What Is a Product Feed in Google Merchant Center?
7 core attributes (id, title, description, link, image_link, availability, price) form the required spine of every Google Merchant Center product data source — and adding correct GTINs drives a 20% average increase in clicks (Google).
What Is a Product Feed in Google Merchant Center?
A product feed — now officially called a data source in Google Merchant Center Next — is the structured connection between your product catalog and Google's Shopping systems. It carries your product data (titles, prices, images, availability, unique identifiers, custom labels, and more) from your catalog into Google, where it powers Shopping ads, free product listings, Performance Max, Demand Gen, and every other merchant program. Without an accurate, well-structured data source, none of those programs run.
The terminology changed in 2023. Google announced Merchant Center Next at Google Marketing Live on May 23, 2023, and completed the migration of all retailers by September 2024. The word "feed" became "data source" throughout the platform. The Feeds tab became Products → Data sources. Feed rules became attribute rules. Every third-party guide written before September 2024 that tells you to "go to your Feeds settings" or "create a new feed" describes a navigation path that no longer exists for most merchants. This guide uses both terms — "product feed" in the title because that is what merchants search for, and "data source" in the body to describe the current Merchant Center Next platform.
At MB Adv Agency, feed quality is the first thing our Shopping team audits on any new account. Data source problems — wrong availability values, missing GTINs, unresolved 2025 spec changes, stale prices — cause more lost impression share than bid strategy errors. Getting the data source right is not a setup task you complete once; it is an ongoing compliance discipline that determines how Google classifies, matches, and surfaces your products. Start with the Google Merchant Center overview if you are new to the platform, then return here for the data-source depth.
Data Sources in Merchant Center Next: Understanding the Rename
When Google rebuilt Merchant Center as Merchant Center Next, it reframed how product data enters the system. The old mental model treated a "feed" as a file — a spreadsheet or XML export you upload. The new model treats a "data source" as a connection — a persistent, configured relationship between your catalog and Google that can use any of five input methods. This is not cosmetic. A supplemental data source overrides specific attributes on existing products without touching the primary source. Attribute rules transform data at the source level without editing the upstream catalog. These concepts do not translate from the old "feeds" model — they are genuinely new architecture.
The naming map:
- Product feed → Primary data source (Products → Data sources in the current UI)
- Supplemental feed → Supplemental data source (requires the Advanced data source management add-on in Settings → Add-ons)
- Feed rules → Attribute rules (formerly "feed rules"; also requires the Advanced add-on)
- Feeds tab → Products → Data sources (the primary nav)
- Programs → Add-ons (Settings → Add-ons)
Merchant Center Next also introduces a new framing for product data origin: "Provided by you" (traditional uploads — file, Sheets, fetch, API) versus "Provided by Google" (automated website crawl, automatic product detection). Google's strategic direction is to reduce merchant dependency on manually maintained feed files by reading product data directly from structured landing pages. The five input methods below reflect both categories. See managing and automating product feeds for the full treatment of attribute rules, supplemental sources, and AI automation tools.
The 5 Data-Source Input Methods
Merchant Center Next supports five ways to connect your product catalog. The right choice depends on your catalog size, how often your data changes, and your technical infrastructure — not on which is the simplest to start with. A 50-SKU boutique and a 50,000-SKU retailer have different requirements; so does a merchant whose prices change every hour versus one whose catalog is stable for months.
