BREAKING: Sanity closes $85M Series C led by Bullhound Capital “The age of the CMS is over” - Sanity, May 2025 NIKE · PUMA · FIGMA · CLOUDFLARE ship content on Sanity 1M+ users · 6,000+ teams · dual HQ Oslo / San Francisco GROQ · Content Lake · Portable Text · content is data
Company Profile · Developer Tools / AI

Sanity.

The four Norwegians who looked at the content management system, decided it was the wrong abstraction, and built a backend for the AI era instead.

Founded 2016 Oslo · San Francisco ~230 employees Series C · $85M
Sanity brand card reading 'Structure powers intelligence' over a glowing orange sphere
SANITY, PHOTOGRAPHED AS IT WANTS TO BE SEEN: a single sentence and a sphere. The pitch fits on a business card; the database underneath does not.
FILED UNDER: CONTENT INFRASTRUCTURE EST. READ: 8 MIN SOURCES: 17 PUBLIC

It is a Tuesday, and somewhere a marketer at Nike changes a single line of copy. The edit does not land on a page. It lands in a database, as a typed block of JSON, and within seconds it is everywhere the brand lives - app, store kiosk, website, and, increasingly, the prompt feeding an AI agent. The marketer never saw the machinery. That invisibility is the whole product. The machinery is called Sanity.

This is what Sanity looks like in 2026: not a place where you build a website, but the place your content lives before it becomes anything. The company calls itself “the Content Operating System for the AI era,” which is a mouthful, and also a fairly precise description of a bet it made years before everyone else was forced to make it. Content, Sanity argues, is data. Treat it like a webpage and you have built a museum. Treat it like structured data and you have built a supply chain.

“Structure powers intelligence.”- Sanity's pitch, reduced to three words
The problem they saw

Every CMS was a haunted house

Before Sanity was a company, its founders ran a Norwegian product studio called Bengler. Clients kept arriving with the same request - build us something ambitious on the web - and the team kept hitting the same wall. Every content management system they tried tangled the content up with its presentation. The words, the images, the structure: all welded to a particular page, in a particular template, in a particular CMS that would be obsolete in three years.

The breaking point, as the founders tell it, arrived with a commission for OMA - the architecture firm of Rem Koolhaas. Building a serious editorial experience on top of the available tools, they found the existing systems “terrifyingly messy.” The origin story Sanity likes to repeat is that the product was created “in a fit of rage.” It is a good line because it is also true: the irritation was specific, technical, and earned.

“Sanity was created in a fit of rage at how messy existing content systems were.”- The founders, on the origin of the company

The insight underneath the rage was almost boring in its simplicity. If content were stored as clean, structured data - separated entirely from how it eventually looks - then it could be queried, reshaped, and sent anywhere. A website today. A mobile app tomorrow. A voice assistant, a billboard, a chatbot, a model. You would write the content once and never again be hostage to the surface it happened to be born on.

The founders' bet

They burned the boat

Conviction is cheap until it costs you something. In Sanity's case it cost the founders their existing business. Rather than build the software on the side while the consultancy paid the bills, they shut Bengler down, moved every employee into the new software company, and put the whole team behind the product. Magnus Kongsli Hillestad - a former finance and private-equity professional who had taught himself to code as a passion project - became CEO. Simen Svale Skogsrud, co-inventor of the real-time backend and the query language, took the CTO seat. Even Westvang runs product as CPO; Oyvind Rostad runs operations as COO.

The unglamorous truth

Sanity is two things wearing one name. Sanity Studio - the editor developers actually touch - is open source and free. The Content Lake - the real-time database humming behind it - is the part you pay for. Give away the cockpit; sell the engine. It is a deliberate arrangement, and it is the entire business model in a sentence.

It is a tidy irony that a company so allergic to lock-in built its commercial moat out of the one piece nobody wants to rebuild themselves: a fast, real-time, referentially-honest backend that never goes down when an editor changes a field. Developers will happily fork an editor. They will not, in their right minds, run their own content database.

The product

Four ideas, no pages

What Sanity actually shipped is less a CMS than a set of primitives for treating content as code. There are four of them, and together they explain why developers talk about the thing the way other people talk about a good knife.

Content Lake

A real-time cloud backend that stores any valid JSON. Schemas live in your code, not as database constraints, so content models evolve without migrations or downtime - and broken references never break your app.

GROQ

Graph-Relational Object Queries: an open query language for JSON that joins, filters and reshapes content at the API level, handing the developer exactly the shape they asked for and nothing else.

