Running a Google Shopping campaign is not the same as running a search campaign. In Search, you write an ad and bid on keywords. In Shopping, Google reads your product data feed and decides for you — which products to show, to whom, at what frequency, and at what effective cost. Your feed is your ad creative, your targeting brief, and your Quality Score input all rolled into one.
That means bad data is expensive. Not in the abstract "you're leaving money on the table" way. In the concrete, measurable way: products disappear from the grid, impression share collapses, and competitors with cleaner catalogs capture the same customers at lower CPCs.
This guide covers everything that changed in Google Shopping feed requirements for 2026, why GTIN data is now one of the highest-leverage inputs in your entire Shopping strategy, and how to build a feed that earns Google's trust at scale.
What Actually Changed in Google Shopping Feeds for 2026
Google rolls out specification changes quietly — a help center update here, a policy page revision there — so many sellers find out about new requirements only after their products start getting disapproved.
For 2026, the significant changes are:
Product ID separation is now mandatory. Google split its product identifier requirements for online and physical channels. If you operate both an online store and a physical retail presence, your product IDs must be distinct by channel. Feeds that don't comply see products disapproved at the channel level.
Shipping attributes expanded. As of April 14, 2026, Google added product-level support for handling cutoff time, minimum order value, and loyalty program labels. These aren't universally required yet, but sellers in competitive categories are already using them to win Premium Shopping placements.
Minimum image resolution raised. The previous 300×300 pixel floor is gone. Google now requires a minimum of 500×500 pixels across all product categories and marketing formats. The practical recommendation is 1500×1500 or higher — products with high-resolution images consistently outperform on CTR, which feeds back into better Quality Scores and lower effective CPCs.
Video link attribute is live. Starting June 30, 2026, sellers can submit product videos directly through the feed using the video_link attribute. This is currently optional but early data from Shopping agencies suggests video-enabled listings see significantly higher engagement rates in the Shopping tab.
Category-specific mandatory attributes are expanding. Fashion and apparel now require gender, age_group, material, and size_type in most markets. Health and beauty categories have standardized attribute requirements. Furniture has enhanced specification requirements. Google is systematically moving toward richer, more standardized attribute requirements across all verticals.
The Core Feed Structure: What Google Still Requires for Every Product
Before the 2026-specific changes, the foundation is the same as it has always been — get the basics right or nothing else matters.
Every product in your feed needs: id, title, description, link, image_link, price, availability, brand, and condition. These are non-negotiable. Missing any of them means product disapproval, not just a warning.
Beyond the required attributes, Google uses a collection of "strongly recommended" attributes that function as de facto requirements in competitive categories:
gtin— The global trade item number (barcode). More on this in the next section.mpn— Manufacturer part number, especially important in automotive, electronics, and toolsproduct_type— Your own category taxonomy, used alongsidegoogle_product_categorycolor,size,material— Critical for variants; feeds without these see poor variant coverage in Shoppingshipping— Accurate shipping speed and cost data directly affects click-through rates
The most common mistake isn't missing the required attributes — most sellers get those right. It's the "strongly recommended" attributes that create the performance gap between a functional feed and a high-performing one.
GTINs: The Barcode Advantage in Google Shopping
Of all the attributes in your product feed, GTIN has the clearest, most quantifiable impact on performance.
Products submitted with valid GTINs receive up to 40% more impressions than equivalent products without them. They perform 20–40% better in Shopping auctions, meaning they achieve the same position at effectively lower CPCs. This isn't an estimate — it's the measured gap between products that Google can confidently identify and those it cannot.
Here's why that matters structurally: Google Shopping's AI system classifies products for relevance, cross-references pricing signals across the ecosystem, and applies its category knowledge to decide where a product fits. When you supply a valid GTIN, you're giving Google a direct link to its own product knowledge graph. It knows the product. It knows the category. It knows the competitive set. It serves the product with confidence.
When you don't supply a GTIN — or supply an invalid one — Google is working blind. It has to infer the product from your title and description alone, which introduces ambiguity. Ambiguous products get lower impression share.
