Custom storefront or Shopify in 2026? The honest maths

The question usually arrives as a technology question: “Should we use Shopify or build custom?”. The answer is an ownership question: who carries the cost of change, and how often do you expect to change?

We’ve shipped all three shapes: Regalia as a custom Next.js storefront, Shopsera as a PHP/MySQL commerce platform, and Ecoumbrix on Shopify for a Canadian sustainability brand. Different answers, same reason: the buying model and the team’s capacity decided it. Not fashion.

What each one actually is

Shopify is a product with a strong default: catalogue, checkout, payments, admin, apps, hosting, and upgrades. A custom storefront is a capability: you own the code, the data model, the buying flow, and the maintenance.

Where Shopify earns its fee

• You can launch before hiring an engineer.
• Checkout, payments, tax, and fraud handling are mature defaults.
• Inventory and order management work on day one.
• The app ecosystem covers reviews, bundles, subscriptions, email, and support.
• Upgrades and security patches are someone else’s job.
• Merchants can run the store without waiting for a deploy.

For a brand validating a category, that’s usually rational. You can learn from real orders instead of debating a checkout flow in a Figma file.

Where the fee stops feeling small

• App creep: reviews, loyalty, bundles, subscriptions, analytics, returns, and popups each add a monthly line item.
• Third-party payment gateway charges may apply on some plans.
• Checkout customisation is limited outside higher-tier plans and checkout extensions.
• Editorial content and unusual product models fight the theme.
• Twelve apps can turn a fast theme into a slow store.
• Migration gets harder as customer accounts, subscriptions, and gift cards accumulate.

What custom really costs in India

These are planning ranges, not quotations. The bigger number is usually not the build. It’s annual ownership: 15–25% of the build cost for maintenance, hosting, security, support, and small changes. If you can’t fund that, custom isn’t cheaper. It’s deferred.

The 80% rule

If Shopify covers at least 80% of your buying experience and admin workflow, start there. If it covers 60% and the missing 40% is where your margin lives, custom starts to make sense. If it covers 85% but the last 15% is one checkout tweak, don’t rebuild commerce for one screen.

Custom wins when...

• Your pricing depends on customer tier, volume, contract, or configuration.
• Products are made-to-order, configurable, or priced by a rule engine.
• You need multi-warehouse, drop-ship, or ERP logic inside the buying flow.
• The brand experience is genuinely editorial and a theme keeps sanding off the edges.
• You need deep integrations with a custom CRM, LMS, booking system, or operations platform.
• Data ownership, procurement rules, or compliance require it.
• You have an internal owner for year two — not just a launch date.

Shopify wins when...

• The catalogue and checkout are standard.
• You’re launching a new category and need real orders quickly.
• The team is small and paid media is the main growth channel.
• You don’t have engineering capacity to maintain a storefront.
• The budget is under roughly ₹5 lakh and you’d rather spend the rest on inventory.
• The product roadmap is mostly merchandising, not software.

The middle path is usually the honest answer

The most common sensible setup isn’t binary. Use Shopify for commerce and a custom Next.js front end for content and merchandising. Or keep hosted checkout and build the configurator your category needs. Or build the admin your operations team actually uses and keep a mature checkout.

Headless is not automatically faster. It adds a second deploy, a second cache, and a second team. Only do it if the experience or content model needs it.

Migration is the hidden project

A redesign is not a migration. Budget for both. The parts people forget:

• 301 redirects for every product, collection, and journal URL.
• Product structured data and canonical tags.
• Analytics events, pixels, and conversion tracking.
• Customer accounts, wishlists, and order history.
• Reviews, gift cards, discounts, and subscriptions.
• Email flows, abandoned carts, and transactional templates.
• A freeze window and a rollback plan.

A 30-day decision sprint

1. Map the real order flow, including returns, exchanges, and failed payments.
2. Price three years of Shopify, apps, and payment fees against a custom build plus maintenance.
3. List every integration and who owns it.
4. Prototype the one page or flow that the standard platform can’t do.
5. Run a small migration dry run with redirects and analytics.

The honest answer

Start on Shopify unless you can name the specific limitation that is costing you money and the person who will own the custom system in year two. “We want it to feel more premium” is a design brief, not a platform decision. A good theme, better photography, and fewer apps can solve a lot of premium problems.

And if you do go custom, don’t build a storefront because you dislike the admin theme. Build it because the buying model needs it. The admin can wait.