Rebuilding a store that sells nothing is a project. Rebuilding a store that sells every day is an operation. The difference is that between the moment you start and the moment you publish, orders have to keep coming in as if nothing were happening. The good news is that Shopify is built for this. The mechanism is simple and almost nobody uses it properly.
The page-builder trap
Most stores we take over share the same history. A standard theme at the start. Then a need for a custom page, and a page-builder app gets installed. Then a second page, a third, and two years later the home page and the product pages live entirely inside the app.
It works, until it gets expensive. Pages load slower because they stack two systems. Theme updates become risky. And the day you stop paying the monthly subscription, your pages disappear.
That is exactly where Fling Toothpicks stood when the mandate started: flavoured toothpicks made in Quebec, five flavours, real brand personality, but a site where the home page and the product pages were built in PageFly on top of the Horizon theme. Plus a bundle builder that bypassed the cart, and no subscription option for a product that gets consumed and rebought.
“You never touch the published theme: the copy is the worksite and the old theme stays as the safety net.”
The duplicated-theme method
The principle fits in one sentence: you never touch the published theme.
You duplicate the live theme. The copy becomes the worksite. It has its own preview address, it pulls the same products, the same inventory and the same customers as the live store, but no visitor ever sees it.
For weeks you rebuild inside that copy. The live site keeps selling, with its current pages, without a single minute of downtime and without anyone on the team having to freeze anything.
When everything is ready and tested, publishing takes seconds. And crucially, the old theme stays in the library, unpublished. If something serious appears in the hour that follows, you republish the old one and the store is exactly as it was. That is the safety net, and it costs nothing.
Rebuilding native instead of in an app
For Fling, the home page was rebuilt as twelve native Horizon sections, no PageFly. The product pages too. The bundle builder was redeveloped from scratch so customers pick flavours and quantities through the normal cart rather than around it. Shopify Subscriptions was integrated into the product page with its two discount tiers, up to fifteen percent savings.
The switch happened on August 26, 2026, from the duplicated theme, without touching the live site before the moment of publishing.
What matters in that story is not the feature list, it is the underlying choice: anything built native belongs to the store. It survives the end of a subscription, it loads faster, and any Shopify developer after us can modify it.
The pre-launch checklist
Seven items, done in the duplicated theme's preview before the switch.
Place a complete test order, from cart to confirmation email. It is the only test that truly counts.
Check the journey on a phone, not just a desktop. Most of your buyers are on a six-inch screen.
Check out-of-stock cases: what the customer sees, what they can still add to cart, and what happens at checkout.
Check tracking. Pixel, conversions and purchase events must be wired to the new theme, otherwise you publish a blind store and your campaigns optimize on nothing.
Compare the URLs. If pages change address, every old address needs its redirect, otherwise you lose the search visibility you had.
Test active discount codes and automatic discounts.
Pick the moment. You don't publish on a Friday afternoon, the night before a big campaign, or during a running promotion.
The following 48 hours
Keep the old theme unpublished for at least two weeks. Watch three numbers: conversion rate, average order value, and orders per hour compared with the same day the previous week. An immediate conversion drop almost always comes from a checkout or tracking problem, not from the design.
This is the kind of redesign we run
Getting out of a page builder, rebuilding native, developing what's missing, and publishing without interrupting sales: it is a bounded project, with a date and a safety net. See our Shopify offer
Frequently asked questions
How long does a zero-downtime redesign take? A few weeks of building in the duplicated theme, then a few seconds to switch. The duration depends on how many templates need rebuilding, not on risk, since the store sells throughout.
Do you lose search visibility by changing themes? Not if the URLs stay the same. Visibility is lost when addresses change without redirects, which is a choice, not a fatality.
Do you really have to drop page builders? For the pages that make your revenue, yes. For a one-off promotional page built in an hour, the tool keeps its use. The problem starts when the home page and product pages depend on a monthly subscription.