Webflow Ecommerce Limitations: What Webflow Cannot Do in 2026

On January 29, 2026, Webflow switched off User Accounts, and with it every native path to recurring revenue on a Webflow store. Subscriptions were built on top of User Accounts, so when Webflow disabled the feature, its dedicated APIs and all of its webhooks on that date, recurring billing went with them. Webflow's own remedy is a third-party tool, with discounts arranged at Outseta and Memberstack.

Four months later the rest of the platform was rebuilt. On May 13, 2026 Webflow merged the CMS and Business plans into a single Premium plan at $25 per month billed yearly, raised the CMS item allowance from 2,000 and 10,000 to 20,000, added AI credits to every Workspace plan, and introduced a $2,500 per month Team plan. The same FAQ that documents all of it has one line for merchants: "Nothing is changing for the Ecommerce Site plans at this time."

None of that makes Webflow a poor place to build a storefront. The Designer is the strongest visual development environment on the market, the hosting is excellent, and Webflow is one of the few hosted builders where the cart, checkout and order confirmation are editable pages rather than a vendor-controlled form. The useful question in 2026 is narrower: where is Webflow investing, and does commerce sit anywhere near the front of that queue? The pricing page answers it plainly. Site plans are headlined as an AI-native platform. Ecommerce sits behind a separate tab, with no new badge, no AI features listed, and item caps that were last raised in November 2022.

Key Takeaways

• Native subscriptions no longer exist for a new store. Recurring billing required User Accounts, and User Accounts was sunset on January 29, 2026, taking the feature, its dedicated APIs and its webhooks with it.
• Products, variants and categories share one budget. Every product, SKU and category counts against 500 items on Standard, 5,000 on Plus and 15,000 on Advanced, and a plain product with no options already costs two items.
• Variants cap at 50 per product across a maximum of 3 option sets. That is roughly five times tighter than Squarespace's 250 combinations and twenty times tighter than Wix's 1,000 variants.
• The May 2026 replatform skipped stores. A content site now gets 20,000 CMS items on a $25 per month Premium plan while ecommerce stays at $29, $74 and $212 per month for 500, 5,000 and 15,000 items.
• The Data API is a catalog and an order ledger, not a commerce engine. It covers products, SKUs, orders and inventory with no cart, checkout, customer, discount or subscription endpoints, and only three ecommerce webhook events.
• One store currency, and Ecommerce is excluded from Webflow Localize. Changing the currency sets every existing product to draft and invalidates configured shipping methods.

What Webflow Genuinely Does Better Than Its Peers

A limitations analysis is only useful if it is honest about what the platform is good at, and Webflow is good at a great deal. Skipping that part produces a comparison no working designer would recognise.

The Designer has no real equivalent

Webflow's Designer is a visual interface over the actual box model. A designer works with real classes, real flexbox and grid, real breakpoints and real cascade behaviour, and the output is markup a developer can read. Nothing else in the hosted-builder category comes close, which is why Webflow retains agencies and in-house brand teams that would never touch a template builder. Hosting, the CDN, SSL and publishing are solid and largely invisible, which is the correct behaviour for infrastructure.

The cart and checkout are real pages, not a locked form

This is where the usual criticism of hosted builders stops applying. In Webflow, the cart, the checkout page, the separate PayPal checkout page and the order confirmation page are Webflow pages you design and restructure. Components can be reordered and restyled, error states have an editable state in the Designer, discount code entry sits on the page, and an Additional Info block collects custom fields at checkout. That is a materially better position than the checkout ceiling on Wix or the half-styleable checkout on Squarespace, and anyone arguing that Webflow's checkout is locked has not used it.

The ceiling here is different in kind. There are exactly two payment providers, there is no server-side logic that runs during checkout, and there is no way to add a step, a conditional flow, or a third-party fraud or tax call inside the transaction. The page is yours; the transaction is not.

The recent platform work is real, and it is aimed elsewhere

Webflow Cloud ships web-app hosting with SQLite, a key-value store and object storage. An MCP server is listed on every Site plan including the free Starter tier, and Webflow Conf 2026 added Source, Campaigns, Assets, Releases, AEO content agents and expanded MCP access covering interactions, CMS querying and Webflow Cloud deploys. It is a genuinely strong year of shipping. The opening keynote contained no commerce announcement at all.