GTIN format requirements are strict. Google accepts:
- GTIN-12 (UPC): 12 digits, primary format in North America
- GTIN-13 (EAN): 13 digits, international standard
- GTIN-14: 14 digits, typically for cases and multi-packs
- ISBN-13: For books
Each format includes a check digit that Google validates algorithmically. Submitting a GTIN that fails the check digit is treated the same as missing the attribute entirely. There is no partial credit.
Three ranges to avoid: GTINs beginning with prefixes 02, 04, or 2 are restricted ranges. Prefixes 05, 98, and 99 are coupon ranges. Both will be rejected by Google Merchant Center as invalid identifiers.
Each variant needs its own GTIN. A blue and a red version of the same t-shirt are different products with different GTINs. Submitting the same GTIN across multiple size or color variants triggers duplicate errors that can suppress your entire product group, not just the duplicated variant.
When identifier_exists: false is the right call. Handmade goods, vintage items, store-brand products, and products manufactured in-house without a GS1-registered barcode can legitimately omit GTINs. In these cases, use identifier_exists: false explicitly rather than leaving the field blank. A blank field is an error. A declared absence is a valid state.
For sellers with large catalogs — think 500 to 50,000 SKUs — sourcing accurate GTIN data at scale is where most feeds break down. Supplier-provided spreadsheets frequently contain incorrect check digits, missing GTINs for older products, or GTINs that were recycled when product lines changed. Running each barcode through GS1 validation before submitting the feed is essential, not optional.
Image Requirements That Keep Products Live
According to Google's own data, approximately 65% of product disapprovals are image-related. This makes images the single most fixable source of suppression in most catalogs.
The 2026 image standards:
- Minimum resolution: 500×500 pixels for all categories (raised from 300×300)
- Recommended resolution: 1500×1500 pixels or higher for competitive categories
- Product frame occupancy: The product must occupy at least 75% of the image frame
- Background for general products: White or neutral, plain background required
- Background for apparel: Lifestyle images with models wearing the product are acceptable and often outperform white-background shots on CTR
- Overlays prohibited: No promotional text, price badges, "sale" flags, or watermarks of any kind
The 75% frame coverage rule catches many sellers off guard. Manufacturer-supplied images sometimes feature a product centered in a large amount of negative space — a frame where the product occupies only 40–50% of the image area. These get flagged in the Diagnostics tab and eventually disapproved.
The overlay prohibition is absolute. Google does not distinguish between a tasteful brand watermark in the corner and a large "50% OFF" badge. Both trigger disapproval. Any text or logo overlay on a primary image will eventually be caught, either on initial submission or during periodic re-crawls.
AI-generated images now require disclosure. As of early 2025, products using AI-generated images must include appropriate IPTC metadata labeling. This requirement is being actively enforced in 2026 reviews. If your catalog uses AI-generated product photography, verify your image metadata workflow includes the correct labeling before submitting.
The Most Common Feed Errors and How to Fix Them
Google Merchant Center categorizes errors on a three-tier scale. Red errors cause immediate product disapproval. Yellow warnings reduce performance but don't necessarily suppress the product. Blue informational items are optimization suggestions.
The errors that cost sellers the most aren't always the most dramatic:
Missing or invalid GTIN. Products that have a manufacturer-assigned GTIN but don't include it in the feed are treated as having a data quality issue. Google cross-references its product knowledge graph — if it knows a GTIN exists for your product and you haven't provided it, the product is flagged even if no explicit error is shown.
Price mismatch. The price in your feed must match the price on the landing page at crawl time. Dynamic pricing, flash sales that run faster than your feed refresh cycle, and currency conversion errors all create mismatches. Google's crawler checks prices independently and suppresses products where the landing page price differs from the submitted price.
Broken or inaccessible landing pages. Every product URL is verified by Google's crawler. Pages behind login walls, pages that redirect differently on mobile than desktop, or pages with excessive load times create "landing page not accessible" errors. These account for a significant share of suppressions in large catalogs.