Portable Text

An open spec for rich text stored as typed JSON blocks instead of HTML strings - so the same paragraph can render on a website, a native app, or inside an AI agent's context window.

Sanity Studio

An open-source editor built in React and TypeScript. Schemas are defined as code; every field and workflow is configurable, extendable, or replaceable. It is a CMS you can rewrite.

“Content is data.”- The two words the other three products are built on

The payoff of all this structure is the part Sanity now leans on hardest. A language model cannot reason about a blob of HTML. It can reason beautifully about clean, typed, structured content. Sanity spent years insisting content should be data for reasons of portability and engineering hygiene - and then the AI wave arrived and made that same structure the difference between content a machine can use and content it cannot. Sometimes the market comes to you.

The receipts

A short history of a long bet

2016
Founded in NorwayBengler's founders shut the agency, regroup as a software company, and start building Sanity.
2017-18
“Content is data”The Content Lake, GROQ and Portable Text take shape as the core of a structured-content platform.
2020-22
Developer adoption climbsSanity Studio's open-source, code-first approach wins over engineering teams; global brands sign on.
2025 · Jan
The AI-first roadmapCanvas AI ghostwriter, Agent Actions and an MCP server reposition Sanity as a backend for agents.
2025 · May
$85M Series CBullhound Capital leads, with ICONIQ Growth, Shopify, Threshold, Heavybit and others. “The age of the CMS is over.”
The proof

Who actually runs on this

Skeptics are right to ask whether any of this survives contact with real users. It does. Sanity reports more than a million users and over six thousand teams. The logo wall is the kind that makes a sales call easy: Nike, Puma, Figma, Cloudflare, National Geographic, Braze, and - tellingly, for a company betting on AI - Anthropic. Lady Gaga's team ships through it. So do thousands of teams nobody has heard of, which is arguably the better signal.

1M+
USERS
6k+
TEAMS
$85M
SERIES C
~230
EMPLOYEES

What content teams report after switching

SELF-REPORTED OUTCOMES SANITY PUTS ON ITS OWN HOMEPAGE · DIRECTIONAL, NOT AUDITED
Release cycles
300% faster
Custom APIs
0
Updates owned by content team
90%
The numbers a vendor chooses to publish are a personality test. Sanity brags about speed, zero glue code, and editors who no longer file tickets with engineering. Read them as ambition with evidence, not as an audit.

The money tells a parallel story. The May 2025 Series C - $85 million, led by Bullhound Capital, with ICONIQ Growth, Shopify, Threshold, Heavybit, Blue Cloud Ventures and Monochrome Ventures alongside - pushed total funding to roughly $173 million. Sanity says it is now fully capitalized and does not plan to raise again. Companies say that until they don't, but the posture is the point: this is a team that would rather build than fundraise.

“The end of the CMS era, and our $85M Series C.”- Sanity, announcing the round and the worldview in one breath
The mission

A backend for things that aren't pages

Strip away the category language and Sanity's mission is narrow and stubborn: be the structured-content backend for everything that isn't a page. Web, mobile, and now the agentic applications that read content, act on it, and run workflows without a human clicking anything. The competitors - Contentful, Contentstack, Storyblok, Strapi, Payload, headless WordPress - are all racing toward some version of this. Sanity's advantage is that it never had to turn the boat around. It was pointed here the whole time.

There is a healthy skepticism worth keeping. “Operating system” is a heavy phrase to hang on a content platform, and Sanity's enterprise governance - the audit trails, the fine-grained roles, the approval workflows that big companies demand - is younger than some rivals'. Structure is a discipline before it is a feature, and not every team wants to model their content as code. Sanity asks more of its users up front. The bet is that the front-loaded effort pays compounding interest later.

Why it matters tomorrow

The edit nobody saw

Return to that Tuesday. The marketer at Nike changes one line, and it propagates - to the app, the store, the site, and the model that will answer a customer's question an hour from now using exactly that line. None of those surfaces was built twice. None of them will be rebuilt when the next surface arrives, because the content was never tied to a surface in the first place. That is the quiet thing Sanity changed: it moved the work upstream, into structure, where it only has to be done once.

The CMS asked you to manage pages. Sanity asked a different question - what if you managed meaning, and let the pages assemble themselves? Four Norwegians got annoyed enough to build the answer, burned their old business to fund it, and waited for the rest of the industry to need it. In the AI era, the industry showed up. The edit still nobody saw. The machinery still invisible. That was always the plan.

Share this profile