MigrationsShopifyWebflowRoute 03 of 4
Shopify to Webflow, with a straight answer about what happens to your checkout
Webflow’s commerce is not Shopify’s commerce, and the gap is structural rather than cosmetic. So this page does the arithmetic before it does the sales pitch: what actually differs, what your CSV will and will not carry, and the three answers a real store gets, only one of which is “move everything”.
Three answers, one of them yours
- Least common
Move the whole thing
A small, stable catalogue. One currency, one stockroom, no subscriptions, no till. The website, not the back office, is what is holding growth back.
- Usually the right answer
Split the site from the checkout
Webflow takes the marketing site, the content and the landing pages. Shopify keeps the cart, the checkout and everything behind it. You get the design platform without re-plumbing the business.
- More often than you would think
Change nothing yet
Subscriptions, point of sale, stock in more than one place, a shelf of apps running the back office. The site is not what is costing you, and moving it would.
The split is the answer more often than the migration is
Shopify is an operations platform with a shop attached. Webflow is a design platform with a storefront attached. Most stores that want Webflow want the first half of that sentence (the pages, the campaigns, the content), and have no business giving up the second.
What a split actually looks like
- Webflow serves the homepage, the content, the landing pages and the campaign work.
- Shopify keeps the cart, the checkout, the customer accounts and every back-office tool.
- Product data is either mirrored into Webflow for display or pulled live, depending on how often it changes.
- One domain across both, with the checkout paths deliberately excluded from every redirect rule.
- Analytics configured across both hosts before launch, so attribution survives the handover.
What it costs you
Two systems instead of one, and a boundary that has to be maintained. Product changes live in one place and have to reach the other. Somebody has to own that, and if nobody does, the mirrored catalogue drifts and the wrong price ends up on a landing page.
It is a real cost and we would rather name it here than discover it with you in month three. It is still, for most stores with a working checkout, considerably cheaper than rebuilding the checkout.
Coming off Squarespace Commerce instead? The same arithmetic applies, start with the Squarespace routeand read this section as its commerce chapter.
Ten differences that are structural, not cosmetic
None of these is a feature waiting to ship. They are shapes the two platforms have held for years, and the third column is what each one costs the business rather than a score out of ten. Anything that varies by plan is written as something to check against your own catalogue, because a number quoted wrong here is a number that costs you money.
| Where they differ | Shopify | Webflow | What it means for you |
|---|---|---|---|
| The checkout itself | Shopify owns it end to end, accelerated wallets, saved payment details, a flow tuned on an enormous amount of traffic. | Webflow has its own hosted checkout with far fewer levers on it. | Your current conversion rate is attached to a checkout you would be replacing. Nothing about parity there can be assumed; it has to be measured. |
| Subscriptions and recurring billing | A mature ecosystem built on platform APIs, with contracts, dunning and customer-managed plans. | Not a native capability. | If you bill anyone repeatedly, this single row usually ends the conversation. Recurring revenue does not get migrated on a hunch. |
| Point of sale and in-person | Real hardware, staff accounts, and stock shared between the shop and the website. | None. | A business with a till stays on Shopify. There is no clever way around this one. |
| Stock across locations | Multi-location inventory, per-location availability, transfers. | A single, simple stock count per item. | Warehouses, shops and third-party fulfilment cannot be modelled. The people packing boxes feel it on day one. |
| Currencies, regions and duties | Regional pricing, presentment currencies and duty handling at checkout. | Sells in one currency. | International pricing is a downgrade, not a port. If more than a trickle of revenue is cross-border, price it before you plan it. |
| Shipping rates and tax | Live carrier-calculated rates, rate rules by weight and zone, and tax handling for the places you are registered. | Simpler rules, and a narrower set of regions. | Exact freight quotes and edge-case tax rules are the two things that quietly break. Check both against where you actually sell. |
| The apps behind the store | Third-party logistics, ERP, loyalty, reviews, helpdesk, returns, an ecosystem built for operations. | Far fewer, and mostly front-end. | Every app running your back office needs a named replacement with a named owner, or an honest "this stops". |
| Catalogue depth and variants | Built for deep catalogues with many options per product. | Item and variant limits that vary by plan. | Count your SKUs and your deepest variant matrix against the current published limits on both platforms before anything is designed. Not after. |
| Order operations after the sale | Order editing, partial refunds, fulfilment workflow, fraud analysis, chargeback handling. | Order records, and not much workflow around them. | The tooling your team uses every day is not part of the storefront, and it is the part they will notice missing first. |
| Customers, accounts and gift cards | Accounts with order history, saved addresses, store credit and gift card balances. | Memberships, built rather than inherited. | Passwords never transfer from any platform, order history does not follow, and outstanding gift card balances are a real liability that has to be honoured somewhere. |
Scroll the table sideways to read every column.
