Ghost: The Open-Source CMS That Made Itself Legally Unsellable — $11M ARR as a Nonprofit, With Revenue and Churn Streamed Live
Ghost, the open-source CMS, chose a nonprofit structure whose legal constitution makes the company impossible to buy or sell. It publishes ARR of $11,055,673, 30,557 paying customers, and 2.92% net churn live on its website, reinvesting 100% of revenue into the product. A record of the "unsellable company" — 41 people, fully remote, founded 2013.
Companies are usually valued by what they could sell for. Ghost discarded that yardstick at the level of its legal constitution. The Ghost Foundation, which develops the open-source CMS for bloggers and publishers, was set up as a nonprofit, and its official site states plainly that its legal constitution ensures the company “can never be bought or sold.” There are no shareholders and no investors. One hundred percent of revenue is reinvested into the product and the community.
And yet the business is not small. Ghost’s About page embeds the company’s own live dashboard: as of August 2026 it displayed ARR of $11,055,673, monthly revenue of $921,306, 30,557 paying customers, and net churn of 2.92%. Not a summary of an annual report, not a founder’s blog recap — raw, continuously updating operating numbers, visible to anyone at any time. Ghost was founded in April 2013 by John O’Nolan and Hannah Wolfe. The team page lists 41 members with nationalities and join dates, and notes the company has been fully remote since 2013.
The numbers
| Item | Figure (August 2026, live on the official site) |
|---|---|
| ARR | $11,055,673 |
| Monthly revenue | $921,306 |
| Paying customers | 30,557 |
| Net churn | 2.92% |
| Monthly requests | 9 billion |
| Total installs | 100M+ |
| GitHub stars | 54,701 |
| Team | 41 people (fully remote since 2013) |
| Founded | April 2013 (John O’Nolan / Hannah Wolfe) |
“Unsellable” is structure, not sentiment
Choosing nonprofit status does not mean running a charity. The Ghost Foundation describes itself as “completely self-sufficient” and able to employ a full team. It pays 41 salaries and runs $11M in ARR. A real business. What differs is the exit design. In a normal corporation, a sale or IPO is always on the table as a way for founders and investors to cash out, and users cannot veto a post-acquisition change of direction. Ghost explains its choice as wanting to be “true to its users, rather than shareholders or investors.”
The design fits the nature of open-source business. A CMS is a high-switching-cost product; users care whether the software will still exist, with the same philosophy, ten years from now. If a sale is constitutionally impossible, that worry disappears structurally. Ghost’s real invention is backing the promise “we will never sell” with a legal document rather than good intentions. In Plausible’s path to $1M ARR (another open-source company built on an ideological stance, against Google Analytics) being “chosen for the philosophy” likewise proved a stronger differentiator than price or features.
Between 100 million installs and 30,000 customers
Ghost’s revenue model gives the software away and charges for the official hosting service, Ghost(Pro). The site describes it as a cycle: free open-source software attracts users, some of whom pay for Ghost(Pro), and that revenue funds development, “a sustainable open source model.”
The numbers show the shape of the funnel. Total installs exceed 100 million. Monthly requests run to 9 billion. Against that, paying customers number 30,557. Only a tiny fraction of the free user base pays, and it works because the denominator is enormous. Open source is a zero-distribution-cost machine for building that denominator, and the segment that doesn’t want the hassle of self-hosting falls naturally into the paid tier. Distribution of the product itself is the customer acquisition, the inverse of burning ad spend to buy users. The same structure appears at the individual scale in azu’s ¥2.29M/year in OSS sponsorships built on 14 years of consistency: what you give away for free becomes the foundation of what you earn.
Showing every number as a sales pitch
Publishing everything from ARR to churn live is a practice of transparency, and also very good marketing. When a publisher picks a CMS, the biggest anxiety is “will this company still exist in five years?” Ghost’s dashboard answers with numbers that update by the minute. With $921,306 in monthly revenue and 2.92% net churn in plain sight, no viability pitch is needed. In a completely different industry, Mirai Shokudo, the Tokyo diner that published its monthly books for four years, won trust and attention the same way: disclosure functions as an asset, not a cost.
