Among a wide variety of products, Obsidian can be considered an outlier. Its CEO is not the founder but an early superuser who, after gradually becoming an ecosystem contributor, was invited by the founders to join as CEO.
Secondly, at a time when collaboration suite features were the path for note-taking apps to expand to larger teams and enterprises, Obsidian hardly relied on the collaboration models of Notion or Google Docs. It doesn’t force you to store data in its cloud, instead enabling collaboration through “sync services” or “shared folders.”
After the AI boom, as nearly all products were migrating to AI, either by rebuilding or adding AI features, Obsidian’s approach to incorporating AI has been remarkably restrained.
The team is fully remote, meets offline just once a year, and has almost no other scheduled meetings, allowing everyone to focus deeply on the highest priority matters.
Yet, it has been profitable from day one and has won over a large base of loyal users. Although they don’t have detailed user statistics, CEO Steph Ango recently disclosed that he estimates Obsidian has over 4 million users, with more than 10,000 organizations using it.
Obsidian’s ARR is around $25 million (a third-party estimate, as the company has not disclosed official figures). Despite having raised no funding, its market valuation is reportedly around $350 million.
The team currently has 9 people, including a cat named Sandy and 3 engineers. They are now hiring a fourth engineer, the 10th person on the team. They add only one person each year, and the full-time team size will be kept at around 10-12 people in the future, with no plans for further growth.
Therefore, Obsidian did not achieve its current success by “turning notes into a collaboration suite” or by “integrating AI everywhere.” Instead, it succeeded by adhering to a set of uncompromising product principles, transforming a niche but high-value user base into a highly sticky, self-propagating, self-expanding, and paying ecosystem.
Co-founder Erica Xu defines Obsidian as an “IDE for thought”—not a tool that thinks for you, but a thinking environment that you can completely control and continuously shape.
The goal from the start was never to “build a large, all-in-one collaboration suite,” but rather to create the “holy grail note-taking app” that meets individual needs. These principles later fostered an almost cult-like product loyalty among geeks, researchers, writers, and knowledge workers.
The two founders, Shida Li and Erica Xu, initially created the outlining tool Dynalist. This product helped them understand the value of “structured information” and “heavy-user workflows,” but it also revealed the limitations of outliners: users wanted more than just hierarchies—they wanted links, backlinks, graph views, and ownership of their local files.
This led to the transition from Dynalist to Obsidian: a shift from “organizing entries” to “connecting thoughts.” Dynalist addressed “how to list things clearly,” whereas Obsidian addresses “how to ensure knowledge remains readable, connectable, and evolvable over time.”
The manifesto on their official website lists Obsidian’s core values in five words: Yours, Durable, Private, Malleable, and Independent. This means your data belongs to you; the file format must be sustainable; privacy is the default; the tool should adapt to the user; and the company must be independent and user-funded, not investor-driven.
Yours: We believe everyone should have the tools to think clearly and organize their ideas effectively. That’s why our tool is free for everyone.
Durable: We believe your data should be future-proof and easily accessible, no matter where you are. That’s why we use simple, open file formats to prevent lock-in and ensure your data can be preserved for generations.
Private: We believe your thoughts and ideas are yours and deserve complete privacy. That’s why your data is stored on your device, where we can’t access it. When using our online services, your data is protected with end-to-end encryption for maximum security.
Malleable: We believe tools should adapt to your way of thinking, not the other way around. That’s why we designed our tools to be highly customizable and extensible, allowing you to tailor them to your unique needs.
Independent: We believe in upholding these principles. That’s why we are 100% supported by our users, not investors.
On his personal website, CEO Steph particularly emphasizes the philosophy of “File over app”: apps become obsolete, while files are worth preserving for the long term. Consequently, Obsidian saves notes as Markdown files on the user’s local device, rather than in a proprietary cloud database.
He says “File over app” is a philosophy: if you want to create durable digital artifacts, they must be files you can control, in formats that are easy to retrieve and read. Use tools that grant you this freedom. Over time, the files you create become more important than the tools that created them. Applications are ephemeral, but your files have a chance to endure. If you want your writing to be readable on a computer in the 2060s or 2160s, it’s important to ensure your notes are readable on a computer from the 1960s.
This positioning led it to intentionally forgo some features friendly to short-term growth (like web-first collaboration, centralized data, and strong lock-in), but in return, it gained a high degree of trust, freedom of migration, and a long-lifecycle mindset.
