How Much Does It Cost to Build an E-Commerce Web App?

The gap between €5,500 and €20,000 is mostly checkout and catalog logic, not page count. A flat catalog with simple products, one currency, and standard Stripe Checkout sits near the floor. The price climbs when products carry variants, bundles, or subscription billing, and when checkout has to calculate region-specific tax and shipping at run time.

The quietest cost driver is order operations — the admin side buyers rarely demo before signing off. Returns, partial refunds, fulfillment status, and stock decrements on sale are real workflows with edge cases, and they account for a large slice of the upper range. A rough split:

• Storefront + cart + Stripe Checkout — entry-level scope, near the low end
• Variants, bundles, or subscriptions — each is a distinct chunk of work
• Tax/shipping rules + multi-warehouse stock — pushes toward the high end
• Returns, partial refunds, fulfillment states — the part most quotes underestimate

To land near €5,500–8,000, keep the first build narrow: one currency, one region's tax rule, standard variants, Stripe Checkout instead of a custom multi-step form, and a lean admin that covers products plus order status. Ship that, sell through it, then add scope against real data.

What pushes a build toward €20,000 is usually decided before any code is written: subscription or custom-pricing logic, multi-warehouse inventory, heavy returns/refunds workflows, and a checkout that reconfigures by region. None of those are wrong to want — they just belong in a later phase once the core store is earning.

The single biggest saving is honest scope. A pizza-ordering flow I built for Pizzeria Bestek (React + Supabase) stayed cheap precisely because the ordering logic was custom but tightly bounded — no speculative features carried into v1. If a feature isn't tied to a real order or a real customer request, it's a phase-two line item, not a launch blocker.

The build quote isn't the whole bill. Recurring costs that don't appear in a fixed price:

• Payment fees — Stripe takes roughly 1.5–3.4% + a fixed fee per transaction. On real volume this dwarfs hosting.
• Hosting — a storefront on Vercel is low-cost or free at small scale, but a managed database, email sending, and image/CDN bandwidth add monthly cost as traffic grows.
• Third-party / SaaS fees — transactional email, tax calculation, search, and analytics are usually per-seat or usage-priced subscriptions.
• App-store cut — only relevant if you later add a native iOS/Android app (excluded here). Apple and Google take 15–30% of in-app purchases; web checkout avoids that entirely, which is one reason a web app is the cheaper starting point.
• Maintenance — dependency updates, payment-API changes, and bug fixes are ongoing. Budget for it as a small monthly retainer or a block of hours, not a one-time line.

For a first e-commerce build, fixed-scope is the right structure. The feature set is knowable up front — catalog, cart, Stripe checkout, admin — so the risk of an open-ended hourly meter falls on no one. You know the number before work starts. Hourly fits better after launch, for iterative changes where the next feature depends on what real customers do.

Realistic timeline is 6–14 weeks. A focused storefront with standard checkout lands around the low end; subscriptions, multi-warehouse stock, and region-specific tax/checkout push toward 14. Working solo is leaner than an agency on a build this size — no hand-offs between separate design, frontend, and backend teams, and one person holding the full picture.

Proof that this scope ships solo: BookBed, a booking SaaS I built on Flutter + Firebase + Stripe with bidirectional iCal sync, and Callidus, a clinic SaaS on React + Firebase — both Stripe-backed products with the same catalog-checkout-admin shape an e-commerce app needs. Per the build note, if your needs are under ~500 SKUs with standard checkout, a well-configured Shopify often beats a custom build; the custom path pays off when the logic is non-standard, and that's exactly what I scope precisely.