The CSV is good, which is exactly why people trust it too far
Shopify’s export is far better than anything Squarespace or Wix will hand you. It carries the catalogue cleanly, and it leaves behind most of what the store has accumulated around the catalogue, which is where the surprises live.
Comes across
- Products, with their options and variants
- Prices, SKUs, barcodes and weights
- Image URLs and alt text
- Inventory quantities, per location
- Customer records, without passwords
- The URL redirects already in your redirect table
That last row matters more than it looks. A store of any age has hundreds of redirects already protecting URLs it changed years ago. They are exported and carried across, or you lose rankings you had already recovered once.
Does not
- Metafields, where half your content often lives
- Automated collection rules, which are logic, not data
- Order history, as anything a new platform can import
- Subscription contracts and their billing state
- App-owned data, reviews, loyalty points, bundles
- Discount rules, gift card balances and store credit
- Theme code, sections and any script-tag installs
- Customer passwords, which never transfer
The two that catch everybody
Metafields. On a mature store, a startling amount of the content (care instructions, sizing, ingredients, specification tables) lives in metafields the basic product export does not include. They are inventoried separately and pulled through the admin API, and a store that has never audited them is usually surprised by the count.
Automated collections. A smart collection is a rule, not a list. It has to be rebuilt as filtering logic on the new platform, and the rule that decides which products appear is easy to get subtly wrong in a way nobody notices until a product is missing from a category for a month.
The redirect map, in Shopify’s own URL shapes
Shopify’s URLs are the most predictable of the four routes, which makes the map mostly wildcards. It also produces the one rule in this family that can break a business inside a minute if it is written carelessly, the fourth row.
| Shopify URL | Destination | Rule | Why |
|---|---|---|---|
/products/leather-tote | /shop/leather-tote | 301 · wildcard | The canonical product path. One rule covers the entire catalogue. |
/collections/bags/products/leather-tote | /shop/leather-tote | 301 · wildcard | The same product answers on every collection it sits in. Those paths get indexed and linked. |
/collections/bags | /shop/bags | 301 · wildcard | Collection pages are usually the strongest commercial rankings on the site. |
/cart · /checkout · /account | Left alone | Do not touch | On a split, these must keep resolving to Shopify. A greedy wildcard here breaks the store in minutes. |
/pages/size-guide | /size-guide | 301 · wildcard | Content pages lose their prefix. Every one of them, in one rule. |
/blogs/news/spring-drop | /blog/spring-drop | 301 · wildcard | A Shopify blog nests under its own handle. Flattening it changes every post URL. |
/policies/refund-policy | /refund-policy | 301 · one-to-one | Policy pages are linked from receipts and emails you cannot edit retrospectively. |
/collections/all | /shop | 301 · one-to-one | The catch-all collection. Often the second-most-linked URL on a store. |
Scroll the table sideways to read every column.
One product, a dozen indexed paths
Shopify answers the same product on /products/handle and on/collections/anything/products/handle. It canonicalises them for you today; the moment the store is gone, every one of those paths still exists in somebody’s index, somebody’s newsletter and somebody’s backlink profile. The map catches them with one wildcard, and it is not optional.
Never redirect the checkout
On a split, /cart, /checkout,/account and the app proxy paths have to keep resolving to Shopify. A greedy wildcard that swallows them takes the store offline while every other page looks perfectly fine, which is the worst possible failure, because nothing appears broken until the day’s revenue does not arrive.
Attribution crosses a boundary now
When the storefront and the checkout sit on different systems, your analytics has to be told they are one journey. Without cross-domain measurement configured before launch, every sale starts attributing itself to a referral from your own site and the reporting your marketing decisions rest on quietly stops being true.
