Shopify Overhauls AI Shopping Infrastructure: Migrates Standalone Knowledge Base to Merchant Admin

SUNNYVALE, Calif. — In a significant structural update designed to streamline AI-driven commerce, e-commerce giant Shopify has announced the complete retirement of its standalone Knowledge Base app. The functionality is being permanently integrated directly into the core Merchant Admin, specifically engineered to power next-generation AI shopping agents.

According to Shopify’s official October 26 changelog, the legacy standalone application will remain accessible for a transitional period before being officially shuttered on January 15, 2027. This migration marks a pivotal shift in how merchants manage non-product metadata, aligning storefront intelligence with the rapid evolution of agentic workflows and automated buyer interactions.

Main Facts: The Shift to Admin-Native AI Knowledge Bases

The core transformation centers on where and how merchant data is aggregated for artificial intelligence. Under the new architecture, Shopify’s Merchant Admin knowledge base—navigated via Sales channels > Agentic > Knowledge base—serves as the central repository for all non-product-related information.

This repository supplies AI shopping agents with the contextual data required to accurately answer complex customer inquiries regarding store operations, such as dynamic shipping times, return windows, warranties, payment options, and foundational brand values.

Automated Publishing and Storefront Scans

The system organizes store metadata across ten distinct categories, encompassing logistical details like delivery and fulfillment, payment methods, returns, second-hand purchase policies, and an open-ended "Additional Facts" bucket.

Data is gathered dynamically through a combination of:

• Automated storefront scans
• Existing Shopify admin settings
• Direct manual input by merchants

Shopify performs regular automated scans of merchant storefronts to unearth new information. When these scans detect potential updates, they are surfaced to the merchant for manual review, allowing them to explicitly approve or reject the newly generated data. However, Shopify notes that automated scans occasionally miss granular storefront adjustments. Merchants retain the ability to manually request a full store rescan, though processing times can take up to 48 hours.

Public Availability via Storefront MCP

Once published, a store’s knowledge base becomes publicly accessible through the Storefront Model Context Protocol (MCP). This allows third-party platforms and AI-native applications—such as Shopify’s own Shop app—to seamlessly query store policies in real time.

Crucially, custom storefronts or those operating on Shopify’s Agentic plan are currently restricted to adding "Additional Facts" to their knowledge base. Furthermore, password-protected storefronts cannot be successfully scanned by automated crawlers; consequently, their knowledge bases will exclusively display data pulled directly from admin settings.

Chronology of the Transition and Deadlines

The timeline for deprecation gives smaller merchants a smooth, automated transition, while imposing strict manual requirements on larger, more complex retail operations.

• October 14, 2025: SEO expert Bill Hunt publishes an extensive analysis testing 122 Shopify stores, revealing widespread gaps in how AI shopping tools handle nuanced customer queries like order cancellations.
• October 26, 2025: Shopify officially publishes its changelog announcing the deprecation of the standalone Knowledge Base app and the rollout of the native Merchant Admin knowledge base.
• Present – January 15, 2027: The extended grace period. Smaller stores have already had their data automatically migrated, while large retailers must manually audit and re-enter data exceeding system caps.
• January 15, 2027: Hard cutoff date. The legacy Knowledge Base app will be entirely discontinued and removed from the Shopify ecosystem.

Supporting Data and Platform Mechanics

Understanding how data flows through the new system requires examining the distinct rules governing default publishing, storage caps, and analytical tracking.

What is Published by Default?

Information generated via automated storefront scans remains strictly unpublished until the merchant explicitly greenlights the items. Shopify’s documentation emphasizes that unapproved machine-read data is never fed to AI shopping agents.

Conversely, critical administrative data—such as core store policies and formal contact information—is automatically published the moment it is created in the admin settings. Merchants retain granular control over this data, with the ability to bulk-publish or unpublish entire categories, individual fields, or global datasets simultaneously.

The 1,000-Fact Threshold and Migration Rules

Shopify has established strict data caps for the new native knowledge base. The system enforces a maximum limit of 1,000 total facts per store (combining both AI-generated suggestions and merchant-authored entries).

This cap dictates how data was handled during the migration phase:

• Smaller Retailers: For stores operating within normal parameters, all facts previously entered or edited within the legacy app were automatically migrated and published to the new database, requiring zero manual intervention.
• Large Retailers (Over 1,000 Facts): Merchants managing massive knowledge bases exceeding 1,000 facts experienced a different protocol. Their data was not automatically transferred. Shopify has confirmed that direct email notifications were dispatched to these larger retailers, instructing them to manually curate and re-enter their desired knowledge base items prior to the January 15, 2027 deadline.

Analytics and Query Tracking

The Shopify ecosystem provides visibility into AI agent performance through specialized metrics pages. The app’s metrics dashboard features a detailed query log tracking customer questions submitted to AI shopping assistants, identifying which specific agent handled each inquiry, and noting whether the agent successfully resolved the query.

Additionally, the Managing Agentic Storefronts webpage aggregates macro-level performance metrics, breaking down sales, orders, session counts, and conversion rates by specific AI channels alongside a functional product search preview.

Official Responses and Technical Evolution

As Shopify pivots toward agentic architecture, developer resources indicate a concurrent shift in underlying protocols. The Storefront MCP tool—historically relied upon for handling policies and FAQs—has officially been marked as deprecated, though it remains functional for backward compatibility.

Shopify is actively directing developers to transition toward Ask MCP. Within this new framework, the ask_business tool dynamically processes and answers buyer queries by synthesizing data straight from a store’s active knowledge base and official policies. Shopify’s technical documentation currently categorizes the "Ask" capability as a "draft" feature, warning developers that availability and response reliability may vary across different store configurations.

Implications for Merchants and the Future of AI Search

The integration of the knowledge base directly into the Merchant Admin is more than a mere administrative housekeeping update; it signals a fundamental maturation of conversational commerce.

Closing the Gap on AI Readiness

Prior to this update, fragmented knowledge management often led to inaccurate or generic AI responses. In an audit conducted in August and September by digital marketing strategist Bill Hunt for Search Engine Journal, 122 Shopify stores were rigorously tested for AI agent readiness.

Out of the 88 stores where Hunt queried the automated policy tools regarding order cancellations, only two returned a fully functional, usable answer. Hunt pointed out that because Shopify lacks a dedicated, native knowledge base category specifically for "cancellations," merchants must proactively input missing cancellation terms manually as "Additional Facts."

Hunt’s findings underscore a vital takeaway for e-commerce operators: simply having a store online no longer guarantees that AI shopping assistants can accurately represent business policies. Merchants must actively audit their native admin knowledge bases, test their policy tools using real-world buyer scenarios, and abandon reliance on outdated app-era query logs.

Visibility vs. Accuracy

It is critical for merchants to understand that optimizing the Merchant Admin knowledge base directly impacts response accuracy, rather than raw search visibility. According to legacy documentation carried over into the modern framework, optimizing these knowledge bases ensures that AI-driven shopping assistants accurately communicate store terms once a buyer interacts with them. However, it does not inherently boost how frequently a store’s products appear in organic AI search results or platform rankings.

As conversational search engines, AI agents, and agentic storefronts rapidly become the primary interface for online shopping, Shopify’s centralization of merchant intelligence ensures that businesses retain granular control over their brand narrative. Merchants operating on the platform have until January 15, 2027, to complete their migrations and ensure their AI agents are speaking accurately on behalf of their brand.