Of course, this works because the numbers are healthy. Disclosure cuts both ways, if growth stalls, the stall is visible to the whole world. Continuing to publish is itself a discipline imposed on management.
Weaknesses and risks
The costs of this structure are equally clear. No capital acceleration, to begin with: unable to issue equity, Ghost can only invest out of its own revenue while competitors stack features and acquisitions with VC money. No stock-based compensation either: hiring depends on salary and sympathy with the mission. Transparency has a blind spot of its own, too: while the 2.92% net churn figure is public, its composition (cancellations versus upgrades) cannot be read from outside. This article can only evaluate what the dashboard shows. In newsletters and paid subscriptions, many competitors are venture-funded, whether conviction can keep beating capital is something the numbers will keep answering.
What generalizes, and what doesn’t
Three elements travel. For products where “we will never sell” affects the buying decision (infrastructure-like products with high switching costs, where longevity is a selection criterion) renouncing the exit is itself a differentiator. Open source plus official hosting still works as a funnel: free distribution builds the denominator, and the convenience-seeking segment converts. Live operating metrics, meanwhile, can erase the “will they survive?” objection in B2B, while doubling as management discipline.
The limits are just as clear. Going nonprofit is irreversible. The founders permanently forfeit any sale proceeds. It is not a design to recommend to everyone, and thirteen years to $11M ARR is not fast by funded-competitor standards. An unsellable company can only be built by founders who can decide, at the start and forever, that they will never sell.
Sources
This article summarizes and analyzes the public sources above. Please refer to the primary sources for details.
You may freely quote or republish this article in news media, blogs, or AI answers, provided you credit "Small Start (small-start.com)" and link to this page. No prior permission is needed. Reprint & quotation policy →
Similar cases

Balsamiq: 16-Year-Old Wireframing Tool Publishes Its Own $6.58M Revenue — Including an "Intentional Decline"
SaaS
MDZ.AI: Bundling $200/Month of AI Tool Subscriptions Into a $49 Plan — $165K/Month in Two Years
SaaS
Kaching Appz: No Office, No Employees, $100,000+ MRR at a 90% Margin on Shopify Apps
SaaS
Canny: $3.5M to $3.3M in Year 7, the First Decline the Feedback SaaS Published Anyway
SaaSMost read
- 1
Peak Monthly Sales of ¥1 Million on minne. A Former Designer Turned Handmade Artist Explains the Craft of "Photos That Sell"
33 recent visits - 2
Six AI videos, ¥153,030 in the first month — one video with 4.22 million views drove two-thirds of TikTok monetization revenue
27 recent visits - 3
From 30 yen in revenue to 8 years later: how running 3 apps in parallel got an indie developer to 200,000 yen a month
22 recent visits - 4
Side-Business Blog "Tsuzuki Blog": From ¥42,000 to ¥1 Million a Month in One Year. Breaking Down the Published Monthly Data
15 recent visits - 5
¥50,000–70,000 a Month in Royalties From 24 Kindle Books: 16 Months of Real Data on "Volume-Production" Self-Publishing, ¥750,000 Cumulative
13 recent visits
Latest articles
- 2026-09-19
Gadgemaga: 594,562 Pageviews, ¥3,784,734 in a Month — The Ledger a Former Civil Servant Published Down to the Last Yen
- 2026-09-18
NinjaPear: After LinkedIn Sued His $10M ARR Company, the Founder Rebuilt in Two Weeks and Reached $15,000 a Month in Six Months
- 2026-09-18
Instatus: $48K MRR with Every Metric Public, After Dropping "Sup" and Repricing Upmarket
- 2026-09-17
SEOTesting: Ditching the Name Sold "2.5 Years' Worth" in 9 Months — Renaming and Rebuilding the Free Tool SanityCheck to $18K MRR and 330 Customers
- 2026-09-17
Hitsuji: Quit With No Website, Savings Down to ¥20,000, Then ¥1.7M a Month in Year One