Headless Shopify is a hammer that’s been on every consultant’s wall since 2021. Most stores aren’t a nail. This is the honest framework for when headless actually earns its keep.
The default story (wrong)
The default pitch goes: “your store is slow, headless will fix it.” This is technically true and operationally misleading. Almost every slow Shopify store I audit is fixable on OS 2.0 with app removal + image pipeline + Liquid refactor — for 1/10 the cost of headless and with no operational surface added.
If “speed” is the entire reason you’re considering headless, do the OS 2.0 speed work first. If you hit a real ceiling at sub-2s LCP and INP that genuinely can’t be lowered further on Liquid, then revisit headless. Most stores never get there.
When headless does earn its keep
Five clear cases:
1. Global multi-region commerce beyond Shopify Markets. If you’re running multiple legal entities in multiple regions, each with their own checkout flow, payment methods, pricing logic, and content — and Shopify Markets is too rigid for your structure — headless gives you the routing flexibility.
2. CMS-heavy editorial commerce. Brands where the content layer (Sanity, Storyblok, Contentful) is the primary surface and product purchase is the call-to-action. Magazine-style commerce, deep storytelling brands, content-first DTC. The CMS is the system of record, Shopify is the cart.
3. Complex configurators that need true SPA performance. Custom jewelry, made-to-order furniture, bespoke fashion. Where every interaction needs visual feedback in 50ms, not a 300ms Liquid re-render. Liquid genuinely can’t match SPA responsiveness here.
4. Native mobile apps sharing the same API layer. React Native iOS/Android apps that use the same Storefront API + cart + checkout as the web storefront. Headless makes the API layer the shared substrate; without it, you build two separate apps.
5. UX polish that exceeds Liquid’s ceiling. React Server Components, Suspense streaming, partial prerendering. These give you a level of perceived performance Liquid can’t reach. Rare, but real for premium brands.
Where headless makes things worse
Three honest costs:
Content editing is harder. Your merchandising team can no longer use Shopify’s theme editor. They learn Sanity / Storyblok / Contentful. Real training cost, real productivity drag for 6-12 months.
Operational surface grows. You add deployment infrastructure (Vercel / Oxygen), CMS infrastructure, build-pipeline maintenance, environment-variable management, security patches across the stack. All things Shopify handles for you on OS 2.0.
Apps stop working. Klaviyo onsite, ReCharge, judge.me, Yotpo — all need API-level reintegration. Each one is custom work; some apps just don’t have headless-friendly APIs.
The honest test
Three questions:
- Have you maxed out OS 2.0 + speed work + app removal? If no, do that first.
- Does your business genuinely fit one of the five “earns its keep” cases? Be honest.
- Can your team absorb the content-editing transition + operational complexity?
If three yeses, headless is the right call. If any no, it’s the wrong call.
If you’re going headless — Hydrogen vs. Next.js
Hydrogen on Oxygen. Shopify’s framework, Shopify’s hosting. Smaller operational surface. Tighter Storefront API integration. Better for stores that want Shopify to manage as much as possible. The default unless you have a reason to go Next.js.
Next.js on Vercel. More flexibility — better for CMS-heavy stores that need rich content infrastructure, multi-region routing, integrations beyond commerce. The default for content-driven brands and stores that want to share the API layer with native mobile apps.
If you went headless and regret it
It happens. The path back is rebuilding the storefront on OS 2.0, decommissioning the headless front-end, routing all traffic to Shopify-managed Liquid. Cost is roughly 60-70% of the original headless build cost. Not cheap, but recoverable. The right time to migrate back: when your team’s frustration with content editing has cost you more than the technical advantages of headless are buying.
For honest headless scoping, see /headless-shopify-developer/. I’ll tell you on the scope call whether your case fits.
