RADOMIR BOŠKOVIĆ
GET IN TOUCH
SHOPIFY HYDROGEN

Shopify, without designing around a theme.

You keep Shopify’s commerce engine — products, customers, orders, payments, checkout. What changes is the storefront in front of it. Instead of fitting your business into a prebuilt theme, the storefront gets designed around your products, your customers and the way you actually sell.

Shopify Hydrogen developer and eCommerce consultant. Ten years running eCommerce for brands, now building headless Shopify storefronts end to end.

TALK ABOUT YOUR STORE Is this right for your store?

SAME ENGINE, DIFFERENT FRONT

Shopify still runs the commerce. Hydrogen runs the experience.

A normal Shopify store puts a theme in front of the commerce engine. The theme decides what a product page can hold, where things are allowed to go, and how far you can push it before you are fighting it.

Headless Shopify separates the two. Shopify keeps doing what it is good at: catalogue, inventory, customers, orders, payments, checkout. The storefront becomes an application of its own, built with Shopify Hydrogen, talking to Shopify over its APIs.

Nothing moves off Shopify. You log into the same admin. Orders arrive the same way. Your accountant sees no difference.

Standard Shopify

ShopifyProducts · Orders · Checkout
ThemeOne fixed template
Customer

The theme decides what is possible.

Shopify Hydrogen

ShopifyProducts · Orders · Checkout
APIs
Your storefrontBuilt for your catalogue
Customer

You decide what is possible.

Same Shopify engine. Much more control over what sits in front of it.

WHAT CONTROL BUYS

A custom Shopify storefront is only worth it if it changes what customers can do.

These are not features from a brochure. Each one is a decision from a store I have built, and each one was possible because the front end was ours to design.

  • The buy action stays on screen

    On mobile the add-to-cart goes sticky on the product page, so the decision never scrolls away while somebody reads the details underneath it.

    Toyland case study

  • Personalisation sits with the purchase

    Sizes, engraving and the purchase action stay together, above the material story. The choice the customer has to make is not separated from the choice they came to make.

    Pansuara case study

  • Ways in that are not collections

    A large LEGO catalogue is shopped by theme, age, price and piece count. Those became real paths: a theme directory, shop-by-age and shop-by-price pages, and facets on every collection.

    Klosslabbet case study

  • A second audience gets its own proposition

    Teak Parcels sells to investor prospects, not to shoppers. It has its own landing experience and its own lead funnel inside the same store, instead of being squeezed into a collection page.

    Pansuara case study

  • The account does more than list orders

    The customer account carries a referral mechanic that pays out real store credit rather than discount codes, with the credit and expiry emails running as Klaviyo flows.

    Pansuara case study

  • One taxonomy, reused everywhere

    Departments, themes and characters, age and price bands: one structure behind the mega menu, the category facets and predictive search, so discovery behaves the same way wherever it starts.

    Toyland case study

THE SHOPIFY APIS

Three doors into Shopify, described as capabilities.

Most of a headless build is Shopify API integrations. You never have to touch this detail yourself — but it helps to know what each door lets you decide.

Shopify commerce core
  • Products
  • Customers
  • Orders
  • Checkout

Storefront API

Everything a shopper browses and buys.

  • Products
  • Collections
  • Pricing
  • Cart
  • Checkout

Customer Account API

Everything that happens after the order.

  • Orders
  • Returns
  • Store credit
  • Referrals

Admin API

Everything the business runs on behind the scenes.

  • Catalogue
  • Inventory
  • Automation
  • Metaobjects
Your storefront experienceOne interface · Three sources of truth
Three doors into the same commerce data. Each one opens a different kind of experience.

Storefront API

Everything a customer needs in order to shop

Products, collections, prices, availability, cart, checkout and search. Shopify holds the truth. You decide how customers meet it: how a product page is laid out, how filtering behaves, what a customer sees when a size is gone, how a bundle is presented.

  • Catalogue, pricing and variants
  • Cart through to Shopify checkout
  • Search, filtering and sorting
  • Availability and stock states

Customer Account API

What happens after the purchase

The account is not just a list of orders. Orders, returns, store credit, referrals, registrations and saved products can all live in one place a customer actually wants to open. Retention gets designed here rather than bolted on afterwards.