Six phases, and phase two is allowed to end it
Most projects run four to eight weeks end to end, across discovery, design, build, QA and launch. Timelines hold because the sequence is the same every time. A store adds a phase in front of that: counting what you sell, before anyone designs anything that would make the counting awkward.
Count what you actually sell
SKUs, the deepest variant matrix, manual and automated collections, metafields, installed apps, order volume, active subscriptions, stock locations and every currency you take money in.
Nothing is designed yet. This phase exists to produce a number that either supports a move or does not.
YoursA product export, and an honest list of the apps you would notice if they vanished overnight.
Give the verdict in writing
Move, split or stay, with the reasons attached, and the revenue at risk under each one.
If the verdict is "stay", the engagement stops here and you pay for an audit rather than a rebuild.
YoursOne conversation with whoever owns the revenue, not only whoever owns the website.
Model the catalogue and the content
Products, options, collections and the filtering rules that were automated collections. Everything living in metafields gets a real field with a real name.
Templates, landing pages and campaign pages designed for the marketing side, which is where a Webflow front end earns its keep.
YoursWhich product attributes your team actually maintains, and which ones nobody has touched in a year.
Build both sides on staging
The storefront in Webflow on a staging domain, with the cart and checkout path wired exactly as it will be in production, including the Shopify side on a split.
Analytics and tag containers configured across both domains before launch, not after the first week of missing attribution.
YoursBrand assets, product photography, and copy for anything being rewritten.
Rehearse a real purchase
Buy something. Then refund it. Then check the confirmation emails, the tax line, the shipping quote, the inventory decrement and the order record at the other end.
Every redirect rule fired against the real URL list on staging, with the checkout paths explicitly excluded and proven to be excluded.
YoursDo the same purchase yourself, the way your customers do it, on your own phone.
Cut over in a quiet week
TTL down ahead of the switch, DNS changed, sitemap submitted, then Search Console, the 404 log and, on this route especially, the revenue graph watched hour by hour.
Nobody moves a store in the run-up to peak. The build can happen in October; the switch waits for a dull fortnight.
YoursNothing on the day, except telling us the moment a number looks wrong.
A real catalogue on Webflow, built rather than ported
Jobe’s Hats is a Webflow storefront we built, not a Shopify store we migrated, and we are not going to blur that line. What it shows is the shape of the “move the whole thing” verdict when it is the right one: a modest catalogue, collection logic rebuilt rather than converted, and a buy path with nothing standing in it.
Jobe's Hats
E-commerce 2024
Faster product pages, cleaner collection logic, and a checkout path with nothing in the way.
- +50%
- sales
- +40%
- revenue
When your store should not move at all
This is not a list of edge cases. Any one of these six is, on its own, a reason to keep the checkout exactly where it is, and you will hear it on the first call, before there is a proposal for anyone to defend.
You bill anybody on a subscription
Recurring revenue lives in contracts, payment mandates and billing state that do not travel. Rebuilding that on a platform without native subscriptions is not a migration, it is a rewrite of your revenue.
There is a till in a shop
Point of sale shares stock, staff and reporting with the online store. Webflow has no equivalent, so a business with counters keeps Shopify, and can still have a Webflow marketing site in front of it.
Stock lives in more than one place
Multi-location inventory, transfers and third-party fulfilment are back-office machinery. Losing them to gain a nicer homepage is a trade nobody in the warehouse agreed to.
You sell across borders
Regional pricing, presentment currency and duties are the difference between a checkout that converts abroad and one that abandons. Downgrading that is measurable within a week.
A shelf of apps is running operations
Returns, loyalty, reviews, helpdesk, ERP, 3PL. Count them, then count how many have a Webflow equivalent. That subtraction is usually the whole answer.
The catalogue is deep and moves daily
Thousands of SKUs, frequent price changes, deep variant matrices and feeds going out to marketplaces. Shopify is built for that shape of work and Webflow is not pretending to be.
If one of those is you, the answer is the split: a Webflow marketing site in front of the Shopify you already have. You get the design platform, the content model and the speed, and the machinery that takes the money is left alone, which is the outcome most stores wanted in the first place.