UCP and ACP Readiness Checklist for Brands

Executive Summary
Google's Universal Commerce Protocol (UCP) and OpenAI's Agentic Commerce Protocol (ACP) are helping shape a new commerce infrastructure in which product discovery, comparison, and purchasing can take place through AI-powered experiences.

For brands, readiness begins with a connected commerce data foundation. Product feed quality, category and attribute structures, globally recognized product identifiers, price and inventory accuracy, variant relationships, trust information, and checkout processes should be assessed as parts of the same system.

Priority actions for brands:

• Audit the existing product feed for AI shopping readiness
• Enrich product titles, descriptions, and attribute data
• Standardize category, color, size, material, and measurement values
• Introduce more frequent updates for price, inventory, and promotion data
• Keep internal product and variant IDs consistent across systems, while validating GTINs, MPNs, and brand data
• Structure shipping, returns, warranty, seller, and customer support information
• Verify consistency between the product feed and cart, payment, and order systems
• Include product-level regulatory notices and compliance information in data flows
• Build a product data infrastructure that can be adapted for Google, ChatGPT, and other channels

Preparing for UCP and ACP is a cross-functional commerce initiative involving ecommerce, marketing, merchandising, engineering, customer experience, and data operations.

AI-powered shopping is creating a new product discovery channel while bringing product data, inventory, checkout, returns, and customer experience into the same decision journey.

On Google's side, the Universal Commerce Protocol, or UCP, is designed to make shopping and purchasing more direct across AI-powered Google experiences.

On OpenAI's side, the Agentic Commerce Protocol, or ACP, together with ChatGPT shopping experiences, provides a framework for products to be understood and recommended by AI assistants and, where supported, connected to a purchasing journey.

Although these approaches belong to different ecosystems, they deliver a clear common message for brands:

Product data must be clean, current, enriched, structured, and understandable by machines.

Why Do UCP and ACP Matter for Brands?

In an AI shopping experience, consumers may discover products by speaking with an AI assistant instead of navigating traditional search results.

For example, a shopper might ask:

"Can you recommend a quiet, energy-efficient air purifier that works well in a small apartment?"

This request contains much more than a product name. It communicates the shopper's need, intended environment, priorities, and decision criteria at the same time.

For an AI system to recommend suitable products, it must be able to understand the brand's product data correctly.

Frameworks such as UCP and ACP can help brands share commerce information with AI systems in a more structured way and prepare their operations for emerging shopping experiences.

Technical integration with UCP or ACP is not enough on its own. If the underlying product data is incomplete, inconsistent, or outdated, products may still be represented incorrectly even when the technical connection works.

What Is the Main Difference Between UCP and ACP?

UCP and ACP address a similar need: creating a more structured commerce connection between brands and AI systems.

However, they focus on different ecosystems and require different preparation steps.

The strategic implication is broader than choosing between two protocols: brands need a reusable commerce data layer that can support multiple AI ecosystems without rebuilding product information for every new channel.

Where Should Brands Begin?

When preparing for UCP and ACP, the first question should not be, "Which API should we integrate with?"

The first question should be:

Is our product data clean and structured enough for AI systems to understand it correctly?

In AI shopping experiences, product data is not used only to list products. It also helps systems classify products, match them with shopper needs, compare them with alternatives, verify price and availability, and display relevant trust signals.

Brands should therefore approach readiness through seven key areas.

1. Assess Your Product Feed Structure

One of the most important steps in UCP and ACP readiness is evaluating the current state of the product feed.

Brands should begin by answering the following questions:

• Which systems and data sources supply the product feed?
• Are internal product and variant IDs stable and consistent across all systems?
• Are the GTINs shown in product barcodes accurate for every product and variant that has one?
• Where a GTIN is unavailable, are the correct manufacturer-assigned MPN and brand values provided?
• Are title, description, category, brand, price, availability, and image fields complete?
• Are product variants grouped correctly?
• Are products represented through accurate categories and attributes?
• Is the feed updated frequently enough?
• Can the business manage channel-specific feed outputs?

In AI shopping, it is not enough for a feed simply to work.

The feed must also be understandable, comparable, and reliable.

For example, a product title such as "Basic T-Shirt" may be technically valid, but it is often not descriptive enough for an AI system. Fabric, fit, intended audience, use case, season, color, and available sizes can all help the system interpret the product more accurately.

A product feed audit should therefore be treated not only as a technical error check, but also as an AI readiness assessment.

2. Strengthen Product Identity Data

Before an AI system can recommend a product accurately, it first needs to understand what the product is.

This requires strong product identity data.