| Input method | How it works | Best for | Update frequency |
|---|---|---|---|
| File upload (SFTP or Google Cloud Storage) | You push a data file (TXT, XML, TSV, XLSX) directly to Google via SFTP or a GCS bucket | Mid-to-large catalogs with automated feed generation; merchants with a PIM or ERP that outputs a data file | On-demand — you control the push schedule. Daily or multiple-daily possible |
| Google Sheets | A live-linked Google Spreadsheet (using the MC template) that Google reads on a schedule | Small-to-mid catalogs; merchants who manage products manually; easiest starting point for testing | Google-controlled fetch on a configured schedule; edits flow through on the next fetch |
| Scheduled fetch | Google periodically fetches a data file from a URL you specify on a schedule of at least every 24 hours | CMS-generated catalogs (Shopify native exports, WooCommerce, BigCommerce); merchants who want Google to pull rather than push | Configurable (minimum daily). Source file must regenerate on your server before each fetch |
| Merchant API (v1, GA July 2025) | Programmatic product-level inserts, updates, and deletes via the Merchant API (v1). Content API for Shopping shuts down August 18, 2026 — any existing Content API integration must migrate to Merchant API now | Large or complex catalogs with real-time price and availability changes; engineering teams with custom integration requirements | Real-time on-demand — highest update frequency; suited to flash sales and live inventory |
| Automatic website crawl | Google reads schema.org product structured data from your landing pages and sitemaps — no file upload required. Also functions as a safety net via automatic item updates (price, availability, condition — on by default) | Simple catalogs with existing schema.org markup; merchants who want to start without a technical integration | Google-controlled continuous crawl; 4–8 hours to reflect removed products |
Source: Google Merchant Center Help, "About data sources" and "Add a data source" (support.google.com/merchants/answer/13982673); Google Merchant API latest updates (Merchant API v1 GA July 2025; Content API for Shopping sunset August 18, 2026).
Supplemental data sources — which override specific attributes on products already defined in the primary source — are covered in managing and automating product feeds. They require the Advanced data source management add-on and are a separate layer on top of the primary source discussed here.
Primary vs Supplemental Data Source
Every Merchant Center account starts with a primary data source, which defines all product records in your catalog. A supplemental data source is an optional overlay that overrides or augments specific attributes on those existing products — without editing the primary source. Supplemental sources require the Advanced data source management add-on in Settings → Add-ons and are not visible in the default Merchant Center UI.
| Dimension | Primary data source | Supplemental data source |
|---|---|---|
| Role | Defines and creates products in your catalog. The foundational source — every product must exist here first | Overrides or augments specific attributes (custom labels, title fixes, price overlays) on products already in the primary source |
| Required to start | Yes — you need a primary source to have any products in Merchant Center | No — optional add-on layer; requires the Advanced data source management add-on (Settings → Add-ons) |
| Can add or remove products | Yes — the only source that creates and removes product records | No — references products via id matching; cannot create or delete records |
| Attribute rules | Attribute rules (formerly feed rules) apply at the primary source level — requires Advanced add-on | Attribute rules also apply at the supplemental level — requires Advanced add-on. See attribute rules documentation |
| When you need it | Always — start here | When overriding attributes without editing the primary (e.g., adding custom labels from a third-party tool; applying a pricing overlay; fixing titles without touching the ERP export) |
Source: Google Merchant Center Help, "About primary data sources" and "About supplemental data sources". Supplemental sources and attribute rules are covered in depth in managing and automating product feeds.
Required Attributes: The Core 7
Google's product data specification defines three tiers of attributes: required (every product needs these to be served at all), conditional-required (become required when a condition is met), and optional (additional attributes that improve performance, enable programs, or support campaign segmentation). Most thin guides flatten this to a binary "required vs optional," which misses the conditional tier — where most merchant disapprovals originate.
The seven required attributes are the minimum for any product to be eligible for any serving in Google's merchant programs. Every product in your data source must include all seven. Missing any one of them disqualifies the product from all serving — not just some placements.