Recurring Revenue Did Not Get Narrower. It Ended.

Most platform comparisons describe subscription support as limited or constrained. For Webflow in 2026 that framing is wrong in the opposite direction. There is no native subscription product for a store built today.

What the sunset actually removed

The subscriptions documentation is explicit that offering subscriptions requires Ecommerce and User Accounts to be enabled together, plus Stripe and an Ecommerce Site plan. New sites lost the ability to enable User Accounts on January 31, 2025. On January 29, 2026 the feature was disabled everywhere, User Accounts data was not migrated to the CMS, content gates were removed, and the dedicated APIs and webhooks stopped responding.

Merchants who were already selling subscriptions were left in an awkward position rather than a clean one. Webflow confirms that existing paid subscriptions remain in Stripe and subscribers continue to be charged, which means the money still moves but the platform no longer knows about it. The store owner keeps a Stripe billing relationship with no product record, no gated content and no in-product management surface attached to it.

The mechanics were already punishing before the sunset

Understanding what the feature actually was matters here, because merchants evaluating third-party replacements often assume they are replacing something sophisticated. Webflow's own documentation describes a deliberately thin implementation:

• No mixed cart. Subscriptions could not be purchased in the same transaction as other products, so a subscription box plus a one-off add-on meant two checkouts.
• No discounts, at all. Discount codes applied only to products with a one-time payment billing method, which removes the entire acquisition-offer playbook that subscription brands run on.
• Immediate revocation on cancel. Cancelling ended access on the spot with no remainder-of-period grace, so a customer cancelling on day two of a paid month lost what they had paid for.
• One of each subscription per customer. The subscribe button was disabled if that subscription was already active on the account, which rules out gifting, multiple boxes and quantity changes.
• Renewals generated no order. Only the initial purchase created an order record, so recurring physical fulfilment had to be tracked by hand from the subscriptions list.

That last point is the one that should decide the question for anyone selling a physical subscription. A platform that does not create an order on renewal cannot drive a pick list, a packing slip, a shipping label or a 3PL handoff, which is most of the operational work in a subscription box business.

What replacing it involves

The practical answer is a membership or billing tool bolted alongside the store: Outseta and Memberstack are the two Webflow named, and there is a wider field of dedicated subscription billing platforms. This works, and plenty of teams run it. It also means the subscriber record, the billing schedule, the gated content rules and the catalog live in different systems with no shared customer object, and every report that combines one-time and recurring revenue becomes a reconciliation exercise.

One Item Budget for Products, Variants and Categories

Webflow's catalog ceiling is structurally unlike its peers, and the number on the pricing page is not the number that matters.

How the counting works

Products, product variants and categories all draw down the same ecommerce item allowance. Webflow's documentation adds the detail that catches people out: every product automatically includes a default SKU even with no options configured, and that SKU is counted separately. A product with no variants therefore costs two items, not one. Once option sets are added, the generated combinations replace that default SKU, so a product costs one item plus one for every variant.

Three option sets, fifty variants

On top of the shared pool sit two hard per-product caps. Webflow allows up to 3 option sets per product and a total of 50 variants, and the documentation gives the worked constraint: five options in the first set leaves room for ten in the second. Squarespace permits six options and 250 combinations against a separate product cap; Wix permits 1,000 variants per product. Webflow's 50 is the tightest ceiling among the major visual builders, and it is the one that decides whether an apparel or footwear catalog fits at all.

The arithmetic on a Standard plan

Put the two rules together on the $29 per month Standard plan and its 500 items:

• A product at the 50-variant maximum costs 51 items. Nine such products consume 459 of the 500, before a single category exists.
• A normal apparel product with 5 sizes and 4 colours costs 21 items. That allows roughly 23 products plus categories, not 500.
• A catalog of simple, option-free products costs 2 items each. Standard tops out near 240 of those once categories are counted.
• Categories are charged too, so a merchandising structure of 20 collections is 20 items off the same budget.

The upgrade path is real but steep: Plus at $74 per month for 5,000 items, Advanced at $212 per month for 15,000. A footwear brand with 40 styles across sizes 6 to 13 and four colourways is already past Standard and well into Plus on variants alone. This is a different failure mode from a simple product cap, because the merchant who hits it is usually not selling many products. They are selling a normal number of products in a normal number of sizes.

