How to Optimise Your Google Shopping Feed for 2026
Your product feed is the single biggest factor separating high-performing Google Shopping campaigns from underperforming ones. Before bidding strategies, audience signals, or campaign structures come into play, the feed has to be right. At Shoparize, we manage Shopping campaigns for 25,000+ merchants across 21 European countries. The single biggest pattern we see separating top performers from the rest is feed quality.
This guide covers the attributes that matter most, the changes Google has introduced for 2026, and what to do about them this quarter. It is written for PPC Managers, Affiliate Managers, and E-commerce leads running Shopping campaigns in European markets. If you want the full checklist with chapter-by-chapter detail, download the Product Feed Fundamentals eBook, free, no email required for the public version.
The feed attributes that matter most
Four attribute groups carry the most weight in Google Shopping performance. Get these right and the rest of your feed falls into place more easily.
Titles are the single most impactful attribute in your feed. Google weighs them heavily when matching products to queries, and shoppers use them to decide whether to click. The first 70 characters are what most shoppers actually see, so front-load brand, product type, and your core differentiator there. A title like “Shoes” loses to “Nike Air Max 90 Running Shoes Black Men’s Size 44” every time. Start with your top 50 products by revenue, rewrite those titles, and you will see click-through rates move within days.
Category-specific formulas help. For apparel: brand + gender + product type + colour + size + material. For electronics: brand + product + model number + key specs. Use digits instead of words to save character space, stay under 150 characters total, and avoid promotional language like “Free Shipping” or “Best Seller.”
Product identifiers (GTIN, MPN, and brand name) allow Google to classify your products accurately. Products without identifiers will not appear on high-performing surfaces like the Popular Products grid and AI Overviews, and will not show product ratings in Shopping results. Incorrect identifiers can trigger disapprovals. If your feed has gaps, fix identifiers first. Submit GTIN + MPN + brand for all manufactured products that have GTINs. For products without them, submit brand + MPN and set identifier_exists to “no.”
Descriptions add the context Google needs to understand what your product actually is. Use the full 5,000-character limit where it adds value. Focus on answering the questions a shop assistant would answer: what is the product, what is it made of, who is it for, what problem does it solve? With AI Mode now interpreting conversational queries that are 2 to 3x longer than traditional searches, descriptions carry more weight than they did in keyword-matching auctions. Sparse descriptions give the AI less to work with when deciding whether to recommend your product.
Custom labels let you tag products by margin, price bucket, or bestseller status for campaign segmentation. If you are not using custom labels yet, start with one. Even a simple price-bucket label gives you meaningful bidding control that flat campaign structures cannot match.
What changed in 2026
AI Mode, launched in January 2026, interprets conversational, longer-form queries and matches them against product attributes. Instead of “running shoes men,” shoppers are now searching for things like “lightweight running shoes for flat feet that work on trails and pavement.” Google’s Agentic Commerce framework organises this shift around three pillars: be understood, be attractive, and be present. The common thread is that product data quality has moved from “good enough to win auctions” to “good enough for AI to recommend you.”
Google is also rolling out new Merchant Center attributes designed specifically for AI discovery: product_use_cases, care_instructions, compatible_with, product_faq, and compatible_accessories. These go beyond traditional keywords and help AI agents reason about your products the way a knowledgeable salesperson would. Target 95%+ attribute completion on your top products, and audit coverage for conversational query fields like material, use case, compatibility, and dimensions.
Universal Commerce Protocol (UCP) is Google’s open standard that allows AI agents to treat a merchant’s store as a programmable service, enabling direct purchases within the AI interface. It is currently a US pilot, but it signals the direction for European markets. The on-switch is the native_commerce attribute in your feed. When UCP becomes available in your market, merchants with complete, structured product data will be ready to participate. Those without it will not.
The Merchant API migration has a hard deadline: August 2026. Google’s Content API for Shopping is being replaced by the Merchant API. If your feed management runs through the Content API, whether directly or via a platform connector, confirm with your development team or provider that migration is planned. If you use a managed feed service, verify that your provider has already completed the switch.
→ Get the full Product Feed Fundamentals eBook here: Download the eBook
Where a CSS partner fits in
A well-optimised feed gives Google the data it needs to match your products to the right queries. How broadly your products compete in the auction determines how much of that optimisation translates into sales.
Google Shopping runs a second-price auction and deduplicates bids at the merchant level, not the CSS level. When a CSS partner like Shoparize bids on your products alongside your own campaigns, Google recognises both bids as coming from the same merchant. Only the highest bid enters the auction against competing merchants. Your CPC is determined by those competitors, never by your own CSS partner. In practice, this means a CSS partner captures auction opportunities your existing campaigns miss: long-tail queries, product variants not getting sufficient budget, and time windows when automated bidding pulls back but demand remains.
Shoparize Managed Ads operates on a pay-per-sale basis. We invest the ad budget and absorb the click costs. If our campaigns produce no conversions, you pay nothing. With Google Shopping CPCs increasing 20 to 30% year-over-year across most European markets, a CPA-based model also serves as a structural hedge against CPC inflation: you shift a portion of your Shopping investment from cost-per-click to cost-per-sale, scaling visibility without proportionally scaling cost risk.
Your feed quality determines your CSS partner’s performance. The same attributes that help your own campaigns help a CSS partner perform: accurate titles, complete identifiers, rich descriptions, correct categorisation, and clean images.
What to do this quarter
- Audit your highest-revenue products. Rewrite titles to front-load brand + product type + differentiator in the first 70 characters. Fill GTIN and MPN gaps. Expand descriptions to answer real shopper questions, especially the conversational queries AI Mode is now interpreting.
- Prepare for the Merchant API migration. Confirm whether your feed runs on the Content API or the Merchant API. If Content API, plan the switch before August 2026.
- Review your attribute completion rate. Target 95%+ on your top products, with particular attention to the new AI-discovery fields: product_use_cases, care_instructions, compatible_with, and product_faq.
Download the full Product Feed Fundamentals eBook, free. It covers every feed attribute in detail, including images, price and availability, categorisation, and the complete AI-discovery checklist.