  • Orders, returns and reorders
  • Store credit and referrals
  • Saved products and lists
  • Registrations and warranties

Admin API

Making Shopify fit the rest of the business

The back-office door. Catalogue imports, inventory sync, workflows that fire when something happens, custom fields through metafields and metaobjects, and connections to the systems you already run.

  • Catalogue imports and inventory sync
  • Metafields and metaobjects
  • Event-driven workflows
  • ERP, CRM, PIM and 3PL connections

DAY TO DAY

Custom does not mean hard to manage.

“Will I need a developer every time I want to change a banner?”

No. What your team can edit is a decision made at the start of the build, and it is one of the first things I ask about.

Shopify already holds structured content through metafields and metaobjects. Homepage modules, size guides, FAQs, buying guides, promotional blocks and navigation can all be built from those, which means they are edited in the Shopify admin your team already uses.

If a content team wants proper drafts, scheduling and previews, a headless CMS goes in alongside Shopify. That is a preference, not a requirement — and it is worth its cost only when somebody is going to use it.

  • Homepage modules

    Banners, rows, featured collections, seasonal blocks.

  • Guides and support

    Size guides, buying guides, FAQs, delivery and returns copy.

  • Merchandising

    Navigation, promotional blocks, campaign pages, badges.

  • Product detail

    Materials, care, specifications and anything else held in metafields.

THE REST OF THE BUSINESS

Hydrogen is the experience layer over the systems you already run.

Going headless does not mean replacing your stack. The ERP, the PIM, the 3PL, the CRM, the email platform, the reviews and the analytics all stay where they are. What changes is that the storefront decides how their data is used, instead of accepting wherever an app widget happens to drop it.

The systems already running the business

  • ERP / PIMStock & products
  • 3PLWarehouse & shipping
  • CRMCustomer records
  • KlaviyoEmail & SMS
  • Judge.meReviews
  • GA4Analytics
  • CMSEditorial content
ShopifyCatalogue · Customers · Orders · Checkout
Hydrogen experience layerConnects every system above · Replaces none
CustomerThe only layer they ever see
Hydrogen does not replace the commerce ecosystem — it becomes the experience layer that connects it.

Reviews where they earn their keep

Review data from a provider like Judge.me is yours to place: the rating beside the price, the reviews that mention fit under the size guide, a filter for photographs only. Not one block at the bottom of the page because that is where the app installs itself.

Events into the flow that should run

A customer does something the business cares about — configures a product and leaves, redeems credit, joins a waitlist. The storefront captures that event with the context around it, sends it to Klaviyo, and the right flow runs. The email can be specific because the storefront knew something specific.

PERFORMANCE

Speed is an architecture decision, not a plugin.

I am not going to promise you a score. Real-store performance depends on your images, your catalogue, your third-party scripts and how much marketing needs to load on every page. What a custom storefront gives you is control over the three things that decide it.

  1. Where pages are rendered

    Rendering runs close to the customer rather than waiting on a round trip to a single origin. Pages that can be cached are cached; pages that have to be personal stay personal.

  2. What code reaches the customer

    A theme ships whatever the theme and its apps ship. Here, every script on the page is there because somebody decided it should be — and it can be removed when it stops paying for itself.

  3. What loads, and when

    The price, the first image and the buy box are what a customer needs immediately. Reviews, recommendations and tracking can wait. That order is something you set, not something you hope for.

Anyone quoting you a guaranteed score before they have seen your catalogue and your marketing stack is guessing.

HONESTLY

A theme is often the right answer.

A theme is probably enough when…

  • Your catalogue is small and shopped by browsing rather than filtering.
  • A standard product page covers everything you sell.
  • You are early, and getting to market cheaply matters more than anything else.
  • Traffic and revenue do not yet justify the cost of a custom build.
  • Your team is small, and a familiar admin matters more than control.

