What Did Shopify Actually Change in Winter '26?

Sidekick, Shopify's AI assistant, got a new extension surface in the Winter '26 release, allowing third-party apps to register as data and action providers directly inside the Sidekick pane. That means apps can now push product attributes, trust signals, and merchant actions into Sidekick's context, not just display them on the storefront.

Before this, Sidekick was largely limited to first-party Shopify data: orders, inventory counts, basic product titles and variants. Useful, but limited. The extension model changes that. Now an app managing your certifications, your bundle logic, or your customer review corpus can surface that data in response to merchant queries, or customer-facing queries if you're running a consumer-mode Sidekick deployment.

Shopify has been pushing toward Sidekick as a full commerce operating layer, and app extensions are how third-party tooling plugs into that layer with structure instead of workarounds.

Why Should a Merchant Care About This Right Now?

Because Sidekick is already advising merchants, and increasingly customers, about which products to surface, what to reorder, and what to recommend. If your product data is thin or unstructured, Sidekick gives incomplete answers. Worse, it skips your products entirely in favor of ones it can reason about confidently.

Here's what's actually happening inside the recommendation engine. Sidekick weights responses based on what it can verify: attributes it can read, actions it can execute, and context it can trust. A product with rich metafield data, a return window, and a verified review count is more "answerably correct" to Sidekick than a product with a title, price, and a three-sentence description.

I've seen this firsthand auditing Shopify stores. Products with thin data don't show up in Sidekick's recommendations at query time, even when they're the right answer. The AI picks the product it can explain.

What Are the No-Code Options?

App-based metafield tools are the fastest on-ramp. No developer required. Several apps in the Shopify App Store now support the Winter '26 Sidekick extension interface, meaning they can write structured product data directly into the namespace Sidekick reads from.

Which Apps Are Already Sidekick-Compatible?

As of early April 2026, the apps with confirmed or partial Sidekick extension support include metafield managers, review platforms, and product data apps that have shipped updates since January 2026. Here's how to evaluate whether an app you already use is compatible:

• Check the app's changelog for "Sidekick extension," "AI assistant integration," or "Winter '26 update."
• In Shopify admin, go to Apps > [your app] and look for a "Sidekick" or "AI" toggle in settings.
• Reach out to the app's support team and ask directly if they've shipped the Sidekick extension API integration.

If the app hasn't shipped it yet, that's also useful information. It tells you to either push them on the roadmap or find an alternative that has.

What Product Data Should You Actually Fill In?

Don't try to fill everything at once. Focus on the attributes Sidekick queries most often.

• Use case tags and descriptors, What problem does this product solve? "For runners with wide feet" is more useful to Sidekick than "athletic shoe."
• Material and composition, Especially for apparel, home goods, and consumables. Sidekick gets asked "what's this made of?" constantly.
• Return window and policy, Trust signal. Sidekick surfaces this when customers are on the fence.
• Review count and aggregate score, If your review app supports Sidekick extensions, make sure this data is flowing through.
• Inventory context, Is this a seasonal item? A limited run? Low stock? Sidekick can use scarcity signals in recommendations if you expose them.

What's the Developer Path for Sidekick Extensions?

Full control. If you have a developer on your team or work with a Shopify developer partner, you can build a custom Sidekick extension that registers your app as both a context provider and an action handler, meaning Sidekick can not only read your data, but trigger actions in your app from a merchant prompt.

How Does the Sidekick Extension API Work?

The extension API, introduced in Winter '26 alongside Shopify's updated App Bridge documentation, works through a registration model. Your app declares what context it can provide and what actions it can handle. Sidekick routes relevant queries to your extension and expects a structured JSON response.

The practical steps look like this:

1. Register a Sidekick extension in your app's configuration using App Bridge.
2. Define context providers: endpoints that return structured product data, inventory metadata, or policy information when Sidekick calls them.
3. Define action handlers: functions Sidekick can trigger from a merchant prompt (e.g., "bulk update return windows for this collection").
4. Scope your data access appropriately, only request the merchant data your extension actually needs.
5. Test against real merchant prompts using Shopify's Sidekick extension sandbox (available in Partner Dashboard as of February 2026).

The extension sandbox is genuinely useful. It lets you simulate Sidekick queries against your extension before shipping, so you're not debugging in production.

What Trust Signals Does Sidekick Actually Weight?

Concrete and verifiable. Sidekick's recommendation logic favors attributes it can confirm from structured data over marketing copy it has to interpret. That's a meaningful distinction, and it changes how you should think about product content.

The signals that show up most in Sidekick's reasoning layer are:

• Return rate and return window, Low return rate is a quality signal. A generous window is a trust signal. Both matter.
• Review count and recency, Not just the average score. Volume and recency tell Sidekick the product is actively purchased and assessed.
• Certifications and compliance attributes, Organic, cruelty-free, food-safe, B-Corp. If these are true for your products and you're not surfacing them as structured data, Sidekick can't use them.
• Availability and lead time, If you're in stock and can ship in two days, that's a competitive signal Sidekick can act on. Expose it.
• Bundle compatibility, Does this product pair well with another? If your bundle app exposes that as structured data via a Sidekick extension, Sidekick can recommend complementary products accurately.

The pattern here is the same one I keep seeing in AI-native commerce generally: the merchants who win aren't the ones with the best marketing copy. They're the ones whose data is structured, complete, and accessible to the AI layer asking questions about it.

How Do You Know If Your Setup Is Actually Working?

Ask Sidekick. Seriously. Open it in your Shopify admin and query it the way a smart customer or merchant assistant would. The answers it gives, and the gaps it shows, tell you exactly where your extension data is missing or incomplete.

Try prompts like:

• "What's our best product for someone who needs X?"
• "Which products have the highest return rates this quarter?"
• "What are our top-reviewed products under $50?"
• "Do any of our products have organic certification?"

If Sidekick can't answer any of those confidently, that's your gap map. Each failed or vague response points to a data attribute you haven't structured or surfaced through an extension yet.

According to Shopify's 2026 Future of Commerce report, merchants using AI assistant features inside Shopify admin saw a 23% reduction in time spent on catalog management tasks. That number is going to climb as more product data flows through Sidekick's extension layer, but only for stores that actually set it up.