May 2026: The Replatform That Went Around Ecommerce

The clearest signal about Webflow's commerce roadmap is not a missing feature. It is a set of changes that deliberately excluded stores.

What content sites received

The May 13, 2026 change sheet is generous. Basic rose to $15 per month yearly with 300 static pages. CMS and Business collapsed into Premium at $25 per month yearly, carrying 20,000 CMS items and 40 Collections, form file upload, well-known files, site search and code components. Webflow Cloud limits were rebalanced to 2 million web-app requests per month. CMS item add-ons were retired because the new allowance absorbed them. AI credits arrived on every Workspace plan, and a Team plan appeared at $2,500 per month with AEO agents, Localize and publishing workflows.

What stores received

Nothing. That is Webflow's word, not an inference. Ecommerce Standard, Plus and Advanced remain at $29, $74 and $212 per month billed yearly, carrying 500, 5,000 and 15,000 ecommerce items, with a 2% transaction fee on Standard and 0% on the two higher tiers. The last time those item limits moved was November 2022, when Plus went from 1,000 to 5,000 and Advanced from 3,000 to 15,000.

Reading the pricing page as a roadmap document

The result is a pricing page that tells the story without editorial help. A marketing site gets 20,000 CMS items for $25 per month. A store gets 500 ecommerce items for $29, and on the same $29 plan a variant of a t-shirt consumes the same budget as the t-shirt. The Site plan section is introduced as an AI-native platform and lists Webflow AI and the MCP server on every tier including the free one. The Ecommerce tab lists three or four bullets per plan: item count, transaction fee and custom cart and checkout, plus unbranded emails on Plus and Advanced. It carries no new badge and no AI capability.

Merchants are usually advised to compare feature lists. The more useful exercise here is to compare investment, and Webflow has made that easy to read.

What the Ecommerce API Cannot Do

Webflow's Data API is well documented, versioned and pleasant to work with. The issue is scope. It manages a catalog and reports on completed orders; it does not expose the machinery a transaction runs on.

A catalog and an order ledger

The ecommerce surface of the Data API consists of product and SKU reads and writes, order listing, retrieval, updating, fulfilment, unfulfilment and refunds, inventory listing and updating, and a single read call for ecommerce settings. For syncing a PIM into the catalog, or pulling orders out into an ERP, that is adequate, and the refund and fulfilment endpoints are more than some competitors offer.

The surfaces that are missing

What is absent defines the ceiling more than what is present:

• No cart endpoints. A cart cannot be created, read or modified through the API, so no external application can hold or resume a basket.
• No checkout or payment-session endpoint. The transaction happens inside Webflow's own pages and nowhere else.
• No customer resource. There is no customer object to create, query or attach a record to, which is a direct consequence of User Accounts going away.
• No discount or promotion endpoints. Discount codes exist in the product but cannot be generated, issued or expired programmatically.
• No order creation. Orders can be read and updated but not written, so a POS, a marketplace or a wholesale portal cannot push an order into Webflow.
• No subscription administration, since the underlying feature no longer exists.

Three webhook events, and what that means for headless

The webhook catalogue offers exactly three ecommerce events: new order, updated order and updated inventory. A store can be told that something already happened. It cannot participate in what is happening. That is the practical reason a Webflow store cannot be driven headlessly: the pieces a headless build needs, a cart to manipulate and a checkout to call, are the two pieces the API does not expose. Teams that want the Webflow Designer for the front end and a real commerce backend behind it generally end up pairing a visual tool with a separate commerce API, which is a reasonable architecture and a very different project from turning on an ecommerce plan.

The Rest of the Ceiling

The remaining constraints are individually smaller, and collectively decide whether a growing store stays.

One currency, and Localize does not cover the store

A Webflow store has a single currency, set once from the business address. Changing it later is not a settings toggle: Webflow's setup guide warns that changing the currency leaves existing price values untouched, saves every product as draft, and requires each price to be re-entered and every configured shipping method rebuilt before republishing. A store with 300 products faces 300 manual price edits and an offline storefront in between.

