TL;DR
- Headless commerce separates frontend from backend. The store and e-commerce engine operate independently, communicating via API.
- For 80% of SMB stores: headless is overengineering. Shopify, WooCommerce, or BigCommerce will suffice and cost less to maintain.
- Headless makes sense with: 10,000+ SKUs, custom checkout, multi-region operations, very high Core Web Vitals requirements.
- Three platforms worth considering: Shopify Hydrogen (for stores already on Shopify), Medusa.js (open source, self-hosted), Saleor (GraphQL-first, enterprise-ready).
- Headless build cost: $5,750-18,400 (25,000-80,000 PLN). Maintenance cost: 2-5x higher than monolith.
Headless commerce: when is it revolutionary, and when is it overpaying?
Headless commerce appears at every e-commerce conference as "the next level." In practice, for most SMB stores, headless is a more expensive tool solving a problem that doesn't exist.
Before I explain when headless commerce makes sense, let me say when it doesn't: for a store with 50-2,000 products, one language version, standard checkout, and a team without full-stack developers. That describes 80% of SMB stores. For them, a monolith (Shopify, WooCommerce, BigCommerce) is better, cheaper, and safer.
This isn't a trendy opinion. It goes against what e-commerce agencies write, because headless means higher project budgets. But honest analysis leads to the same conclusion.
How does headless commerce work in practice?
In a traditional store (monolith), frontend and backend are one application. Shopify renders pages, handles cart, checkout, payments, and CMS in one system. You don't need to configure anything for the store to work.
In headless commerce architecture, the frontend lives separately: it's a Next.js, Nuxt, Remix application, or static site (Astro, Gatsby). It communicates with the e-commerce backend via API (REST or GraphQL). The backend handles products, orders, payments. The frontend displays them however it wants.
Benefits of this separation:
- Complete frontend freedom: you can build any UX without platform template limitations
- Performance: frontend as static site + CDN = fastest possible Core Web Vitals
- Omnichannel: the same backend serves store, mobile app, in-store kiosk
- Independent deployment: changing frontend doesn't require touching backend and vice versa
Drawbacks:
- Complexity: two systems instead of one, two deployments, two monitoring setups
- Cost: more expensive build, more expensive maintenance, requires developer knowing both stacks
- Time to market: MVP in monolith: 4-6 weeks. Headless MVP: 10-20 weeks minimum
Three headless platforms worth considering for SMBs
Shopify Hydrogen
Shopify Hydrogen is Shopify's official React meta-framework for building headless storefronts. Advantage: Shopify's backend features (checkout, payments, inventory) remain. You only change the frontend. Good choice if you already have a Shopify store and want full frontend control without backend migration.
Drawbacks: vendor lock-in to Shopify (backend and hosting). Shopify Plus costs (required for full Hydrogen capabilities) are $2,300/month.
Medusa.js
Medusa is an open-source e-commerce platform for developers. Self-hosted, GraphQL and REST API, modular architecture (you can swap payments, shipping, CMS separately). No monthly platform fees.
For whom: teams with full-stack devs who want complete control over code and data. Good option for stores with custom business logic (custom pricing, B2B, wholesale).
Drawbacks: no managed hosting (you must deploy yourself), smaller community versus Shopify/WooCommerce, fewer ready integrations with regional systems.
Saleor
GraphQL-first platform, open-source, strong in enterprise. Advanced multi-channel sales management, multi-currency, multi-language out of the box. Managed hosting available in Cloud.
For whom: stores with serious multi-region and omnichannel requirements. Comes into play with budgets above $23,000 (100,000 PLN) per project.
Headless Shopify (Hydrogen): when does it make sense?
Hydrogen is the right choice in four specific situations:
1. Core Web Vitals are an SEO blocker
If Google PageSpeed Insights and GSC show LCP above 3.5s and INP above 300ms, and all monolith optimizations have been exhausted: Hydrogen with SSG/ISR can deliver 0.8-1.5s LCP. This matters for competing on Google's first page for competitive keywords.
2. You need custom checkout
Shopify Plus allows checkout customization, but only within platform limits. Headless frontend: checkout can be fully custom (multi-step, product configurator, conditional fields), without template restrictions.
