Pro | 99 dollars | 2.9 percent |
| Circle Professional | 89 dollars | 2 percent |
| Circle Business | 199 dollars | 1 percent |
| Mighty Networks Launch | 79 dollars | 2 percent |
| Mighty Networks Scale | 179 dollars | 1 percent |
| Patreon | None | 10 percent |
| Substack | None | 10 percent |
Kajabi sits outside this comparison because its model is different: a high fixed fee from 179 dollars a month, with no platform take rate on top and card processing at 2.9 percent falling to 2.7 percent on higher tiers. Whop publishes only processing-style fees and no separate platform take at all.
A word on what we could not confirm. Patreon’s page states its 10 percent rate and then adds that payment processing, currency conversion and payout fees apply on top, without publishing those components; its help centre refused our requests. Substack states that writers keep 90 percent of revenue minus credit card fees, and does not publish the card rate it passes through. Skool’s page does not say whether processing sits inside or outside its stated transaction fee. We are not going to guess at any of these, so the table above compares only what each company actually publishes.
The crossover nobody tells you about
Put the two halves together and the interesting result appears. Here is what each platform costs, per month, at five revenue levels.
Table 4. CEOtudent derived analysis: total platform cost as a percentage of monthly revenue. Computed from the published fee schedules in Table 3. Card processing excluded throughout.
| Platform and tier | At 500 dollars | At 1,000 dollars | At 2,500 dollars | At 5,000 dollars | At 10,000 dollars |
|---|---|---|---|---|---|
| Patreon or Substack | 10.0 percent | 10.0 percent | 10.0 percent | 10.0 percent | 10.0 percent |
| Skool Hobby | 11.8 percent | 10.9 percent | 10.4 percent | 10.2 percent | 10.1 percent |
| Skool Pro | 22.7 percent | 12.8 percent | 6.9 percent | 4.9 percent | 3.9 percent |
| Circle Professional | 19.8 percent | 10.9 percent | 5.6 percent | 3.8 percent | 2.9 percent |
| Circle Business | 40.8 percent | 20.9 percent | 9.0 percent | 5.0 percent | 3.0 percent |
| Mighty Networks Launch | 17.8 percent | 9.9 percent | 5.2 percent | 3.6 percent | 2.8 percent |
| Mighty Networks Scale | 36.8 percent | 18.9 percent | 8.2 percent | 4.6 percent | 2.8 percent |
Read down the first column. At 500 dollars a month of revenue, Circle’s Business tier consumes 40.8 percent of everything you collect, and Mighty Networks’ Scale tier 36.8 percent. At that level the percentage-only platforms are less than a quarter of the cost. Read across the bottom rows and the ranking inverts completely: at 10,000 dollars a month, the fixed-fee tiers cost under 3 percent while Patreon and Substack still take a full 10 percent, which is more than three times as much in absolute dollars.
The crossovers are precise, and they are worth memorising.
Table 5. CEOtudent derived analysis: the monthly revenue at which each upgrade starts paying for itself. Solved from the published fee schedules.
| Decision | Crossover revenue |
|---|---|
| Skool Hobby to Skool Pro | 1,268 dollars |
| Patreon or Substack to Circle Professional | 1,112 dollars |
| Patreon or Substack to Skool Pro | 1,394 dollars |
| Mighty Networks Launch to Scale | 10,000 dollars |
| Circle Professional to Circle Business | 11,000 dollars |
Now join this to the population data, which is the step nobody takes because the population data did not exist until we built it.
The first-tier upgrade decision is easy for most operators: at implied revenue, only 15.6 percent of the paid communities in our sample fall below the 1,268-dollar Skool crossover, so the large majority are past the point where a fixed-fee plan beats a percentage. Remember that implied revenue is a ceiling, so the true share below the line is higher, probably considerably. Still, the direction is clear.
The second-tier upgrade is a different story. The Circle Business and Mighty Networks Scale tiers do not pay for themselves until 11,000 and 10,000 dollars a month respectively. In our sample, 58.1 percent of paid communities do not reach 10,000 dollars a month of implied revenue, and 60.3 percent do not reach 11,000. Those are ceilings. The real proportion of communities for which the premium tier is a net loss is higher than three in five, and plausibly much higher.
That is the finding. The upper tiers that these platforms market most aggressively are, for the majority of the communities actually listed on them, more expensive than the tier below. Not marginally: at 2,500 dollars a month, Circle Business costs 9.0 percent against Circle Professional’s 5.6 percent, which is a 61 percent increase in platform cost for a downgrade in economics.
The benchmark that does not exist
We tried to find a churn benchmark for paid communities. The result is worth reporting as a finding in its own right.