The second half of this is easy to miss. Webflow Localize is the company's flagship internationalisation product, sold as an add-on from $9 per month. The Ecommerce overview states that Ecommerce Site plan features do not support Localize, and that on sites with both enabled, products, categories and ecommerce pages cannot be localised. A merchant can translate the marketing site and not the catalog. Given that local-currency pricing measurably affects conversion, this is the constraint most likely to cap an otherwise healthy brand's international growth.

Two payment providers, and their rules are the store's rules

Webflow Ecommerce supports Stripe and PayPal and nothing else. Web payments such as Apple Pay and Google Pay route through Stripe, and Buy Now buttons only work on Stripe-enabled stores. There is no provider abstraction, no regional processor, no alternative gateway for a category one of the two declines. Whether a merchant can sell at all is therefore decided by Stripe's restricted businesses list and PayPal's acceptable use policy, not by Webflow. One geographic quirk deserves a flag: Stripe does not support application fees in Brazil, India or Malaysia, so merchants based there must be on Plus or Advanced rather than Standard.

Automatic tax stops at four jurisdictions

Sales tax and VAT are calculated automatically through TaxJar for the United States, Canada, the European Union and Australia. Everywhere else, the documentation says taxes must be calculated manually outside Webflow, and notes an intention to allow manual regional rates in future. Per-product tax classes exist, but the exemption class is limited to the US and certain EU countries. For a store selling into the UK, Japan, Singapore, Switzerland or Latin America, tax is an off-platform process.

Digital goods are a link you host and police yourself

Digital downloads are a file URL attached to a product, with support described as limited to a single externally hosted file per product variant. Webflow does not host the file, does not expire the link, does not cap download counts and does not bind the download to the purchaser. The documentation puts responsibility for the URL being valid, reliable and legal on the store manager. For a course, a font licence or a software key, that is a licensing model built on trust.

B2B and wholesale are a workaround

There is no wholesale price list, no customer group, no net terms, no minimum order quantity and no quote flow. With User Accounts gone there is no logged-in customer construct to attach a trade price to. The usual workaround, a password-protected page or a second site carrying trade pricing, now needs a third-party membership tool to gate it, and the second site carries its own Ecommerce plan because Webflow bills per site.

Costs, Compliance and the Cost of Leaving

The parts of a platform decision that never appear in a feature matrix are usually the ones that determine the outcome.

What a store actually pays

Webflow bills per site and the headline figures assume annual commitment; monthly billing costs more. A realistic Standard store pays $29 per month for the Ecommerce plan plus 2% of every order including tax, plus Stripe or PayPal processing on top, plus a Workspace plan if more than one person needs Designer access. Conversion testing is the $299 per month Optimize add-on. A membership tool to replace subscriptions is another line. The 2% disappears at Plus, which means the effective decision point for a growing store is whether $74 per month is cheaper than 2% of revenue, and for anything above roughly $2,250 of monthly sales it is.

Compliance sits with the processors

Because payments are Stripe and PayPal only, acceptable-use questions collapse to a single paragraph: whatever those two permit, Webflow permits, and whatever they decline is simply not sellable on the platform. There is no second route and no appeal inside Webflow.

What migration costs on the way out

Products, variants and categories export cleanly to CSV, and the Data API can pull orders. Customers do not export, because since January 2026 there is no customer record to export. Order history exists in the Designer and through the API but is not a normalised customer ledger. Subscribers, for stores that had them, live in Stripe. The design work does not transfer at all, since Webflow's output is bound to its own hosting and CMS. Teams should plan a Webflow exit as a rebuild of the storefront plus a data reconciliation, not a migration.

Who Should Stay on Webflow, and Who Should Not

The honest verdict is not that Webflow is bad at commerce. It is that Webflow has decided commerce is not the competition it wants to win, and merchants should read the roadmap rather than the feature list.

Choose Webflow Ecommerce if

• The site is the product. A brand, editorial or campaign site where a modest shop is one section of a larger content property is exactly what Webflow is built for.
• The catalog is small and shallow. Under about 100 products with few or no variants fits comfortably inside a Standard or Plus item budget.
• One currency and one market cover the business, and tax obligations fall inside the US, Canada, the EU or Australia.
• Design control at checkout matters more than logic at checkout. Webflow wins this comparison against every other hosted builder.
• Revenue is one-time. No subscriptions, no memberships, no recurring fulfilment.