3. Store + mobile app + POS with one backend
Three frontends (web, mobile, point of sale) communicating with shared inventory and orders: this is a natural headless use case. Alternative: three separate systems with manual synchronization. Headless is cheaper long-term here.
4. Content managed by CMS with advanced workflow
Large stores with high content change frequency (daily promotions, banners, campaign landing pages) benefit from combining Shopify (backend) with headless CMS (Sanity, Contentful, Storyblok). Marketers can change content without deployment.
When headless commerce hinders instead of helps
From my experience: most companies come with a headless idea after reading technical articles, before checking whether they have a problem that headless solves.
Specific signals that headless is NOT the right answer:
Small product catalog (< 2,000 SKUs) without expansion plans. Monolith is fast enough if properly configured. Image optimization, cache, CDN, and lazy loading solve 90% of performance problems without changing architecture.
No in-house full-stack developers. Headless requires maintaining two systems in two technologies. If you don't have a developer who knows Next.js and the store API: every change is an expensive external order. Monolith is cheaper to maintain in this scenario.
Timeline under 3 months for MVP. Headless can't be sensibly delivered in 4 weeks. If you need to sell quickly: monolith, deploy, optimize after launch.
Budget below $9,200 (40,000 PLN). Below this amount, you can't build solid headless with proper architecture, tests, and documentation. What you'll get for less: a prototype that won't handle traffic and will generate technical debt instead of revenue.
Decision table: headless commerce or traditional e-commerce?
| Criterion | Headless | Monolith |
|---|---|---|
| Product catalog | 10,000+ SKUs | up to 5,000 SKUs |
| Language versions | 5+ languages, multi-currency | 1-3 languages |
| Checkout | Custom, multi-step | Standard platform checkout |
| Frontends | Web + mobile + POS | Web only |
| CWV requirements | LCP < 1s, INP < 100ms | Standard |
| Own dev team | Full-stack minimum 2 people | Any |
| Project budget | $9,200+ (40,000 PLN+) | $920-4,600 (4,000-20,000 PLN) |
| MVP timeline | 3-6 months | 4-8 weeks |
If your store meets 2-3 criteria on the left: time to talk about headless. If it meets 0-1: monolith.
More about e-commerce options and integrations with BaseLinker regardless of architecture.
Cost of migrating to headless: realistic calculation
Cost of building headless store from scratch:
| Scope | Cost |
|---|---|
| Architecture design + stack selection | $1,150-2,300 (5,000-10,000 PLN) |
| Headless frontend (Next.js/Hydrogen) | $3,450-6,900 (15,000-30,000 PLN) |
| Backend API + integrations | $1,840-4,600 (8,000-20,000 PLN) |
| Headless CMS (Sanity/Contentful setup) | $690-1,840 (3,000-8,000 PLN) |
| Testing, deployment, documentation | $1,150-2,760 (5,000-12,000 PLN) |
| Total | $8,280-18,400 (36,000-80,000 PLN) |
Monthly maintenance cost (after build):
Shopify monolith: $115-345/month (500-1,500 PLN/month) for hosting, maintenance, minor changes. Headless: $460-1,380/month (2,000-6,000 PLN/month) for two systems, two deployments, more configuration with changes.
Cost of migrating existing store:
Migrating product data, orders, customers is an additional $1,150-3,450 (5,000-15,000 PLN). Plus risk of store downtime if migration isn't carefully planned.
Cost summary: headless pays off with store revenue above $115,000/year (500,000 PLN/year), where the difference in performance (conversion) or flexibility (faster time to implement changes) justifies higher costs.
Where to start if you decide on headless?
Concrete path if your store meets the criteria:
Step 1: Audit current store
Before changing architecture: measure current Core Web Vitals (Google PageSpeed Insights, GSC). Measure time to implement typical changes (how long does it take to add a new homepage section?). Measure pain (what specifically doesn't work and how much does it cost monthly).
Step 2: Proof of Concept
Don't migrate the entire store. Build a headless frontend for one section (campaign landing page or product category). Measure performance. Compare with monolith.
Step 3: Gradual migration
Strangle pattern: headless frontend replaces monolith page by page. Backend stays. At any point, you can stop migration without disrupting store operations.
If you want to assess whether headless makes sense for your store or need a quote, we'll help conduct an audit. Service details in our services. We don't sell headless to everyone, because in most cases it's not the right answer.