The most-cited membership-community churn figure we could locate, quoted as 5.8 percent monthly, comes from a page whose own sourcing statement describes it as an editorial estimate anchored to published industry ranges. No sample. No method. No underlying data. The publisher sells churn-analysis software. That number is not evidence of anything, and it is currently circulating as though it were.
The best-documented subscription benchmark set we found comes from a payments provider publishing aggregated network data, and even there we hit a problem we cannot resolve. The page labels its figures as median annual churn rates and reports an overall figure of 3.60 percent, with an education vertical at 4.99 percent. But its own explanatory text elsewhere discusses converting a 2 percent monthly rate into an annual one, and if 3.60 percent were genuinely annual, the average subscriber would stay for more than 27 years. The figures are internally inconsistent with the label. We are not going to build lifetime-value arithmetic on a number whose time unit is ambiguous, and neither should anyone else. Getting that wrong is not a rounding error; it is a factor of roughly twelve.
What we can report, because it is audited and stated in filings: Coursera’s Net Retention Rate for Enterprise Customers was 91 percent for the three months ended 30 June 2026, compared with 95 percent for the same quarter a year earlier, and the company attributes the decline primarily to customer attrition. Udemy’s business segment net dollar retention was 93 percent as of 31 March 2026, in its final quarter as a standalone company, with the filing noting that churn outpaced expansion; retention among its largest customers was higher at 97 percent. Those are large, professionally-managed subscription businesses with dedicated retention teams.
If that is the baseline for well-resourced operators, a solo community operator planning around better numbers is planning around hope. The pattern-recognition point here is the same one we made about passive income maintenance costs: the recurring-revenue model that looks self-sustaining on a spreadsheet has an attrition rate that has to be actively fought every month.
The reality check on the population
One more anchor, from a source with no interest in selling you a platform.
The United States Census Bureau’s Nonemployer Statistics count every business with no paid employees, which is the category almost every community operator falls into. For reference year 2023 there were 30,427,808 such establishments generating 1.753 trillion dollars in receipts. We downloaded the underlying data file rather than relying on a summary, and the revenue-size bands sum exactly to the published total with no residual.
Table 6. What solo businesses actually earn. United States Census Bureau Nonemployer Statistics, reference year 2023, all sectors. Shares derived; bands verified to sum to the published total.
| Annual revenue band | Establishments | Share of all |
|---|---|---|
| Under 25,000 dollars | 18,439,574 | 60.6 percent |
| Under 50,000 dollars | 22,971,578 | 75.5 percent |
| 250,000 dollars and above | 1,420,567 | 4.7 percent |
| Average receipts per establishment | 57,611 dollars | – |
Three in five American solo businesses gross under 25,000 dollars a year. Three in four gross under 50,000. That is the actual population that community-as-a-business-model advice is addressed to, and it is a long way from the framing most of that advice uses.
For scale on the addressable side: the Bureau of Labor Statistics counted 9,844,000 unincorporated and 6,645,000 incorporated self-employed workers in July 2026, and Eurostat recorded 19,430,400 solo self-employed people across the EU-27 for 2025, which is 67.9 percent of all EU self-employment. The population of potential operators is enormous. The population earning meaningful money from it is not.
What to actually do with this
Pick the platform for the revenue you have, not the one you want. This is the single most expensive mistake in the data. Below roughly 1,100 dollars a month, a percentage-only platform is cheaper, full stop, and it costs nothing to start. Above roughly 1,400 dollars, move to a fixed-fee plan. Do not buy the premium tier until you are clearing 10,000 dollars a month, which the majority of listed communities never do.
Price at 25 to 49 dollars unless you have a specific reason not to. That is where a third of the market sits, and the weak price-to-size correlation means you are not obviously buying a smaller community by charging at the upper end of that band.
Do not model on member counts. Every public number you see, including ours, counts members rather than payers. Your own dashboard is the only place the real figure exists. If you are estimating what someone else is earning, you are estimating a ceiling.
Budget for churn you cannot benchmark. There is no credible community-specific churn figure available. Audited retention at far larger subscription businesses is in the low nineties annually and falling. Plan for replacement as a permanent monthly cost, not a launch-phase problem.
Check whether you need the community at all. A community is a recurring-delivery obligation, which is the most demanding revenue structure a solo operator can take on. If your goal is recurring revenue rather than community specifically, our comparison of recurring revenue models sets out the alternatives, and micro-SaaS unit economics covers the numbers for the least labour-intensive option.
