The diversity of customer needs and the complexity of data that digital content is expected to handle keep growing. Yet many teams are stuck on legacy technology that makes it hard to update content, let alone scale the experience to new channels.

A headless CMS stores content as structured data and delivers it through APIs to any frontend, or "head": a website, an app, a portal or an AI agent. Because the content isn't tied to one presentation layer, teams can reuse it across channels, markets and integrations.

In this article, we walk through 8 headless CMS use cases in the order teams ask us about them. Each comes with a real team, a public result and what the setup means for AI. Every team here builds on Hygraph, our agentic content platform.

#When to choose a headless CMS

Companies typically go headless for two reasons. The first is freedom to build. A headless CMS lets developers pick their own frontend frameworks and connect the tools they need, such as a PIM, a commerce engine or a search service. Traditional CMSs can lock developers into fixed templates, so business requests get turned down for technical reasons.

The second is reuse. With a content as a service approach, content lives in the CMS as modular blocks and flows to any channel through APIs. The same API that feeds your website now feeds AI assistants and agents too, so how well your content is structured matters more than which frontend shows it.

A headless CMS is probably not the right fit if:

• You run one simple website and have no developer time to build or maintain a frontend.
• Your editors need a page builder only and nothing will ever reuse the content elsewhere.
• You need a site live this week with a ready-made theme.

#8 headless CMS use cases and examples

Here's how 8 teams put a headless CMS to work, from several brands on one platform to streaming and publishing.

1. Several brands from one platform: Vision Healthcare

Vision Healthcare, the consumer health group behind VitaminExpress, grows through a buy-and-build model. Each acquired brand used to come with its own development team and its own stack, which made every new brand a new technical project.

Today, Vision Healthcare runs 13 brands on one composable stack, with Hygraph as the content layer next to commercetools for commerce. The in-house platform team in Haarlem migrates acquired brands off their legacy platforms and onto this shared foundation. Every model carries a tenant field, and role-based views mean editors only see the content of the brands they work on, as the Vision Healthcare case study shows.

The result is central control over the structure with room for each brand to run its own content. If you're weighing the same setup, here's our guide to running several brands from one CMS.

The API-first foundation also pays off for AI. It let the team ship a custom AI-powered compliance tool in a matter of days. Diederik Oostburg, who leads eCommerce engineering and technology at Vision Healthcare, covers compliance tooling and AI in practice in the webinar.

2. AI and agent-first experiences: MAD Design Group

MAD Design Group runs a portfolio of premium design brands, from fireplaces to outdoor heating. Its digital team is one of our most active MCP users: they handle content creation, translation, content architecture and web development with Claude, connected to Hygraph through the MCP server.

Hygraph is the tracker for their AI content pipeline. Status fields mark each stage of a piece, from research to outline to draft. Each agent picks up the entries at its stage, does its job and moves them on. Agents only save drafts. A person approves in Hygraph, and that approval is what triggers translation.

Governance is built in from the start. Each job gets the narrowest token it needs, read only for research and content only for writing, and every new AI workflow starts in a staging environment before it touches production. We cover the full setup in our guide to what makes a CMS agentic.

Here, AI is the way the content gets made, and Hygraph keeps it structured, reviewable and in one place. Ederson Morche, Head of Digital at MAD Design Group, walked through the setup in our Mastering MCP for CMS webinar. To see what agents can do inside your own workflows, explore Hygraph AI Agents.

3. eCommerce across regions: Lindex

Lindex, one of Europe's leading fashion companies, has about 440 stores in 16 markets and sells online in 32 eCommerce markets. Its legacy CMS slowed every update: a small text change took two to three days before translations even started, and editors needed a developer to unlock modules.

With the move to Hygraph, developers and editors rebuilt the content model together, breaking one big module into smaller, clearer blocks. Content is now created once and delivered to both web and app, with no duplicate work between the two teams. New markets launch in 2 weeks instead of 2 months, and campaign setup dropped from a week to a day or two, according to the Lindex case study.

AI is already part of the daily workflow. Lindex built its own translation service, trained on its tone of voice, and local market teams review the AI translations in Hygraph with the full page in front of them.

Read more about running eCommerce content on a headless CMS.

4. B2B portals: Topps Tiles

Topps Tiles, the UK's largest specialist tile retailer with 280+ stores, built the Topps Trade App with mobile consultancy Apadmi so tradespeople can check stock, trade pricing and loyalty offers from the job site. Its existing CMS couldn't deliver content to a native app, and the business wanted each trade type to see what matters to them.

Hygraph now powers the content in the app's Flutter frontend, from the homepage to banners and carousels. Variants in the content model tailor the app by trade type, so a tiler and a builder get different views of the same app without anyone duplicating content. Product data stays in Magento and comes in through Remote Sources, and Bloomreach handles segmentation.

It's paying off: the Topps Trade App now accounts for 50% of Topps Tiles' digital trade sales. See how teams build B2B portals for partners, suppliers and trade customers.

Topps Tiles doesn't describe using AI here, but the structure is ready for it. Because trade variants live in the content model and product data comes from one source, any new channel, an assistant included, gets the same typed answer for each trade customer.

