"Should we use Shopify or build custom?" is the most common e-commerce question we get. The answer is almost always Shopify — until it isn’t. This post is the framework we use to draw that line in 2026.

TL;DR

Shopify (or Shopify Plus) is the right choice for almost every e-commerce business under $5M GMV/year. Custom platforms (Medusa, Saleor, headless on Next.js + Stripe + Sanity) make sense above that threshold and only when the standard Shopify checkout/catalog model doesn’t fit your business. Building custom because "we want flexibility" without a specific business reason is the most expensive mistake in e-commerce.

Why Shopify wins for most stores

Shopify in 2026 is mature platform infrastructure: PCI-compliant checkout, fraud prevention, app ecosystem with 8,000+ integrations, multi-currency, multi-language, tax calculation, 99.99% uptime. None of this is interesting to build but all of it is required to operate. Building it from scratch costs $200k–$500k of engineering plus ongoing maintenance — for a feature set that converts 99% of e-commerce flows.

Shopify is the right call when

  • Annual GMV is under $5M (Shopify) or under $20M (Shopify Plus)
  • Your business model is standard product → cart → checkout — not subscription-with-complex-tiers, not B2B with negotiated pricing, not custom-quote-then-fulfill
  • You want to focus engineering on differentiation (product, brand, marketing automation) — not on rebuilding checkout
  • You need to launch in <3 months — Shopify lets you go from zero to live store in 2 weeks
  • Tax, fulfilment, payment compliance is something you’d rather pay for than maintain

Where Shopify breaks down

Shopify is opinionated. Its checkout flow, product model, and customer object are designed for a typical retail e-commerce store. When your business doesn’t fit that model, you fight the platform.

Custom is worth considering when

  • Annual GMV is over $5M and Shopify’s 0.5–2% transaction fees become real money
  • Your checkout needs custom logic Shopify Plus can’t accommodate (B2B punchout, multi-step quotes, complex bundling)
  • Subscription billing is more complex than Shopify Subscriptions supports — multi-product subscriptions with mid-cycle changes, pause/resume with custom rules, hybrid one-time-plus-subscription cart
  • You need deep ERP/PIM integration that Shopify’s API rate limits make painful (B2B catalog with 100k+ SKUs and frequent inventory updates)
  • Your pricing model is genuinely unusual — usage-based, configurable products, dynamic pricing rules

The middle ground: headless Shopify

Before going fully custom, consider headless Shopify (Shopify Hydrogen or custom frontend on Storefront API). You keep Shopify’s checkout, payment, fraud, and admin while replacing the storefront with a custom Next.js or Remix app. This is the right answer for many "we’ve outgrown the theme but not the platform" cases.

Headless Shopify gets you 80% of the custom-stack flexibility for 20% of the build cost. The remaining 20% is the parts where Shopify’s checkout flow is non-negotiable — which is also the part where rebuilding it is the most expensive and risky.

Going fully custom: when and how

When you do need custom, the modern stack in 2026 is: Medusa.js or Saleor for the e-commerce backend, Next.js or Astro for the storefront, Stripe for payments, Sanity or Strapi for product content management, Algolia or Meilisearch for search. This stack has matured significantly — Medusa 2.0 in particular has caught up with most Shopify Plus features for B2C/B2B common cases.

Build cost realistic range: $80k–$200k for a custom commerce platform with the standard feature set. Maintenance: 1 engineer ongoing minimum for security patches, integration updates, and platform evolution. Compare that to Shopify Plus at $2k/month for the platform — the math is clear unless your scale or model genuinely demands custom.

What we ship at Schedars

Most of our e-commerce work is on Shopify (theme customization, custom apps, Plus migrations). Three projects in the last 18 months were custom builds: one B2B SaaS-meets-commerce hybrid (custom on Next.js + Stripe Connect), one subscription box company with complex pause/skip logic (custom on Medusa), one premium-brand storefront where the existing Shopify Plus build couldn’t hit the perf budget the brand wanted (headless storefront on Hydrogen).

In every custom case, the client had a specific business reason — not "we want flexibility". That’s the test we apply: is there a feature you need that Shopify Plus + apps cannot deliver? If yes, custom or headless. If no, theme + apps + careful integration architecture.

Bottom line

Custom e-commerce is rarely the right answer in 2026 — Shopify and Shopify Plus have absorbed enough flexibility that "we have a unique business" usually means "we haven’t looked at the right Shopify app". Go custom only when you can name the specific feature gap and quantify the GMV impact of closing it.

Weighing Shopify vs custom for your specific case? Tell us the business model and scale — we’ll tell you what we’d build and why.