We build headless sites. Our own site runs on Payload inside a Next.js app, and we set up Sanity and other headless systems for clients. So this isn't an argument against headless. It's an argument against choosing it by default, because for many startups it's more system than the job needs.
If you want the bigger picture of what a good editing setup looks like on any platform, start with our hub on CMS setup for marketing websites. This article answers one narrower question: do you need headless at all?
When is a headless CMS overkill?
A headless CMS is overkill when its main benefits, content reuse across channels, flexible modeling and full front-end control, solve problems you don't have. For a single marketing site with a blog, case studies and a handful of landing pages, edited by one to three marketers, a visual builder with a built-in CMS usually does the job with less to maintain.
The pattern we see most often: a startup goes headless because it sounds more serious, then discovers that every new page layout needs a developer, preview takes extra work to build and the marketing team edits less than before. The CMS wasn't wrong. It was the wrong size. A headless CMS gives you a content store and nothing else, so the front end, preview and page building all have to be built.
What are the downsides of a headless CMS for a small team?
The downsides of a headless CMS for a small team are a second system to run, a front end that depends on developers, and editing features, like preview and page building, that you have to build rather than get for free.
Pros
- Content is structured and reusable across a site, an app and email
- Full control of the front end, performance and code
- Flexible content models with real relationships
- Easier to redesign without moving content
Cons
- A separate front end to build, host and update
- Layout changes often need a developer
- Preview and visual editing take setup work
- More vendors, more bills and more things that can break
None of those cons is fatal for a company with an engineer or a studio on retainer. They're expensive for a two-person marketing team without one. That's also why website hosting for startups gets more complicated on headless: you're hosting an app, not just a site.
Headless vs traditional CMS or a builder: which fits a startup?
For most early-stage startups, a visual builder like Framer or Webflow fits better than either a headless or a traditional CMS, because it bundles design, CMS and hosting in one tool that marketers can use. Headless fits once content or engineering needs outgrow that bundle.
Situation | Better fit | Why |
|---|---|---|
One site, small team, no developer | Framer or Webflow CMS | Marketers edit layout and content in one place |
Designer-led team, frequent page launches | Framer or Webflow CMS | New pages without tickets |
Content also feeds an app or docs | Headless CMS | One source, many front ends |
Complex models: integrations, changelog, locales | Headless CMS | Relationships and structure at scale |
Engineers own the site in the product repo | Headless CMS | Code review, types and one stack |
Deep plugin needs, existing content site | Traditional CMS | Mature ecosystem already in place |
Plan limits matter too. Framer and Webflow both cap CMS items and collections by plan, so check the current plan pages if you publish at volume. We build on Framer CMS when the team wants to own layout, and our guide to keeping Framer's design while moving content to a headless CMS covers the middle path.
What are the signals that justify going headless?
Three signals justify going headless: the same content must appear in more than one place, the content model has relationships a builder handles awkwardly, and engineers will own the site as part of the product. One strong signal is enough; none means stay on a builder.
Content has more than one destination
Case studies, docs or a changelog also need to appear in your app, help center or emails. A headless CMS publishes once and every front end reads it.
The model has real relationships
Integrations link to categories, use cases and customers; posts link to authors and products; pages exist in several languages. Our guide to content modeling shows what that looks like.
Engineers own the stack
Your developers want the marketing site in the same repository, with typed content, code review and the product's design system.
You're hitting builder limits
Item caps, reference limits or rich text that can't hold your components. Our guide to moving from Webflow CMS to headless lists the specific limits.
Localization is central
Many locales with translators working on fields, not whole pages. Our headless CMS localization guide compares the options.
What does going headless cost compared with a builder?
Going headless costs more up front, because you build a front end, preview and page blocks that a builder gives you for free, and it adds ongoing developer time to maintain them. A builder costs a predictable monthly plan and less engineering.
Quotes we see for moving a WordPress site to headless Next.js typically range from about £13,000 to £30,000 and 6 to 12 weeks, and they vary widely with page count and content. That's before ongoing upkeep: framework upgrades, CMS upgrades and the occasional broken webhook. For a team that will use the flexibility, it's worth it. For a team that won't, it's money spent on capacity they never touch.
Hypothetical example: two startups, same size, different answers
For example, suppose two 15-person startups each want a new marketing site.
The first sells a single product, has one marketer and a freelance designer, and launches a landing page every few weeks. Framer or Webflow CMS is the right call: the marketer builds pages from designed sections and nobody waits on a developer.
The second has an integrations directory, a public changelog that also shows inside the product, and docs in three languages, with two frontend engineers who want it all in the product repo. Headless is the right call, because content has several destinations and engineers will own it.
Can you start on a builder and go headless later?
Yes. Starting on Framer or Webflow and moving to headless later is a sensible path, as long as you model content cleanly from the start. Structured collections with clear fields move to a headless CMS far more easily than pages built from freeform text.
The migration is real work: export, schema mapping, media re-hosting, redirects and a new front end. Our headless CMS migration guide covers the steps. But it's work you do when the need is proven, not on a guess. If you're not sure which side of the line you're on, a short paid test can answer it: our guide on what a trial week can realistically ship explains how we use that week.