The most important fields include:

• Internal product or item ID
• GTIN, including formats such as UPC, EAN, JAN, or ISBN
• Manufacturer Part Number (MPN)
• Product title
• Product description
• Brand
• Category
• Product type
• Color
• Size
• Material
• Dimensions
• Capacity
• Variant information
• Parent and child product relationships

These identifiers serve different purposes. An internal product ID or SKU connects the product across the merchant's own feed, inventory, analytics, and checkout systems. A valid GTIN identifies a trade item globally and can help platforms match the same product across different sellers. When a product has no assigned GTIN, the correct combination of manufacturer-assigned MPN and brand can provide an alternative identity signal.

Identifiers should never be guessed or replaced with a store-specific SKU. They must also be assigned at the correct variant level, since different colors, sizes, or configurations may have their own GTINs or MPNs.

Product titles and descriptions should not be short text fields written only for SEO. They should clearly communicate the product's intended use, distinctive features, and customer value.

For example, "backpack" is too broad to describe a product effectively.

"Black waterproof city backpack with a 13-inch laptop compartment" gives an AI system much more useful context.

This richer structure makes it easier to understand which customer need the product can meet.

3. Standardize Categories and Attributes

Products must be comparable within AI shopping experiences.

Standardized category and attribute structures are essential for making that comparison possible.

When similar characteristics are expressed differently across products in the same category, AI systems may struggle to interpret them consistently.

For example, the material field might contain "cotton" for one product, "100% cotton" for another, and "cotton fabric" for a third. Each value may be technically accurate, but the data is not standardized.

This inconsistency can make reliable product comparison more difficult.

Brands should standardize:

• Category names
• Attribute names
• Attribute values
• Color naming
• Size systems
• Material information
• Measurement formats
• Variant relationships

This work supports not only AI shopping readiness, but also feed performance across Google, Meta, TikTok, marketplaces, and advertising platforms.

4. Keep Price, Inventory, and Promotion Data Current

The product information shown in an AI shopping experience must reflect current commercial conditions.

An incorrect price, an unavailable product, or an expired promotion can directly damage the customer experience.

Dynamic data should therefore receive particular attention when preparing for UCP and ACP.

Priority fields include:

• Current price
• Sale price
• Promotion start and end dates
• Availability
• Pre-order status
• Delivery availability
• Currency
• Country- or region-specific pricing

If an AI assistant recommends a product and the shopper later discovers that it is out of stock, the issue is more than a feed error. It becomes a customer experience problem that can reduce trust in the brand.

Brands should therefore review how frequently price and inventory data is updated.

A daily feed update may be sufficient for some businesses. Categories with fast-changing inventory, however, may require intraday updates or API-based synchronization.

5. Prepare Returns, Shipping, and Customer Support Information

Most brands begin their UCP and ACP preparations with product data, which is the right place to start. However, operational information and trust signals are also becoming increasingly important.

In an AI shopping experience, a shopper will not ask only, "Which product should I buy?"

They may also ask:

• Can I return it easily?
• How quickly will it be delivered?
• Is the seller trustworthy?
• Does it include a warranty?
• How can I contact customer support?
• How do returns work if I buy it as a gift?

Brands should therefore structure and maintain:

• Return policy
• Return window
• Return costs
• Shipping time
• Delivery options
• Seller information
• Customer support information
• Warranty information
• Post-purchase communication channels

For Google UCP, return policy and customer support information in Merchant Center are particularly important because they help establish confidence throughout checkout and post-purchase experiences.

For OpenAI and ChatGPT experiences, making returns, shipping, seller, and other trust signals available in a structured form can also support more accurate product evaluation.

6. Evaluate Your Checkout and Purchasing Flow

Product discovery is only the first stage. Brands must also prepare for the shopper's purchasing journey.

Brands should therefore assess their checkout infrastructure as well.

Key questions include:

• Do product IDs match the IDs used in the checkout system?
• Do the cart and payment flows operate consistently?
• Are price, inventory, tax, and shipping calculated correctly at checkout?
• Does the experience adapt correctly for different countries and currencies?
• Can order status, cancellation, and return information be tracked across systems?
• Is the engineering team ready for API-based checkout integrations?

On the Google UCP side, checkout integration is intended to connect shopping journeys more directly with Google's AI-powered surfaces. Consistency between product data and the checkout system therefore becomes critical.

As product discovery, evaluation, and purchasing experiences evolve on OpenAI's side, brands must also maintain a strong merchant-owned checkout experience on their own website or application.

In short, AI shopping readiness does not end with the feed file. Product data, inventory systems, cart, payment, shipping, and post-purchase operations must be assessed as one connected journey.

