When (Not) to Leave Shopify.

Most brands shouldn't leave Shopify. That's a strange thing to write on an agency site that does both — but it's the honest place to start. Most of the urge to leave Shopify, in our experience, isn't actually a Shopify problem. It's a craft problem, or a copy problem, or a brand problem, or a budget problem dressed up as a platform problem. The replatform is the most expensive version of the answer to a question that's usually different.

We do leave Shopify when it's right. We left Shopify for Pegoretti in 2026, and we'll get to why below. But three out of four brands that arrive talking about replatforming are, in our reading, better served by a properly considered rebuild on the platform they already have. This piece is about telling the two apart.

If you want the longer version of the same argument applied to headless commerce specifically, we wrote that one too — see Headless Commerce, Honestly elsewhere on this site. The Shopify question and the headless question are different conversations, and we keep ending up in both.

Where our Shopify view comes from

Pegoretti — the case study below — wasn't on Shopify. It was on Cargo.site, a different templated platform with its own constraints. Our Shopify view comes from elsewhere: three years as consultant e-commerce directors for Above Category in Sausalito. That meant a lot of hours inside the platform, inside the theme code, inside the app marketplace, and inside the bill at the end of each month.

Three things stood out then, and they show up in almost every "should we leave Shopify" conversation we've had since.

Template suffocation. Shopify themes are flexible until you want to try something genuinely outside the box they were built around. Then you're either fighting the theme, replacing the theme, or living with the version of the idea you didn't want. Most of the time a custom theme solves this — and we'll come back to that. The point at which it doesn't, is the point worth examining.

App marketplace lock-in. Shopify's app marketplace is one of the platform's real strengths and one of its more expensive traps. You start with an email app, a reviews app, a loyalty app, a subscription engine, a shipping app, a tax engine, a search app. Each is £30 to £150 a month. They integrate with each other through more apps. The bill stacks quietly. We've seen mature stores paying four-figure monthly subscription totals before anyone's totted up the receipts.

The subscription tax you don't see until you do. Those fees aren't a one-time number; they're a permanent tax on staying on Shopify the way you've built it. A custom build replaces a lot of them with things you build once and own outright. The day-one cost is higher; the eighteen-month maths sometimes surprises brands either way.

None of which means leaving. It means knowing what you're actually paying, and for what.

What Shopify gets blamed for (and shouldn't)

Three accusations come up almost every time.

"Our site looks templated." Almost always a craft problem rather than a platform problem. We've worked on Shopify sites that look like nothing else on the internet; we've seen custom builds that look like everything else. The templated feeling comes from default theme choices, default photography, default copy, default everything-stacked-vertically structure. All fixable on Shopify, without leaving Shopify. The brand whose look is keeping you up at night could probably be solved with a custom theme, a fresh design pass, and a copy rewrite. Not a five-figure replatform.

"We've outgrown it." Sometimes true. Often hyperbole. The real question is what specifically you've outgrown. If the answer is "the URL structure feels limiting" or "the checkout looks like Shopify's checkout" — that's a configuration question, not a platform question. If the answer is "we genuinely cannot model the shape of our business inside Shopify's data structure" — different conversation, and one we should have. Most teams haven't sat with the question long enough to know which they are.

"We need something more custom." Custom usually means a custom theme on Shopify, which is a perfectly valid path that gets you most of what the custom-build pitch promises at a fraction of the cost. If you need a Shopify theme that doesn't look like anyone else's, you can have one. The platform isn't the constraint there. The brief is.

The honest reasons to actually leave

Three reasons make us nod when we hear them.

Your business doesn't have a checkout. If your sale isn't an add-to-cart-and-pay flow — if it's a conversation, a commission, a bespoke quote, a private viewing — Shopify's checkout architecture is the wrong shape for the actual transaction. You can hack around it for a while. Eventually the hacks cost more than the rebuild would have.

You share one inventory across genuinely distinct channels. Web shop, wholesale portal, in-store POS, mobile app, embedded checkout on third-party publisher sites — all pulling from one source of stock truth, with different rules, different prices, different presentation. If that's the actual shape of the business, a commerce engine in the middle (with various fronts pulling from it) is the architecture that fits. Shopify can be that engine; it can also not be, depending on scale and specifics.

Strategic ownership. Rarer than the first two, but real. A brand that doesn't want a load-bearing part of its business sitting inside another company's roadmap and pricing changes. A brand where the e-commerce IS the product. A brand whose investors or board have specific views on platform risk. These are usually larger businesses with specific reasons. If you're doing under a million a year in revenue and you're asking this question, the answer is almost certainly: the question is too early to ask.

What it actually costs when you do leave

The replatform isn't just the build fee. The headless piece covered the full picture in detail; the headlines:

A real custom build, properly done, is mid-to-high five figures at the considered end, six figures at the larger end, and several months either way. AI development tools (Claude Code, Cursor, the rest of the current toolkit) compress experienced-engineer time meaningfully and the running cost is genuinely lower than it was a year and a half ago — but they don't change the order of magnitude. They make a £150k project a £100k project, not a £20k one.

Then there's the ongoing engineering. A maintained custom store needs a developer for the rest of its life. Adding a homepage banner becomes a code change. Launching a holiday campaign means commits and deploys. Apps you currently install in five minutes become custom-built integrations.

And every Shopify app you use today disappears the day you leave. The email platform, the shipping integration, the subscription engine, the loyalty programme, the tax engine, the search, the reviews — every single one gets rebuilt, reintegrated, retested. This is the part brands underestimate hardest in the buy decision, and the part that drags the longest in the actual build.