The CEO and the student in this
The student instinct here is to look for the benchmark: what should I charge, what churn is normal, what does a good community earn. That instinct sent us looking, and the honest report is that for paid communities specifically, most of those benchmarks do not exist. What circulates instead is vendor marketing and editorial estimates wearing the costume of data.
The CEO instinct is different and more useful. A CEO facing a market with no published comparables does not wait for one. They price the inputs they can verify, they build the distribution themselves when nobody has, and they make the decision on the arithmetic that survives scrutiny. The platform fee schedules are public. The crossover points are solvable in one line of algebra. The size distribution can be compiled by anyone willing to spend an afternoon on it.
That is the real transferable skill in this piece, and it is worth more than any of the individual figures. In a market where the numbers are missing because everyone selling into it prefers them missing, the operator who does the arithmetic anyway has an advantage that does not depend on anyone publishing anything.
Learn like a student: check whether the benchmark you are about to plan around has a method behind it, and accept it when the answer is that nobody knows. Lead like a CEO: build the number yourself, price the decision on what you can verify, and stop paying for a tier that the data says you have not earned yet.
FAQ
What is a realistic monthly price for a paid community in 2026?
In our sample of 473 paid communities the median was 28 dollars a month, and 78.5 percent charged under 50 dollars. The largest single band was 25 to 49 dollars at 34.5 percent. Communities charging 200 dollars or more were 1.7 percent of the sample. Prices above 100 dollars exist but are rare enough that they should be treated as a specific strategy rather than a default.
How many members does a typical paid community have?
The median paid community in our sample displayed 250 members, and the median across all 600 including free communities was 257. The 90th percentile was 1,822. Note that these are displayed member counts, which include free and lapsed members, so the number of paying subscribers is lower and often much lower.
Which platform is cheapest?
It depends entirely on your revenue, and the answer flips. Below about 1,100 dollars a month, Patreon or Substack at a flat 10 percent with no fixed fee is cheapest. Above about 1,400 dollars, a fixed-fee plan such as Skool Pro or Circle Professional wins and the gap widens quickly. The premium tiers do not pay for themselves until 10,000 to 11,000 dollars a month.
Can I trust the revenue figures other people publish about their communities?
Treat them as unaudited self-reports. No pure-play community platform files audited financials, and the companies in this category that were publicly listed have mostly delisted since late 2024. Where a figure comes from a platform’s own marketing, it is a claim rather than evidence, and where it comes from a third party dividing one public number by another, it is a computation rather than a disclosure.
What churn rate should I plan for?
There is no methodologically-documented churn benchmark specific to paid communities that we were able to verify, and the most widely-circulated figure describes itself as an editorial estimate. The nearest defensible anchors are audited and disclosed in filings: a Net Retention Rate for Enterprise Customers of 91 percent at Coursera for the quarter ended 30 June 2026, down from 95 percent a year earlier, and business segment net dollar retention of 93 percent at Udemy as of 31 March 2026. Plan conservatively and measure your own.
Is a paid community still a good business model?
For a minority of operators it is clearly working, and the concentration in our data shows how small that minority is: the top 10 percent of paid communities hold 61.2 percent of implied revenue. It is a legitimate model with a punishing distribution, which is different from being a bad one. The decision should turn on whether you can sustain a recurring delivery obligation indefinitely, not on the median figures, because the median is not where the money is.
Sources
United States Census Bureau, Nonemployer Statistics, reference year 2023, published data file NS2300NONEMP.
United States Bureau of Labor Statistics, Employment Situation, Table A-9, July 2026.
Eurostat, Labour Force Survey, dataset on self-employment by professional status, reference year 2025.
Etsy Inc., quarterly report on Form 10-Q for the period ended 30 June 2026, filed with the United States Securities and Exchange Commission.
Shopify Inc., quarterly report on Form 10-Q for the period ended 30 June 2026, filed with the United States Securities and Exchange Commission.
Coursera Inc., quarterly report on Form 10-Q for the period ended 30 June 2026, filed with the United States Securities and Exchange Commission.
Udemy Inc., quarterly report on Form 10-Q for the period ended 31 March 2026, filed with the United States Securities and Exchange Commission.
Thinkific Labs Inc., second quarter 2026 results and investor presentation, period ended 30 June 2026.
Published platform pricing pages for Skool, Circle, Mighty Networks, Patreon, Substack, Kajabi and Whop, together with the standard published card-processing schedule, all collected 3 September 2026.
CEOtudent original dataset: 600 communities compiled from the Skool public discovery directory, 3 September 2026.
This content was compiled with the support of AI following in-depth research, then written and prepared for publication by the CEOtudent editorial team.
This post is also available in:
