Shopify Canvas, the AI Store Builder: What It Means for Merchants and Custom Builds

Shopify Canvas, the AI store builder launched on 1 October 2026, lets merchants design a store by chatting with Sidekick. What it does, what it can't do yet, and where engineering and apps still matter.

Shopify Canvas, the AI store builder Shopify launched on 1 October 2026, changes how a Shopify store gets designed. Instead of picking a theme and editing it section by section, a merchant opens one workspace that shows every page of the store, describes what they want to Sidekick, Shopify's AI assistant, and watches the store change. Shopify's own changelog describes Canvas as a new design surface that lays out every page of your store in one interactive workspace, where you can pan, zoom, edit directly and work with Sidekick to build a custom store.

We build custom Shopify work for clients and run our own Shopify app, SellRank, so we have a practical interest in what this means. This post sets out what Canvas is according to Shopify and the reporting on the launch, what it does not do yet, what it changes for merchants and for anyone who pays for custom theme work, and where human engineering and apps still carry the load. We have tried to be fair: Canvas is a real step forward for many stores, and it also has clear limits at launch that Shopify itself has acknowledged.

What Shopify Canvas is, in plain terms

Canvas is a design surface inside the Shopify admin. The two primary sources for what it does are Shopify's changelog entry of 1 October 2026 and TechCrunch's report from the same day, which was based on Shopify's briefing. Taken together, they describe a tool with these characteristics:

• Every page of the store is laid out in one workspace that you can pan across and zoom into, rather than editing one template at a time.
• You build by chatting with Sidekick. You describe a change in plain language and Sidekick makes it.
• According to TechCrunch, what you see is not a static mock-up but the store's real code being rendered, so you can test interactivity and animation and view pages at different screen sizes.
• TechCrunch also reports that Sidekick takes screenshots so it can see what you see as it works, and that you can click on elements to edit them directly rather than only through chat.
• Shopify told TechCrunch it gave Sidekick direct access to theme files and simplified its theme architecture so that structure, logic and design are easier for Sidekick to understand and change.

That last point matters more than it first appears. Canvas is not a separate website builder bolted onto Shopify. It edits the same kind of theme files that a developer would edit, which is why the output can be genuinely custom, and also why some of the things a developer worries about, such as code quality, performance and maintainability, do not go away just because an AI wrote the code.

How Canvas fits with Sidekick and Shopify's earlier AI tools

Canvas did not appear from nowhere. In May 2025, as part of its Summer '25 Edition, Shopify introduced an AI store builder that generated a storefront from a short description of the business, along with the Horizon theme and a block generator that could create design elements from a text prompt, as TechCrunch reported at the time. Sidekick itself has been gaining abilities steadily: Shopify's changelog over the past year shows it working inside the online store editor, building Shopify Flow automations from a description, and gaining a fullscreen mode. On the developer side, Shopify's developer changelog says all app developers can now build Sidekick app extensions, so apps can connect their own data and workflows to Sidekick.

TechCrunch frames Canvas as bringing those Sidekick capabilities together into a full site builder. The difference from the 2025 store builder is that the earlier tool produced a starting point you then edited in the usual theme editor. Canvas is meant to be the place where the editing happens, conversationally, across the whole store at once.

What Shopify Canvas does not do yet

This is the part every merchant should read before rebuilding a live store in Canvas. According to TechCrunch, Canvas is desktop-only at launch, and it does not yet support the following. Shopify told TechCrunch it plans to add support for them over time, without giving dates.

• Third-party themes. If your store runs on a theme you bought from the Theme Store or had built by an agency, Canvas does not support it at launch.
• App blocks and extensions. Many stores get reviews, size charts, wishlists, bundles, upsells and similar features from app blocks. Those are not supported in Canvas yet.
• Markets. Stores that sell to several countries with different settings per market cannot manage that in Canvas yet.
• Translations. A store in more than one language cannot handle its translated content in Canvas yet.
• Rollouts. Shopify's Rollouts feature, which lets a merchant schedule and test changes, is not supported in Canvas at launch.
• Theme updates. Canvas does not yet support theme updates.

Put plainly, Canvas at launch suits a new store, or a simple store willing to start from Shopify's own theme foundations, much better than it suits an established store with a purchased or custom theme, a stack of apps embedded on its pages, and more than one market or language. That is not a criticism. It is a normal launch state for a new tool, and Shopify has said these gaps are on its list. It does mean the sensible timeline for most established stores is to watch, experiment on a copy, and move later.

What Canvas means for merchants

For a merchant, the most useful way to think about Canvas is as a much faster way to produce and test design ideas. Several things that used to need a developer or a long session in the theme editor can now be described in a sentence. That shifts where a merchant's time goes.

• Faster first versions. A founder starting a new store can get to a credible, branded design much sooner, which matters most for small businesses that could not justify a design budget.
• Cheaper experimentation. Trying a different homepage layout or a new product page structure becomes a conversation instead of a ticket. That makes it easier to try more ideas, provided someone measures which ones work.
• Fewer small change requests. The stream of minor requests, such as moving a banner, changing spacing or adding a text block, is the kind of work a conversational editor handles well.
• A clearer view of the whole store. Seeing every page at once makes inconsistencies obvious: a button style that differs on one template, or a collection page that never got the same care as the homepage.

There are also new responsibilities. When it becomes easy to change the store, it becomes easy to change it too often, without a record of what changed and why. A store that edits its homepage every few days in Canvas and never measures the result has not become more agile. It has become harder to learn from. The discipline of changing one thing, recording it, and comparing results matters more, not less, when changes are cheap.

What Canvas means for custom theme builds

The obvious question for anyone who sells or buys custom Shopify development is whether Canvas replaces it. Our honest reading is that it replaces some of it, changes more of it, and leaves the hardest parts where they were.

The work most exposed is the straightforward kind: taking a brand guideline and a set of reference sites and producing a good-looking theme with standard sections. If a merchant can describe that to Sidekick and get a working result in a sitting, paying for weeks of that work becomes hard to justify. Agencies and freelancers whose main offer is visual theme customisation should expect that part of the market to shrink.

The work least exposed is everything a design surface does not see. A storefront is the visible layer of a system that includes product data, inventory, pricing rules, checkout extensions, integrations with ERPs and logistics providers, analytics, consent, performance budgets and the habits of the team that runs it. Canvas operates on the theme. It does not decide how your product data is structured, how your warehouse sync handles an oversell, or whether your tracking survives a checkout change.

In between sits work that changes shape. Developers who used to hand-build sections will increasingly review, correct and harden what an AI produced. That review matters. Code that renders correctly in a preview can still be slow on a mid-range Android phone on a patchy connection, can still fail an accessibility check, and can still be difficult for the next person to maintain. These are the same reasons we wrote about theme debt making stores slower year after year; AI-generated theme code can accumulate that debt faster, because it is produced faster.

Where human engineering still matters

Here is our list of the places where we expect experienced engineering to remain necessary, with or without Canvas. None of them is a claim that Canvas is weak. They are simply outside what a design surface is for.

• Data structure. Metafields, metaobjects, variant design and collection structure decide what a theme can show. A badly modelled catalogue produces a confusing store no matter how good the design is.
• Integrations. Connecting Shopify to an ERP, a warehouse, a shipping aggregator, a CRM or a custom pricing engine is backend work with failure modes that need monitoring.
• Checkout and post-purchase. Checkout UI extensions, Shopify Functions for discounts and delivery customisation, and tracking on the Thank you page are separate surfaces with their own rules.
• Performance. Core Web Vitals depend on what loads, in what order and how heavy it is. Somebody has to measure the real store on real devices and push back on what slows it down.
• Accessibility. Colour contrast, focus order, alt text and screen-reader behaviour need testing, not only a visual check.
• Measurement. Knowing whether a change helped requires clean tracking, a baseline and patience. A design tool does not provide that by itself.
• Review and ownership. Somebody needs to understand the code well enough to fix it at 11pm during a sale, when a chat interface is not the right tool for a production incident.

Where apps still matter

Apps on Shopify do two broad kinds of work: they add things to the storefront, and they do work behind it. Canvas affects these differently.

Storefront apps that render through app blocks are the most affected in the short term, because Canvas does not support app blocks and extensions at launch. A store that relies on an app block for reviews or a size guide cannot place it through Canvas until that support arrives. Over time, as Shopify adds app block support, we would expect these apps to become available inside Canvas like any other block.

Apps that work behind the storefront are mostly unaffected, because they do not depend on how the theme was built. Inventory sync, order routing, fulfilment, accounting connectors, and merchandising tools that write data into Shopify itself keep working the same way. Sidekick app extensions also point to a future in which merchants ask Sidekick to do something and an app does the work, which is a sign that Shopify sees apps as part of the AI picture rather than replaced by it.

The collection page: design versus decisions

One distinction is worth drawing out because it is easy to blur. Canvas can make a collection page look excellent: the grid, the filters, the product cards, the banner at the top. It does not decide which products appear in the first row, and on most collection pages that decision affects sales more than the layout does.

That decision lives in the collection's sort order, which is store data rather than theme design. Shopify offers its built-in sort options, such as best selling, newest and price, or a manual order set by a person. A collection page designed in Canvas will still show products in whatever order the collection is set to. If that order is all-time best sellers with an out-of-stock product at the top, the beautiful new design will show that product first.

This is the problem we built SellRank to solve. SellRank ranks the products in a collection by real performance, such as views, add-to-carts, orders, revenue, conversion rate or profit, and writes that order into Shopify's own manual collection sort. It needs no theme edits and no code, so how the theme was built, by a developer, an agency or Sidekick in Canvas, does not change how it works. Design and merchandising are separate jobs, and a store benefits from doing both well.

A sensible plan for an existing store

If you run an established Shopify store, here is a practical way to approach Canvas over the next few months.

1. 1Leave your published theme alone through the festive and BFCM peak. Stability is worth more than novelty in October and November.
2. 2List what your store depends on that Canvas does not support yet: a third-party theme, app blocks, markets, translations. If any apply, treat Canvas as an experiment for now.
3. 3Try Canvas on a duplicate theme or a test store. Ask it to rebuild one page you know well, such as a collection page, and compare the result with your current page on a real phone.
4. 4Check the output the way you would check a developer's work: page speed, accessibility, how it behaves with long product titles and missing images, and whether your tracking still fires.
5. 5Watch Shopify's changelog for app block, markets and translation support. Those are the signals that Canvas is ready for a full store.
6. 6When you do move, change one area at a time and record what changed and when, so you can measure the effect rather than guessing.

A sensible plan for a new store

For a founder starting from scratch, the calculation is different. A new store has no legacy theme, often no app blocks yet and usually one market and one language, which is close to the situation Canvas handles well at launch. For that founder, Canvas is a reasonable place to start. Spend the money you might have put into visual design on the things Canvas does not do: clean product data, good photography, honest shipping and returns information, and tracking that works from the first order.

Keep the code reviewable. Even if you never plan to hire a developer, the day may come when you need one, and a theme they can understand will cost far less to extend than one they have to untangle.

Questions to ask before paying for a custom build now

If you are about to commission custom theme work, Canvas gives you useful questions to put to any developer or agency, including us.

• Which parts of this project could Canvas do today, and why are you proposing to build them by hand?
• Which parts depend on things Canvas does not support yet, such as a third-party theme, app blocks, markets or translations?
• How will the theme stay editable by our team, in the theme editor or Canvas, after you hand it over?
• What performance and accessibility targets will you test against, and on which devices?
• What happens to our tracking, checkout extensions and integrations during the build?
• Who owns the code, and how will the next developer understand it?

A good partner will welcome these questions. Some of the honest answers will be that a part of the project is better done in Canvas by your own team, and that the developer's time is better spent on data, integrations and measurement.

What we are watching next

Canvas's long-term importance depends on how quickly Shopify closes the launch gaps. Support for third-party themes would open it to most existing stores. Support for app blocks would let merchants arrange the features they already rely on. Markets and translations would make it usable for stores that sell across borders or in several languages, which includes many Indian brands selling abroad. Rollouts support would let merchants test Canvas changes on part of their traffic before committing, which is exactly the kind of discipline cheap changes require.

We will update this post as Shopify publishes those changes in its changelog. Until then, our view is simple. Canvas lowers the cost of design. It does not lower the cost of getting the data, integrations, checkout, speed and measurement right, and those are where most stores win or lose.

What Canvas means for Indian D2C stores

Most of the clients we work with are Indian brands, and a few points apply to them in particular. Many Indian D2C stores run on a purchased theme that has been customised over several years by different freelancers. That is exactly the setup Canvas does not support at launch, so for these stores Canvas is something to evaluate later rather than adopt now.

Indian brands that sell abroad usually rely on Shopify Markets for currency, pricing and shipping per country, and some run their storefront in English and Hindi or another regional language. Until Canvas supports markets and translations, those stores should keep using the existing theme editor for anything that varies by market or language.

Performance deserves particular attention. A large share of Indian shoppers browse on mid-range Android phones and mobile data, where heavy pages lose people before the first product loads. Whatever Canvas produces, test it on that kind of device and connection, not only on a fast laptop in the office. If a design looks good but loads slowly on a typical customer's phone, it is not a good design for that store.

Finally, the festive calendar. With Diwali falling on 8 November this year, Indian stores are already in their busiest stretch. That is one more reason to treat Canvas as a January project for an established store, and to spend October on stock, collection pages and delivery promises instead.

Talk to us about your store

If you are weighing Canvas against a custom build, or want a second opinion on what your store actually needs, tell us about it on our contact page and we will give you a straight answer, including when the answer is that you do not need us. And if your collection pages look good but show the wrong products first, SellRank has a free plan for one collection so you can see the difference on your own store.