Headless Shopify or a Shopify theme? A working decision framework

A founder asked us last week: "Should we go headless with Shopify, or just customize a theme?"

Our answer is usually the boring one. For most brands, at most stages, a well-built theme wins on speed to market, cost of ownership, and merchant sanity. Headless earns its keep in a specific set of situations — and outside those situations, it quietly eats budget for years.

Here's the framework we walk merchants through, before we quote.

The comparison at a glance

	Shopify theme (Liquid + Dawn)	Hydrogen (Shopify's headless)	Custom headless (Next.js + Shopify API)
Time to launch	4–8 weeks	10–14 weeks	12–18+ weeks
Ongoing dev cost	Low	Medium	High
Merchant-owned edits	First-class	Limited without dev	Limited without dev
Custom PDPs / bundles	Workable via app + sections	Native	Native
Content velocity	Sections + metaobjects	Any React pattern	Any React pattern + CMS of choice
Peak performance	Very good with discipline	Excellent	Excellent
Best for	Brands prioritising speed, iteration, editor autonomy	Brands committed to Shopify's stack that need headless flexibility	Brands with content-heavy storefronts, non-Shopify CMS, or unusual product logic
Failure mode	Fighting Liquid on the tenth edge case	Locked to Hydrogen's abstractions	Slow to ship simple content edits

The pattern is consistent: as you move right, you gain flexibility and lose speed. Not just at launch — every content update, every A/B test, every seasonal campaign moves slower once the storefront lives outside Shopify's admin.

The four criteria that actually matter

Every merchant we've quoted headless work for asks about the same four things. In our experience, these are the criteria that predict the right answer.

1. Traffic scale. Under 100k monthly sessions, headless usually doesn't earn back its cost. The performance ceiling on a well-built theme is high enough that few brands are actually held back by it. Above a million monthly sessions, the argument changes — LCP and INP compound into real revenue, and the engineering budget is easier to justify.

2. Content velocity. How often does the storefront change? A brand shipping a new campaign every two weeks — new PDPs, new landing pages, new module variants — is a very different animal from a brand shipping four campaigns a year. High velocity favors editor- first tooling: Shopify sections + metaobjects, or a headless build with a CMS the marketing team actually likes.

3. Custom PDP ambition. How weird is your product? Configurators, bundle builders, subscription previews, virtual try-on, marketplace listings from multiple vendors — the further your PDP is from a "choose a variant and add to cart" pattern, the more Liquid fights you and the earlier headless earns its keep.

4. Brand and storefront ambition. For brands where the storefront is the marketing — editorial homepages, magazine-style category pages, motion-heavy product stories — the theme model is usable but constraining. Headless with a proper CMS (Storyblok, Sanity, or similar — see how we choose between them) is usually the honest answer.

The two questions that decide it for eighty percent

Once we've named the four criteria, the decision is usually made by two of them. The rest are noise.

Question 1: Who edits the storefront day-to-day, and how often? If the answer is "marketing, weekly," your risk is not performance. It's how many things they can safely change without a dev. A theme with a well-modelled sections library and metaobject- driven PDPs beats a headless build for that team, every time.

Question 2: Is there a product experience Liquid can't cheaply build? A configurator that needs to talk to a pricing engine. A bundle builder that lives outside Shopify's native bundles. A marketplace layer. If yes, you're a headless candidate.

Two yeses to velocity + editor autonomy → theme. A yes to custom product experience Liquid can't reach → headless. Everyone else sits in the middle, and we default to theme.

Where Hydrogen fits

Hydrogen is Shopify's headless framework. It's the right tool for brands that:

• Are committed to Shopify long-term
• Want headless flexibility without leaving the Shopify ecosystem
• Value tight integration with Shopify's storefront APIs, cart, checkout, and Oxygen hosting
• Have an in-house or agency team fluent in React

Hydrogen is not the right tool if you want to swap Shopify for a different commerce backend later, or if your content lives primarily in a non-Shopify CMS and you'd rather not have two source-of-truths. In that case, a custom Next.js storefront hitting the Storefront API gives you more room, at the cost of more engineering.

The specific things that surprise merchants

Every headless Shopify project we've run has surfaced one of these around week six. We now flag them at scoping.

1. Checkout stays Shopify. Shopify's checkout is not fully customizable, even in headless. This is by design — it's PCI-hardened and protects conversion rate. Merchants sometimes discover this too late and expect us to "just build our own." We don't recommend it, and we will push back hard.

2. Every content edit is now a code deploy. Unless you pair the headless storefront with a CMS the marketing team owns, every copy change goes through a developer. This is the single biggest source of regret in headless builds.

3. The theme apps you rely on may not have headless equivalents. Review widgets, upsell modules, quiz builders — many were built for Liquid and don't have Storefront-API-friendly versions. Every one of them becomes a custom build or a compromise.

4. Performance is not free. Headless doesn't automatically ship faster than a well-optimised theme. Getting to top-quartile Core Web Vitals still takes work — server components, image discipline, edge caching. Skipping that work leaves you with the cost of headless and none of the payoff.

The scoping shape

For most Shopify projects we ship, the shape is:

• Weeks 1–2: Discover + a working prototype of the two hardest screens (typically the PDP and the storefront homepage). This settles the theme-vs-headless question with data, not opinion.
• Weeks 3–8: Build against the locked scope. Ship the sections library, metaobject models, and any custom logic.
• Weeks 9+: Launch, monitor, and iterate against a stated goal.

If you want the long version of that shape, see the five phases we run every project through.

The answer

Theme unless there's a specific reason we can point to. Hydrogen when you're committed to Shopify and Liquid can't reach the product experience. Custom headless when the storefront itself is a content product and your CMS lives elsewhere.

If you're weighing this call for a specific store, we're happy to walk through it. See how we scope Shopify ecommerce, or send us a short brief. We reply within one business day.