The value you already have

It's easy to miss the value you've already got while chasing the shiny new thing.

Our e-commerce experience didn't begin with WBD. It began with Always Riding — a D2C cycling boutique we founded in the UK in 2008, sold in 2018, and the crucible in which we learned everything. It was a destination, the place where brands now famous in the cycling world (PAS Normal Studios, Café du Cycliste, Pedla, ORNOT, among others) first found their UK audience. And for a long while, Always Riding ranked number one on Google for "cycling jerseys" — a high-volume, high-intent term that drove a steady proportion of new-customer business each month.

We knew the site was well-optimised; we'd spent years on the long-form copy, the meta descriptions, the meta titles (more important then), the internal linking between pages. We assumed redirects would carry the SEO equity across when we migrated platforms.

They didn't. Despite a careful move with redirect URLs in place, we lost most of that ranking. The new site replicated the pages but there was an essence the old site had that the new one didn't — obvious to Google, even when it was hard for us to put a finger on at the time. Eventually we got the position back. The interim wasn't fun.

Looking back, the move was probably still the right call technically. But we'd have been a lot more careful about identifying what was actually working before we touched it. Some of what was on the page wasn't just content — it was years of compounding signal that doesn't show up in a platform export.

That's the cost most brands don't see until it's gone: the slow-built SEO equity that's been doing free work for you for years. A redirect is a redirect; the thing Google has actually been ranking might not survive the move, no matter how careful you are. Bear that in mind when you're weighing the replatform fee against the status quo — there's a line on the cost side that nobody puts in the quote.

What we'd say no to

A brand comes to us and says: we've outgrown Shopify, we need a custom build. We ask: what specifically isn't working?

If the answer is the site looks like everyone else's — we'd say no to the rebuild. The site looks like everyone else's because the design and the copy haven't been thought about properly. Spend a third of the replatform budget on a real design pass, a real copy rewrite, and a custom theme, on the Shopify you already have. Most brands have never had this kind of work done. The result is closer to what you wanted than the rebuild would be.

If the answer is our conversion is dropping — we'd say no. That's a CRO question, often a homepage-trust question, sometimes a brand-misalignment question. Almost never a platform question.

If the answer is our agency told us we need to go custom because we're at scale now — we'd ask which agency, and what they're proposing to charge for the migration. The migration is sometimes the fee. The urgency, in those conversations, can be invented.

What good looks like, when you stay

The Shopify build we'd recommend to a brand that doesn't actually need to leave looks like this. A custom theme, designed from scratch around the actual brand. A considered structure — homepage, product page, collection, account, post-purchase — each rethought in the brand's voice rather than the theme defaults. Copy that sounds like the brand and not like a template. Photography that's the brand's, not stock. A typography system. An actual point of view, expressed in pixels.

Most brands have never had a build like this on Shopify. They've had a theme installed, brand colours dropped in, the logo swapped, and shipped. Every brand whose Shopify site reads as templated feels like they've outgrown Shopify, because they've never seen Shopify fit their brand properly. The fit, not the platform, is the gap.

Pegoretti — when a templated platform isn't right

Pegoretti was on Cargo.site — a curated, template-based platform for creative studios. We replatformed them onto a fully custom Next.js + Sanity build in 2026. Same principle as leaving Shopify, applied to a different templated platform: the business shape didn't match the platform shape.

Pegoretti commissions a custom bicycle frame for each customer. There's no inventory in the conventional sense — every frame is unique, with bespoke lead times, paint configurations through the Color Wall, and a real conversation with the Bottega in Verona before any money changes hands. The transaction isn't a checkout. It's the start of a relationship.

We could have stayed on Cargo.site and built around it. We'd have spent another year hacking a portfolio platform into something it doesn't want to be. The same would have been true on Shopify — Pegoretti's business doesn't fit a checkout-shaped platform of any kind.

So we left. The new site runs Next.js on the frontend, deployed to Vercel. Sanity holds the editorial content — the journal, the Bottega story, the Color Wall finishes. A custom commerce flow, built in-house, handles the commission process. It's the shape Pegoretti's actual business is, rather than the shape any commerce platform wants every business to be.

The lesson transfers across platforms. If your business doesn't have a checkout, the platform built around a checkout will fight you forever. The same exit, the same logic, applies to a brand on Shopify in the same situation. But — and this is the point of the whole piece — Pegoretti is the exception, not the rule. Pegoretti's business doesn't have a checkout. Yours probably does. If yours does, the answer is almost certainly: get the Shopify build right.

Three questions before you spend the rebuild fee

If you're seriously considering leaving Shopify — or your agency is recommending it — three questions worth answering honestly first.

One. What specifically about Shopify is blocking us — name the actual limit, not the feeling? If you can't name it precisely, the answer isn't a replatform yet.

Two. Have we done a proper design and copy pass on the Shopify build we currently have? If no, do that first. It's a smaller cheque and you may find it solves what you thought was a platform problem.

Three. What's the total cost of ownership over eighteen months, including the engineering retainer the new platform will need? The honest answer is usually two to three times what brands expect when they first ask.

One last thing

We're platform-agnostic — we build on Shopify, WooCommerce, Magento, and headless stacks like Pegoretti's. We do more Shopify than the rest combined, and we're good at it. We're also happy to tell a brand it should stay where it is, if that's the honest answer.

If you're weighing a rebuild and want a real second opinion before you sign — written by a person, no AI, no upsell — there's a sketch flow on this site at /sketch for the toe-in-the-water version, or a full Discovery at /discovery if you'd rather dive in. If our answer to your three questions is "Shopify with a redesign," that's what we'll tell you. We'd rather you make a good decision than a fast one.