What Is a Headless CMS, and When Should You Choose One?

A headless CMS separates the back end that manages content from the front end that displays it. How it differs from a classic CMS, the advantages it brings, its costs such as preview and SEO, and when it should be chosen.
Technical summary
Technology
Headless WordPress, Strapi, Directus, Sanity
Topic
Headless architecture
Level
Start
In this article7 sections

Most headless CMS projects struggle not because of the technology but because of wrong expectations: the team wants a faster site and ends up with an editor experience where preview doesn’t work and every page layout change waits for a developer. A headless CMS is a content management approach in which the back end where content is managed is separated from the front end where it’s displayed. “Headless” refers to the system having no theme or page template layer of its own. Content is served through an API (REST or GraphQL); a website, mobile app, kiosk screen or partner portal uses that content in its own interface.

How it differs from a classic CMS

In a classic CMS (e.g. WordPress running with a theme), content, templates and HTML generation all happen in the same system. The editor edits the page and the system generates it. In a headless setup, the editor enters content into structured fields (title, summary, technical specifications, images, related products); the front end decides how that content will look.

Classic CMS Headless CMS
Front end CMS theme Independent application (Next.js, mobile, etc.)
Content format Mostly pages/HTML Structured fields and relationships
Number of channels Usually a single website Multiple channels
Page design freedom (editor) High (page builder) Controlled, with defined blocks
Development needed Less Front-end development required

The most decisive row in the table is “content format”. In a classic CMS, a product page is an HTML block containing a title, text and images; the system doesn’t know which part of that text is a technical specification and which is a description. In a headless setup, each field is defined separately. This way the same product record can be shown as a full page on the website, a card in the mobile app and a comparison table in the dealer portal. In return, the content model has to be designed correctly from the start.

The advantages of headless

  • Manage once, use everywhere: The product description is entered once; the website, catalog app and dealer portal are fed from the same source.
  • Independent front end: The front-end team works with modern tools without being held back by the CMS’s theme constraints; a front-end redesign doesn’t affect content.
  • Performance: The front end can be served very fast with static generation and a CDN.
  • Smaller attack surface: Content can be read through the API only, without exposing the admin panel to the public internet.

The attack surface advantage doesn’t come automatically. If the admin panel is still left open to the internet, the API read key is embedded in the front-end code and that key can also access draft content, nothing is gained. Restrict access to the admin panel with a VPN or IP restrictions, and use separate tokens for published and draft content.

Costs and risks

  • The front end must be developed; there’s no ready-made theme.
  • Preview, redirects, sitemaps and SEO metadata are set up separately on the front end.
  • Two systems (CMS and front end) are hosted and maintained separately.
  • Editors work with predefined blocks instead of free-form page design.

The gap we see most often in the field is redirects. If a site moving from a classic CMS to headless forgets to 301-redirect old URLs to their new addresses, the search visibility built up over the years is lost. The redirect list, sitemap generation and canonical tags must be ready in the first version of the front end; SEO work that’s left for “later” is usually done after traffic has dropped.

Headless WordPress preserves existing WordPress knowledge and editor habits; it serves content via the REST API or WPGraphQL. Plugin and security maintenance continue. Strapi is a Node.js-based, open-source system that runs on your own server, where you can define content types from the admin panel. Directus adds an admin panel and API on top of an existing SQL database; it suits teams that want to keep data independent of the application. Sanity, as a managed service, offers real-time collaborative editing and powerful content querying; pricing is usage-based and data location should be evaluated.

Choose based on the operating model, not the feature list. Strapi or Directus running on your own server gives you control of the data but also hands you responsibility for version upgrades, backups and monitoring. A managed service takes on this load; in return come usage-based costs and the question of where the data is kept. For content that contains personal data, data location must be clarified at the contract stage.

When should you choose it?

  1. If the same content will be used on at least two different channels.
  2. If the front end requires app-like interaction or very high performance.
  3. If the content is strongly structured (products, documents, regulatory articles, catalogs).
  4. If the front-end and content teams want to work at independent speeds.

If there’s a single corporate website, a small marketing team and frequent page design changes, a classic CMS is usually more efficient. For this decision, you can also read the WordPress or Next.js guide.

If none of these four conditions apply, performance alone isn’t a sufficient reason to go headless. A well-cached classic site is fast too. The difference comes from implementation quality, not the architecture.

Checklist for a successful headless project

  • Has the content model been validated with real content samples?
  • Do preview and scheduled publishing work for editors?
  • When content is updated, are the related pages regenerated via webhook?
  • Can API read tokens access only the content types they need?
  • Are the backup, version upgrade and media storage plans ready?

The first item on the list is the one most often skipped. The content model is designed on a blank slate, and when real product and page samples are entered it turns out the fields aren’t enough or are unnecessary. Before finalizing the model, enter your ten most complex pieces of content as drafts; the missing fields show up there, not after go-live.

Frequently asked questions

Does a headless CMS hurt SEO?

No; SEO depends on how the front end is rendered. If content is generated as HTML on the server (SSG/SSR), search engines crawl it without problems. Metadata, canonicals and sitemaps must be set up separately on the front end.

Can editors see the page before publishing?

Yes, but preview has to be set up deliberately: the CMS sends draft content to the front end’s preview mode with a secure token. If this isn’t set up, the editor experience suffers badly.

What happens to existing content when moving to a headless CMS?

Once the content model is designed, existing pages are transferred into structured fields with scripts. Free-form layouts built with a page builder may need to be restructured by hand.

For headless architecture setup and CMS selection, see our Headless CMS and API-first web architecture service.

Support

Contact us for support or a discovery session on this topic.

Contact form
[email protected]

Eretek Web

We build and maintain corporate websites, WordPress and CMS platforms, intranet and SharePoint portals, document management, and form and approval workflows, and keep them secure and up to date.

Knowledge Center

Other articles