Hydrogen earns its cost when…

  • Discovery is the problem: a big or complex catalogue with several honest ways in.
  • The product page has to do something a theme cannot — configuration, personalisation, bundles, B2B pricing.
  • You serve more than one audience, or more than one market, from one store.
  • Retention matters: accounts, store credit, referrals, subscriptions, repeat ordering.
  • Shopify has to stay in sync with an ERP, PIM, 3PL or CRM.
  • App sprawl is already slowing the store down and you keep building around it.

I will not recommend headless just because I can build it. If a theme does the job, that is what I will tell you. A Shopify Hydrogen migration is worth doing when there is a commercial reason for it, and not before.

HOW IT GETS BUILT

Commercial decisions first. Code last.

  1. 01

    Commercial decisions

    What you sell, to whom, at what margin, and what has to change for the store to earn more. Before anything is drawn.

  2. 02

    UX

    Navigation, discovery, the product page, the account — specified as states, including loading, empty, no match and error, not only the happy path.

  3. 03

    Architecture

    The data model, metafields and metaobjects, integrations, and what is cached versus what has to be personal.

  4. 04

    Build

    Hydrogen and TypeScript against your real Shopify catalogue, deployed on Oxygen or Cloudflare.

  5. 05

    QA

    Real products, real prices, real stock states, on real devices. Checkout tested end to end before anyone else sees it.

  6. 06

    Measure

    Analytics that answer commercial questions, then the next round of changes based on what they say.

PROOF

Three Hydrogen builds, described accurately.

Pansuara storefront homepagePRE-LAUNCHPansuaraA DTC teak brand. Storefront, customer account, referral mechanic and lifecycle flows are built and configured. No customer traffic has run through it yet, so there are no results to quote.CASE STUDY Klosslabbet live storefrontLIVE · HANDED OVERKlosslabbetA Swedish LEGO retailer, built and launched on my own, live since September 2025 and handed over that December with documentation. The client has operated it since.CASE STUDY Toyland storefront redesignIN DEVELOPMENTToylandA toy and baby retailer’s storefront redesign: the commerce experience designed end to end, then built in Hydrogen against the real Shopify catalogue. Not yet live.CASE STUDY

QUESTIONS I GET ASKED

The five that come up every time.

Is Hydrogen more expensive than a theme?

To build, yes. A theme is bought and configured; a custom Shopify storefront is designed and built. Running it is a different matter — hosting is modest, and stores often drop paid apps whose job the storefront now does properly. The question is whether the extra build cost buys enough revenue or enough saved time to be worth it. If it does not, say no.

Can my team still edit content without a developer?

Yes. Homepage modules, size guides, FAQs, buying guides, promotional blocks and navigation are built on Shopify metafields and metaobjects, so they are edited in the Shopify admin your team already knows. A headless CMS can be added if the content team wants drafts, scheduling and previews.

Do I keep the Shopify checkout?

Yes. Checkout stays Shopify’s, with your payment providers, your tax and shipping configuration, and Shop Pay if you use it. Headless changes the storefront in front of checkout, not the checkout itself.

What happens to my apps?

It depends on the app. Anything that works through Shopify’s APIs — reviews, email, subscriptions, loyalty, search — generally carries over, and usually fits better because you decide where its data appears. Apps that work by injecting a widget into a theme do not, so we go through your app list early and decide which ones stay, which get replaced and which were never earning their keep.

Can you migrate an existing store to Hydrogen?

Yes, and it is less dramatic than it sounds. Products, customers, orders and checkout stay exactly where they are in Shopify; the storefront in front of them is what gets rebuilt. A Shopify Hydrogen migration is usually staged, with URL structure, redirects and SEO planned before anything is switched over.

IS IT RIGHT FOR YOU?

Signals that a custom storefront would pay for itself

  • You have more products than your navigation can honestly handle.
  • Customers keep asking for a way to browse that your theme cannot give them.
  • The product page needs to do something it was never built to do.
  • You pay for apps that each add weight, and none of them quite fits.
  • Your catalogue or stock lives in another system and keeping it in sync is manual.
  • You sell to two different audiences, and one storefront is serving neither well.
  • The traffic is there and the conversion rate is not.

If two or three of those sound like your store, it is worth a conversation. If none of them do, keep the theme — and I will tell you so.

TALK ABOUT YOUR STORE