| Attribute | What it is | Key spec constraints |
|---|---|---|
id |
Your unique, stable product identifier (SKU or internal ID) | Max 50 characters. Changing id creates a new product record and loses all performance history — keep it stable |
title (or structured_title) |
Product name displayed in listings. Front-load brand + key attributes | Max 150 characters; about 70 characters visible. structured_title required for AI-generated titles |
description (or structured_description) |
Product description; plain text only (no HTML markup). Put the most important information in the first 160–500 characters | Max 5,000 characters. structured_description for AI-generated content |
link |
The landing-page URL where the product is sold. Must match the advertised price and availability exactly | Price or availability mismatches against this URL trigger Preemptive Item Disapproval |
image_link |
Main product image URL. No promotional overlays or watermarks | Current enforced minimum: 100×100 px (non-apparel) / 250×250 px (apparel). Universal 500×500 minimum enforced January 31, 2027 (warnings active since April 14, 2026). Recommended: 1,500×1,500 px or larger |
availability |
Product availability status — exactly four accepted values | in_stock, out_of_stock, preorder, backorder. Must match the landing page in real time. preorder = unreleased products ONLY; temporarily-OOS returning items use backorder |
price |
The publicly-available selling price. Include VAT/GST for EU and non-US markets. For US: Google now auto-calculates sales tax (the tax attribute was removed July 2025) |
Must match the landing page exactly. Loyalty/member prices go in loyalty_program, NOT in this field |
Source: Google Merchant Center Help, "Product data specification" — required attributes section; web-verified 2026-06-30. Missing any of these seven attributes disqualifies the product from all merchant program serving. Disapproval patterns from attribute errors are covered in debugging feed disapprovals and errors.
Conditional-Required Attributes: The Most Misunderstood Tier
Conditional-required attributes are where the most common merchant disapprovals originate. These attributes are not required for every product — but they become required the moment a specific condition is true. Google does not always generate an immediate disapproval when a conditional attribute is missing; it can surface later as a listing restriction or a policy flag. The MB Adv Agency feed management approach treats the entire conditional set as effectively required: submit gtin whenever it exists, always include item_group_id on variants, always set availability_date on preorder and backorder items.
| Attribute | Required when… | What happens without it |
|---|---|---|
brand |
Any new physical product (except books, movies, music — which use GTIN or ISBN instead) | No fallback — product is disapproved or severely limited without it. Also required as the GTIN-fallback partner (brand + mpn) |
gtin |
A manufacturer-assigned GTIN exists for the product. Google's own benchmark: products with correct GTINs receive a 20% average increase in clicks | Missing GTIN warning. Use brand + mpn if no GTIN exists. Set identifier_exists: no only for genuinely identifier-less goods |
item_group_id |
Submitting product variants (same product in different colors, sizes, or materials) | Without it, Google cannot group variants into a single listing with selectable options. Also triggers the apparel set requirement: color, size, gender, age_group |
availability_date |
availability is set to preorder or backorder |
Preorder and backorder listings without an availability_date are disapproved. ISO 8601 format required |
Apparel set: color, size, gender, age_group |
Apparel, footwear, and accessories variants in key markets. Triggered by item_group_id on apparel products |
Apparel variant listings are disapproved or receive severely limited serving without the full set |
certification |
EU/EEA products requiring energy efficiency labeling (since April 2025). The old energy_efficiency_class attributes were deprecated |
EU energy-labeled products (appliances, lightbulbs, tyres) lose the energy-label annotation in Shopping. Switch to certification with authority: EC and the EPREL registration number |
Source: Google Merchant Center Help, "Product data specification" — conditional attributes section; web-verified 2026-06-30. The apparel set (color, size, gender, age_group) applies to apparel variants in key markets including the US, UK, DE, FR, and AU.
Optional Attributes: The High-Impact Set
Google's product data specification includes dozens of optional attributes. The nine below are the highest-leverage for a general ecommerce feed — the attributes that most directly affect categorization accuracy, listing display quality, campaign segmentation, and program eligibility. For fashion merchants running Shopping campaigns with variants, the attribute priorities look different than for a home goods merchant with a stable catalog. The fashion PPC context is particularly dense with attribute requirements: size, color, gender, age group, material, and pattern all apply.
| Attribute | What it does | Why it matters |
|---|---|---|
google_product_category |
Maps your product to Google's product taxonomy | Improves category matching and search relevance. Required for restricted categories (adult, alcohol, healthcare). Google assigns a category automatically if omitted, but your explicit value takes precedence |
product_type |
Your internal product category taxonomy (free-form) | Creates product-type product groups in Shopping and PMax for granular bid segmentation. Submit alongside google_product_category, not instead of it |
additional_image_link |
Up to 10 additional image URLs per product | More images produce richer listing display — lifestyle images go here (main image_link must be product-only on white or neutral background) |
sale_price |
Publicly-available sale or discounted price, shown with a strikethrough of the regular price | