Shopify alternative: own your e-commerce stack

When Shopify stops being cheap

You started on Shopify because it was the obvious answer. Quick to launch, thousands of apps, good documentation. For the first year or two, the fees felt like a fair trade for not having to think about infrastructure.

Then revenue climbed.

At £100k/yr, a 2% transaction surcharge is £2,000. Annoying but manageable. At £500k, it's £10,000. At £1M, that single line item — the additional transaction fee Shopify charges when you're not on their own payment processor — is **£20,000 a year**. That's before your app subscriptions, your plan fee, or the 2.0% + 25p Shopify Payments takes on every card it does process.

This is the point where a growing number of UK merchants start searching for a Shopify alternative. Not because Shopify is bad — it isn't — but because the economics of renting a platform shift dramatically as you scale.

This post is for the founder whose monthly fee statement just made them wince. We'll show you the real numbers, explain what the headless alternative actually involves, and be straight about who should make the switch and who shouldn't.

Breaking down the fee stack

The transaction fee trap

Shopify charges an additional fee on every order processed through a third-party gateway. On the Basic plan, that's **2%**. On the Grow plan it's 1%, and on Advanced it's 0.5%. The only way to eliminate it entirely (on standard plans) is to use Shopify Payments — their own processor, powered by Stripe.

If you use Shopify Payments on the Basic plan, you pay **2.0% + 25p** per transaction instead of the transaction fee plus your gateway's fee. On £1M GMV, that's roughly £20,000 in processing costs plus the fixed 25p per order. For a store doing 20,000 orders a year at an average £50 order, that's 20,000 × 25p = £5,000 extra on top of the 2%.

So even the "free" route isn't free — it's just a different fee structure. You're trading the 2% surcharge for 2.0% + 25p per order directly to Shopify/Stripe.

App creep: the fee you didn't budget for

Shopify's core product is a storefront and checkout. Most things beyond that — reviews, loyalty programmes, advanced search, subscriptions, returns management, B2B pricing — require apps. Each one is £30–£150/month.

A typical established Shopify store runs six to 12 apps. At an average of £80/mo per app, that's £960–£1,150/month before you've paid for the platform itself. Over a year: **£11,500–£13,800 in app subscriptions** for features that, in an owned system, would be built into your stack from day one.

Checkout restrictions below Plus

Until you're on Shopify Plus (from £1,800/mo), you cannot customise the checkout. You can't add custom fields, split shipping logic, or run a multi-brand checkout. You can't add your own logic to the order confirmation page. The checkout belongs to Shopify.

On Plus, you get Checkout Extensibility — but you're now paying at minimum £21,600/year for the privilege of modifying the page your customers see when they hand over money.

Page speed: themes vs. server components

Shopify themes run on Liquid, a templating language that pre-renders HTML on Shopify's CDN. That's fast — for a simple store. But every app you install can inject additional JavaScript, additional requests, and additional render-blocking code into your storefront.

A Shopify store with eight apps, a product upsell widget, a chat widget, and a loyalty pop-up will score somewhere in the 40–65 range on PageSpeed Insights. Not terrible. But not competitive with a Next.js storefront using React Server Components, which shifts expensive rendering work to the server and ships minimal JavaScript to the browser.

Properly architected Next.js storefronts consistently score **90+ on PSI** for both mobile and desktop. That matters for conversion rates and — increasingly — for how AI search systems rank product pages.

What owning your storefront looks like

A STACKD Owned Storefront replaces the Shopify rental model with a system you own outright. The stack:

- **Next.js + Payload CMS** — your storefront and content management. Payload is open-source, Next.js-native, and runs inside your own Next.js app. No separate login, no extra service. - **Frappe / ERPNext** — your inventory, purchase orders, fulfilment logic, and supplier management. Open-source, self-hosted or cloud-hosted. One Frappe instance can serve multiple brand storefronts. - **Stripe** — your payment processor. You have a direct Stripe account, no intermediary. You pay Stripe's rates (1.5% + 20p for UK cards) and nothing else. - **n8n** — your automation layer. Order webhooks to Frappe, stock-level sync, email/SMS sequences, returns workflows. Built once, runs indefinitely.

The one-time cost

Building this system costs **£22,000–£32,000** as a one-time project. That's not a guess — it's the range across the Owned Storefront builds we scope. Larger multi-brand or multi-market configurations sit at the top of that range.

After that: roughly **£30/month** in hosting (Vercel + a small VPS for Frappe + n8n), plus Stripe's card processing fees, plus your own time to manage content in Payload CMS.

When does the break-even happen?

On £1M GMV/yr with Shopify Basic and Stripe (non-Shopify Payments), you're looking at roughly **£38,000+ per year** in combined fees and subscriptions. The Owned Storefront costs £22k–£32k to build and £360/yr to host. Add Stripe fees (same as you'd pay anywhere): approximately £15,000/yr.

Total year one cost of ownership: roughly **£37,000–£47,000** (build + hosting + Stripe).

Year two and every year after: roughly **£15,360** (hosting + Stripe only).

Against Shopify's £38,000+/yr, the system pays for itself between 12 and 24 months depending on your build cost and GMV. At £500k GMV the numbers still work — your Shopify fee stack at that scale is typically £20,000–£25,000/yr. At £250k GMV, the case is weaker and you should probably wait.

Be honest about the trade-offs

Shopify wins in several ways that matter, and we'd be doing you a disservice to ignore them.

**Ecosystem**: 8,000+ apps, thousands of Shopify developers, deep integrations with Meta, Google, TikTok Shop. If you need a new selling channel tomorrow, it's three clicks on Shopify.

**Speed to market**: A developer who knows Shopify can have a functioning store live in days. The Owned Storefront path takes weeks to scope and months to build.

**Support**: Shopify has 24/7 chat support. An owned system means you're responsible for your own uptime — or you pay someone like us to be on call.

**Beginners**: If you're under £100k GMV and still finding product-market fit, Shopify is the correct choice. The overhead of owning infrastructure before you know your store will work is foolish.

The STACKD path is for operators who have product-market fit, are past the point where the fee maths hurt, and want to stop sending a percentage of every sale to a platform indefinitely. It's for the founder who's thought about it, run the numbers, and decided that ownership makes more sense than perpetual rent.

What to look for in a Shopify alternative

If you're evaluating options beyond STACKD, here's what actually matters:

**No platform transaction fee** — BigCommerce charges none on any plan. WooCommerce (self-hosted WordPress) charges none. ShopWired (UK-based) charges none. The moment a platform takes a percentage of every sale, the maths changes.

**Checkout control** — Can you modify the checkout without a £2,000/mo enterprise plan? If not, it's a constraint you'll hit.

**Data portability** — Can you export your full product catalogue, customer records, and order history in a standard format? If the platform makes that hard, you don't own your data.

**Headless option** — Does the platform support a headless architecture where your Next.js frontend talks to an API? Shopify does (via Storefront API and Hydrogen), but it still takes the transaction fee unless you're on Plus. An open-source headless CMS like Payload removes the platform dependency entirely.

**Inventory management** — Most SaaS storefronts have weak inventory and no supplier management. If your operation is complex — multiple warehouses, purchase orders, drop-ship — you need an ERP layer. Frappe/ERPNext is the open-source answer here.

Frequently asked questions

**Is Shopify worth it for UK merchants?**

For merchants under £200k/yr GMV: yes, almost certainly. The platform, apps, and support are hard to beat at that scale. For merchants at £500k+ GMV, the fee maths becomes painful enough that alternatives deserve serious consideration. The answer depends entirely on your current spend and whether you're willing to invest in a build.

**What is the cheapest Shopify alternative for UK businesses?**

WooCommerce (self-hosted) has no platform fee and no transaction fee, but you bear all hosting, security, and maintenance costs. BigCommerce has strong built-in features and no transaction fee. Both are cheaper in recurring costs but require developer investment to maintain. A custom-built Next.js storefront is more expensive upfront but cheapest over a five-year horizon at meaningful GMV.

**Does Shopify charge transaction fees if I use Stripe?**

Yes. If you connect Stripe directly as a payment gateway on a standard Shopify plan, Shopify charges an additional transaction fee of 2% (Basic), 1% (Grow), or 0.5% (Advanced) on top of whatever Stripe charges. To avoid this, you must use Shopify Payments — which is powered by Stripe anyway — or upgrade to Shopify Plus.

**What is headless Shopify, and is it worth it?**

Headless Shopify decouples your frontend (what customers see) from Shopify's backend (product data, cart, checkout). Your Next.js app calls Shopify's Storefront API. You get full design freedom and better performance, but you still pay Shopify's fees and are still limited to their checkout on standard plans. For most merchants, it's expensive to build and doesn't solve the fee problem unless you're on Plus.

**How long does it take to migrate from Shopify to an owned stack?**

A full migration — products, customers, order history, SEO redirects, and new storefront build — typically takes 12–20 weeks. The build itself is eight to 14 weeks; data migration and QA is two to four weeks on top. It's not a weekend job, but it is a finite project rather than an indefinite subscription.