5. Localization across countries: Voi

Voi, the shared electric micromobility company, operates in 12 countries and publishes in 10 languages. On WordPress, even a chatbot script change went through its agency, and complex templates turned design tweaks into projects.

The team migrated to Hygraph in one to two months. Right after the move, website traffic doubled and maintenance costs halved, as the Voi case study reports. With the Lokalise integration, launching content in a new language now takes just under two weeks, and the brand team owns content updates end to end. The team shared the full story in a webinar on scaling across 12 countries.

Each locale is a structured set of fields connected to Lokalise, so translation tools, and AI agents later on, can work field by field instead of page by page. That's the groundwork for headless CMS localization at any scale.

6. Media and streaming: Telenor

Telenor is a leading Nordic mobile, broadband and TV service provider. It had built a homebrew content solution to launch its video streaming service, and as the service grew, developers spent their time manually editing thousands of metadata entries each month.

A headless CMS treats all content as structured data, which makes it a strong fit for metadata management. Telenor uses Hygraph's schema builder to define the metadata structure and adds, edits and deletes entries in bulk with GraphQL mutations, while the content team edits by hand in the UI.

Today, Telenor programmatically adds 2K new videos a month and makes millions of monthly API calls at under 100ms latency. More than 90% of its branded metadata is served from Hygraph, with content federation pulling from external APIs. That clean metadata layer is exactly what recommendation engines and assistants need. Explore more headless CMS setups for media and streaming.

7. B2B publishing and knowledge: BioCentury

BioCentury offers a robust knowledge base for C-level biotech executives and investors. Its monolithic setup and rigid, template-based publishing process made it hard to deliver more content and improve the user experience.

Hygraph lets BioCentury create any form of content it imagines using structured content. The team publishes and maintains content programmatically, with Hygraph connected to CRM and eCommerce systems. The time from idea to publishing has decreased by 81%, and content engagement is up 120%, as the BioCentury case study shows.

For AI, this kind of structured, verified data is the source answers can cite with confidence. Learn more about B2B publishing platforms on a headless CMS.

8. Omnichannel member platform: Samsung

Samsung Electronics Germany's membership platform started as a mobile app, but a web version became critical as the loyalty program expanded. Replicating the app on Adobe Experience Manager's rigid templates was slow: any component that differed from the global templates had to be built statically by an external agency and pass a lengthy global governance process.

Moving to a headless CMS lets the team structure content once and deliver the same experience to the mobile app and a single-page application on the web. With Hygraph's API-first CMS, frontend developers build and test local features without touching the global backend.

Content managers now add and edit content on their own. Page updates take half the time, and Samsung reports a 15% higher engagement rate on frequently updated pages. One content model feeds web and app, so an AI assistant becomes one more channel on the same model. That's the idea behind an omnichannel headless CMS.

#What to plan for when adopting a headless CMS

Moving to a headless CMS is a shift in how your teams work. Here's what to get right early.

Content modeling

A headless CMS usually starts as a blank slate, so you can create any schema you need and balance what developers and editors expect. That flexibility also makes it easy to build structures that are more complex than you need.

Before you move, assess your content and map out how it fits the new system. An efficient schema keeps both developers and editors productive. Our guide to content modeling is a good place to start.

APIs

Headless CMSs typically provide REST or GraphQL APIs that let developers retrieve, create, update and delete content and deliver it to any framework, platform or device.

Check each API's limits before you choose. Rate limits on requests per minute or hour matter, because exceeding them can lead to throttling or service interruptions.

Integrations

A headless CMS should connect with the rest of your stack: commerce platforms, marketing automation, analytics and more. Hygraph offers three ways to extend it:

• Ready-made apps in the Hygraph Marketplace, built by Hygraph or its partners.
• The App Framework to extend the Hygraph UI with your own integrations.
• Remote Sources to pull data from external systems into the Hygraph API without migrating it, so your frontend uses one endpoint.

Governance

The more teams, brands and markets share a CMS, the more you need clear rules. Plan roles and permissions, review and approval workflows, and audit logs from day one. Read how Hygraph handles content governance.

AI readiness

Structured content is what assistants and agents can actually use, so a clean model is your first AI investment. Hygraph's MCP server connects Claude, Cursor, VS Code and Windsurf to your content model, with delete and unpublish blocked. Output from Hygraph AI Agents lands in draft and waits for your team's review.

#Where headless CMS is heading

Look across these 8 examples and one pattern stands out: a single structured layer of content serves websites, apps, portals and, more and more, AI agents.

In our survey of 400 technology leaders, 84% said their current CMS keeps them from unlocking the full value of content. The issues they reported most were that only a small group of people can make changes, that new data and content types are hard to add, and that the CMS is hard to integrate with other systems. The teams above solved each of these with structured content.

We spent the past few years putting AI into the CMS. Now the CMS is going into AI. Hygraph, the agentic content platform and headless CMS, is built for that: one governed content layer that your channels, your teams and your agents all draw from.

Find the setup closest to yours in our use cases, or talk to our team.