Mismatched domain. Your product links must use the domain verified in Merchant Center. Subdomains, CDN domains, and URL shorteners that resolve to a different domain than the verified one trigger this error across every affected product.
Title keyword over-optimization. This isn't technically an error, but it affects Quality Score. Titles that repeat keywords — "Blue Widget 12-Pack Wholesale Cheap Blue Widget UPC 123456789" — perform worse than titles that accurately describe the product in natural language. Google's title parsing now actively penalizes artificial repetition.
Most of these errors share a root cause: data that looked correct in a spreadsheet but wasn't validated against what Google's crawler actually sees. The fix is consistent: validate feed data against live page data before submitting, not after you see disapprovals.
Building a Clean Feed at Scale
For catalogs with dozens of SKUs, manual data entry is feasible. For catalogs with hundreds or thousands — particularly catalogs that span multiple suppliers, product categories, or retail channels — feed quality is fundamentally a data enrichment problem.
The challenge is that barcodes are the anchor. A valid GTIN, when submitted against a product data API, returns a complete product record: accurate title, brand, description, category hierarchy, weight, dimensions, and image metadata — everything Merchant Center needs. The GTIN is the key that unlocks the rest of the data.
The gap most sellers discover at scale: supplier-provided data is wrong often enough to cause systematic problems. GTINs have incorrect check digits. Brand names vary ("Nike" vs "Nike Inc" vs "NIKE"). Descriptions are either missing or raw warehouse jargon. Images don't meet the 500×500 threshold or have gradient backgrounds where Google requires white.
An enrichment pipeline that validates GTINs first — running each barcode through check-digit verification before querying product data — dramatically reduces the rate of feed errors. From there, standardizing brand names against a reference set, normalizing category taxonomy to Google's required google_product_category values, and auditing image dimensions as a final pass produces a feed that Google processes with confidence.
The operational question is whether this runs manually per batch or continuously as catalog items change. For high-velocity catalogs — particularly in FBA or multichannel retail — real-time feed updates are increasingly the standard. Google's feed update frequency expectations have shifted: daily minimum cycles are the floor, with high-volume sellers expected to maintain near-real-time sync through the Content API rather than file-based uploads.
What a High-Performing Feed Looks Like in 2026
A Google Shopping feed that performs well in 2026 has five characteristics:
Complete GTIN coverage for all products with manufacturer-assigned barcodes — valid check digits, correct variant assignment, no reused GTINs across different products.
Images at 1500×1500 or higher, white background for general products, product occupying 75% or more of the frame, zero overlays.
Feed refresh rate matched to catalog change velocity — daily minimum, real-time for dynamic pricing catalogs.
Product IDs cleanly separated by channel for sellers operating both online and physical retail.
Category-specific mandatory attributes populated for the product's vertical: size, gender, and material for apparel; enhanced specifications for furniture; standardized attributes for health and beauty.
The businesses that invest in feed data quality aren't just avoiding disapprovals. They're buying down their effective CPC — achieving the same ad positions at lower bids, or better positions at the same bids, by improving Google's confidence in their product data. At any meaningful ad spend, the ROI on a clean feed is one of the most measurable numbers in ecommerce marketing.
The right starting point is always the barcode. Valid GTINs unlock everything downstream: better impressions, better auction performance, and a feed that Google doesn't have to guess at.
SKU Monster provides barcode-to-product-data enrichment for ecommerce sellers building Google Shopping feeds at scale. Our API returns validated GTINs, standardized titles, brand data, and category taxonomy for UPC and EAN barcodes — the structured inputs your Merchant Center feed actually needs.
Ready to Try SKU Monster?
If you're managing product data at scale — whether you're on Amazon, Shopify, eBay, or WooCommerce — SKU Monster gives you structured titles, descriptions, images, and pricing for any EAN, UPC, or ASIN in seconds.
No manual entry. No scraping. Just clean product data via API.