Is 'we need a non-commerce content experience wrapped around the store' (editorial, brand storytelling) a legitimate headless trigger, or can a theme handle that too?
An editorial layer is not a headless trigger on its own. A Shopify template holds up to 25 sections and 1,250 blocks with eight levels of nesting, and metaobjects are readable in Liquid and through the Storefront API (Shopify, September 2026). A headless CMS attaches to a Liquid theme without replacing the storefront.
The decoupling people actually want is the CMS, not the storefront
These two get collapsed into one decision constantly, and they are separate.
A headless CMS means your content lives in a system built for editors, with its own model, workflow, versioning and roles, and it is delivered by API. A headless storefront means your customer-facing pages are rendered by your own application instead of by the Online Store.
You can have the first without the second. A Liquid theme can fetch and render content from an external CMS, and Shopify's own metaobjects give you a structured content model natively, accessible "in themes using Liquid and through the Storefront API" (Shopify Help Center, September 2026).
Shopify's enterprise content draws the same line: "Shopify themes use an integrated CMS that connects content, design, and commerce in one system, making them faster to manage and easier to operate for most storefront-led brands," and "for teams focused on speed, simplicity, and storefront performance, themes often deliver more value, while headless makes sense when multi-channel content reuse is a core requirement" (shopify.com/enterprise, September 2026).
Multi-channel content reuse. Not editorial ambition.
What a theme will actually do for an editorial program
The ceilings are published, and they are high:
| Capability | Published ceiling |
|---|---|
| Sections per template | 25 |
| Blocks per template, across all sections | 1,250 |
| Nesting depth for theme blocks | Up to eight levels |
| Structured content | Metaobjects, with fields, readable in Liquid and the Storefront API |
| Editor composition | Theme blocks, described by Shopify as "blocks that are defined within your theme and can have dynamic functionality, such as nesting" |
(All per Shopify Help Center, September 2026, and the Horizon changelog of 21 May 2025.)
A long-form editorial article with pull quotes, inline product modules, a gallery, a video and a shoppable footer sits well inside those numbers. A campaign hub with six landing pages, each with a distinct layout, sits inside them too.
The three things that genuinely strain a theme
Be honest about where it does break, because "the theme can do it" is not a universal answer.
- Editorial workflow. Draft states, scheduled publishing, approvals, multiple editors, localized variants of the same article. Shopify's native content tooling is thin here, and this is the real reason brands buy a CMS. It is also solved by a headless CMS with a Liquid front end.
- Content reused outside the website. Newsletters, an app, retail screens, a partner site. Once the same article renders in three places, an API-delivered content model earns its cost, and this is the condition Shopify itself names.
- Very large content archives with editorial taxonomies. Thousands of articles with faceted browse and cross-references start to strain what blogs and metaobjects were designed for.
None of those three requires a headless storefront. All three are arguments for a headless CMS.
The order of operations we recommend
- Write down the five editorial capabilities you are missing, in editor language, not architecture language.
- Check each against theme blocks and metaobjects. Most of the layout items resolve here.
- Whatever remains is almost always workflow. Price a headless CMS against a Liquid front end.
- Only if the content must render outside the website does the storefront itself come into scope.
- If you reach step four, you now have a multi-surface problem, which is the legitimate headless case.
When NOT to decouple the storefront for content
- When the complaint is that the blog looks bad. That is a theme problem with a theme fix.
- When one person publishes twice a month. Workflow tooling for a workflow of one is a cost with no beneficiary.
- When the CMS decision has not been made yet. Choose the content model first. Deciding the rendering layer before the content model is backwards, and the rendering layer is the expensive half.
- When SEO is the stated driver. Moving the storefront to a custom application puts every technical SEO behavior the Online Store gives you for free onto your own team.
The Deploi point of view
Our own position, from building on Shopify. Separate from the facts above.
- Our take: An editorial requirement justifies a headless CMS, and a headless CMS does not require a headless storefront. We will say that even though Headless CMS is one of the six services Deploi sells, because the version of that project we would rather ship is a content model behind a Liquid theme, and it is the cheaper one.
- What we’ve seen: When we take the "we need editorial" brief apart, the items sort into layout and workflow, and layout is nearly always solvable in the theme. Workflow is the real gap and it is the one that never appears in the original brief, because the people who feel it are editors and the people writing the brief are usually not.
- Where we disagree: The category treats "content-led brand experience" as a headless trigger, which conveniently makes the largest possible version of the project the correct one. Shopify's own guidance sets the bar at multi-channel content reuse, and a brand publishing to one website does not clear it.
- What this page adds: the specific published ceilings a theme has to be pushed past before the argument holds (25 sections, 1,250 blocks, eight nesting levels, metaobjects in Liquid), and the separation between a headless CMS and a headless storefront that makes most editorial briefs a content project rather than an architecture project.
Reviewed by Martin Dejnicki, Director of SEO & AI Search. Facts verified 2026-09-14.
Where we worked this out
Our decision records