More importantly, Obsidian didn’t become a cold, sterile text tool just because it’s “file-first.” It layered capabilities like bidirectional links, graph view, Canvas, Properties, Web Clipper, Bases, and a CLI on top, but always ensures these features resolve back to the user-controlled file layer. Bases, introduced in 2025, brought a “database view” into Obsidian, yet it’s still built upon local Markdown + properties. Without betraying its file-based philosophy, it added Notion-style structured views to its value proposition.
Community-Driven Growth Link to heading
When most SaaS companies claim to have a community, they often just mean they have a Discord server. Obsidian is different; its community acts as an outsourced layer for product capabilities. In an interview, CEO Steph said the plugin architecture is “genius,” not because it’s technically flashy, but because it allows the core application to remain simple while letting the community address long-tail needs. Every user can have a different version of Obsidian, meaning the company doesn’t have to build “ten thousand features” itself, yet can still cover “ten thousand workflows.”
This brings a second-layer benefit to its growth: the spontaneous formation of a content ecosystem. It’s not the typical PLG where the “product goes viral on its own,” but rather a form of craft-driven PLG: users produce tutorials, plugins, templates, and themes out of a desire to “build their own systems better,” which in turn attracts new users.
Its business model is also remarkably restrained. The official licensing page states: Obsidian is free for personal, commercial, non-profit, educational, and governmental use. Paid options consist of optional support and services, rather than locking core features behind a paywall. Revenue primarily comes from four sources: Sync, Publish, Catalyst, and optional commercial support licenses.
According to Steph, although less than 1% of users use the Sync feature, this is sufficient for them to run the company and pay employee salaries (an estimated 80% of their revenue comes from this). Sync Standard is $4/user/month (billed annually), and Plus is $8/user/month (billed annually), emphasizing end-to-end encryption, version history, offline sync, cross-platform support, and shared vaults.
Its core users are not the general public, but a group of people highly sensitive to data ownership, customizability, and long-term readability—researchers, writers, developers, content creators, knowledge workers, and many heavy individual users within large corporations.
In the realm of AI, when all other products are making AI their core narrative, Obsidian’s approach is very restrained. It doesn’t position AI as a substitute for thinking, but rather as an auxiliary layer, an ecosystem layer, and a source of leverage for the team.
In a recent conversation, Steph mentioned that if AI were to be integrated into Obsidian, it must comply with the company’s Manifesto, especially its privacy principles. They are unwilling to send user data to OpenAI servers without explicit user consent.
Therefore, Obsidian’s earliest AI capabilities were not developed officially but were pioneered by the community. Examples include plugins like Text Generator. Later, the official roadmap introduced more cautious AI/automation plans, such as the Web Clipper Interpreter (for extracting web content using natural language), Keychain (for securely storing API keys for plugins), and a CLI (to facilitate scripting, automation, and integration with external tools). However, these efforts are fundamentally about laying the infrastructure for AI connectivity at the product’s periphery.
I believe Obsidian has truly turned its “values” into product constraints. Every core trade-off—be it local-first, Markdown, end-to-end encryption, the plugin ecosystem, or not taking investment—is interconnected. This creates a highly consistent image in the user’s mind and reduces a great deal of repetitive debate within the team.
It sells “convenience under principles,” not a “betrayal of principles.” Sync and Publish are viable because they add convenience without violating the fundamental premises that “your files are yours, data is local by default, and it’s all portable.” The product’s expansion is extremely restrained, always revolving around enhancing thinking and organizational capabilities, rather than sliding towards becoming an “all-in-one collaboration workspace.”
They also haven’t pursued making it “understandable for everyone.” The product has never been the most beginner-friendly tool, but this has paradoxically fostered a strong sense of ownership among its core users. For instance, Markdown, local files, and third-party plugins all raise the entry barrier, but this “love that requires effort” has forged a deeper loyalty. The AI boom hasn’t made it lose its original heart either.
In her personal blog, CEO Steph says that Obsidian doesn’t sell notes; it sells control. This is what knowledge workers truly want—not just to improve efficiency, but to ensure that their thoughts, notes, structures, and memories are not ultimately locked into a system they cannot control. So when they posted an engineer recruitment ad on X yesterday, it instantly ignited discussion. So far, this recruitment post has reached 8 million views.
For products with a philosophy similar to Obsidian, I think Ghost, which I introduced before in “1.1 Billion USD Revenue, 525 Million USD Profit, The Most Profitable Creator Platform,” is one of them. It was also built based on the founder’s own needs, and its ARR has now exceeded 10 million USD.
End!
Memo: Signal, not noise!
Scan the QR code or click “Read the original text” to continue reading