7. Include Regulatory and Compliance Data in the Feed

Some product categories require legal notices, restrictions, or compliance information to be shown to customers.

This can be especially important for:

• Health and personal care products
• Electronics
• Baby and children's products
• Food products
• Home and safety products
• Age-restricted products
• Products subject to regional regulation

In AI shopping experiences, these notices need to appear in the correct place and format.

Brands should assess whether they can provide the following at product level:

• Safety warnings
• Legal disclosures
• Usage restrictions
• Age or regional restrictions
• Certification information
• Warranty terms
• Product compliance information

When this information is missing, a product may be technically listable but still encounter eligibility or compliance issues during recommendation or checkout.

UCP and ACP Readiness Checklist

The practical readiness checklist for brands can be summarized as follows:

Product Feed Readiness

• Are internal product and variant IDs stable and consistent across systems?
• Are valid GTINs supplied for products and variants that have them?
• When no GTIN is assigned, are the correct MPN and brand values available?
• Do titles clearly explain what each product is?
• Do descriptions include the intended use and distinguishing features?
• Is category information accurate?
• Are attribute fields complete and standardized?
• Are variants grouped correctly?
• Are the images sufficient and matched to the correct products?
• Is the feed update frequency appropriate for the product category?

Dynamic Data Readiness

• Is pricing current?
• Are sale prices represented correctly?
• Are promotion dates transferred accurately to the feed?
• Does availability reflect actual inventory?
• Are pre-order and replenishment details managed correctly?
• Can country, region, and currency differences be managed?

Trust and Operational Information

• Is the return policy clear and current?
• Are shipping times and delivery options clearly defined?
• Is seller information accurate?
• Is customer support information available?
• Can warranty information be supplied at product level?
• Can review and rating data be used?

Checkout Readiness

• Are product feed IDs compatible with the checkout system?
• Is the cart flow consistent with product, price, availability, and shipping data?
• Can order status, cancellation, and return processes be tracked?
• Is the engineering team prepared for API-based integrations?
• Have payment and security processes been assessed for AI-powered commerce flows?

Compliance and Regulation

• Are product-level legal notices defined?
• Have category-level restrictions been identified?
• Can safety, age, regional, and certification information be included in the feed?
• Can mandatory consumer disclosures be delivered in the required format?

Which Teams Should Own This Preparation?

Effective preparation requires collaboration across several functions.

The ecommerce team owns the product catalog, category structure, pricing, availability, and operational processes.

The performance marketing team monitors how the feed is used across Google, Meta, TikTok, and other advertising platforms, as well as its effect on performance.

Product or merchandising teams contribute to attribute quality, product descriptions, and category accuracy.

The engineering team manages APIs, checkout, inventory synchronization, data transfers, and system integrations.

The customer experience team ensures that returns, shipping, support, and post-purchase information are presented accurately.

The feed management or data operations team is responsible for delivering this information to each platform in the required format.

Together, these teams should manage AI shopping readiness as a shared product data and commerce program with clear ownership for data quality, synchronization, and customer experience.

How Can Optifeed Support Brands in This Process?

Optifeed helps brands adapt their product data to the requirements of different platforms.

In the context of UCP and ACP readiness, Optifeed's role is to make product data cleaner, richer, more current, and compatible with the channels where it will be used.

Optifeed can support brands with:

• Product feed analysis
• Detection of missing or inaccurate product data
• Validation of product IDs, GTINs, MPNs, and brand data
• Category and attribute optimization
• AI Enrich improvements for product titles and descriptions
• Product variant structure optimization
• Price and inventory synchronization
• Channel-specific feed adaptation
• Feed preparation for Google, Meta, TikTok, marketplaces, and AI platforms
• Product data enrichment for AI shopping experiences
• Reduction of feed-related visibility and eligibility issues

This approach helps brands improve the quality of the product data they already use across advertising platforms while also preparing for UCP, ACP, and emerging AI-powered shopping experiences.

Conclusion: UCP and ACP Readiness Starts with Product Data

UCP and ACP are among the developments signaling a new phase for ecommerce brands.

In this environment, having products visible on a website or advertising platform will not be enough. Products must also be understood correctly by AI systems, matched with relevant shopper needs, and presented with accurate and trustworthy information.

A clean product feed, reliable GTIN and MPN data, standardized categories and attributes, current price and inventory information, accurate variant relationships, clear returns and shipping details, and checkout consistency are the foundations of this preparation.

The objective for brands can be summarized simply:

Make your product data understandable, comparable, and trustworthy for AI systems.

Optifeed helps brands manage this process in a more controlled, scalable, and platform-compatible way.