Choose something else if

• Recurring revenue is part of the model. There is no native path, and the replacement is an integration project.
• The catalog has real variant depth. 50 variants and 3 option sets per product is a wall, not a guideline.
• Selling happens in more than one currency, or the catalog needs to be localised alongside the site.
• Anything needs to write into the store. No cart, checkout, customer or order-creation endpoints means no POS, no marketplace sync, no wholesale portal.
• The roadmap matters. A platform whose commerce product was explicitly excluded from its own replatform is a bet on stasis.

Where Swell is the clear frontrunner for these specific gaps

Swell is the strongest answer to the three ceilings this analysis keeps returning to, and the reasons are specific rather than a feature list. First, subscriptions and recurring billing live in the core API on every plan, as a first-class object where renewals of physical subscriptions generate real orders, which is the exact capability Webflow removed. Second, carts, orders and customer accounts are addressable API resources, so an external system can create a cart, drive a checkout and write an order, which is what makes a genuinely headless build possible rather than a catalog sync. Third, on the Standard and Unlimited plans multi-currency is a pricing capability rather than a display conversion: priced currencies let a catalog carry real prices per market and check out in that currency, while Starter and Basic get multi-currency display.

The tradeoff runs the other way too, and it should be stated. Swell has no visual page builder in Webflow's class; the front end is built by developers or with a storefront template. On cost, Swell's published plans are revenue-tiered rather than flat: Starter at $29, Basic at $79, Standard at $299 and Unlimited at $2,250 per month billed yearly, each with a trailing-twelve-month revenue ceiling of $50K, $250K, $1M and $5M and an overage above it of 2%, 1.5%, 1% and 0.4%, with custom pricing beyond $10M in annual sales. A small store with no subscriptions and a shallow catalog will find Webflow simpler and cheaper. A store with recurring revenue, deep variants or multi-market pricing will find that the ceilings arrive faster than the savings accumulate.

Frequently Asked Questions

Can a new Webflow store sell subscriptions in 2026?

Not natively. Subscriptions required User Accounts, new sites could not enable User Accounts from January 31, 2025, and the feature was disabled everywhere on January 29, 2026 along with its APIs and webhooks. A new store needs a third-party membership or billing tool such as Outseta or Memberstack, with the catalog and the subscriber record living in separate systems.

What happened to subscriptions that were already running?

Webflow states that existing paid subscriptions remain in Stripe and subscribers continue to be charged. The platform side is gone: no gated content, no subscriptions element, no in-product management. Merchants in that position hold a live Stripe billing relationship with no corresponding product or customer record in Webflow.

How many products can a Webflow store really hold?

Fewer than the plan number suggests, because products, SKUs and categories share the allowance and every product carries a default SKU. On the 500-item Standard plan, option-free products cost two items each, a five-size four-colour product costs 21, and a product at the 50-variant maximum costs 51. A SKU-heavy apparel catalog exhausts Standard at around ten products.

Is Webflow's checkout really customisable?

Yes, and this is a genuine strength. The cart, checkout, PayPal checkout and order confirmation are Webflow pages whose components can be reordered, restyled and extended with custom fields and error states. The limit is behavioural rather than visual: two payment providers, no server-side logic during the transaction, and no ability to add a step or an external call inside checkout.

Can Webflow Ecommerce run headless?

No. The Data API covers products, SKUs, orders, inventory and a settings read, with no cart, checkout, customer or order-creation endpoints and only three ecommerce webhook events. An external application can sync a catalog and read completed orders, but it cannot hold a basket or complete a purchase, which are the two things a headless storefront needs.

Does Webflow support multiple currencies?

No. A store has one currency, and changing it sets every existing product to draft, requires each price to be re-entered manually and invalidates configured shipping methods. Webflow Localize, the platform's internationalisation add-on, explicitly does not cover ecommerce products, categories or pages.

Did the May 2026 pricing change improve anything for stores?

No. Webflow's FAQ says nothing is changing for the Ecommerce Site plans at this time. Content sites moved to a Premium plan at $25 per month with 20,000 CMS items, while the three ecommerce tiers kept their 2022-era item limits and their existing prices of $29, $74 and $212 per month billed yearly.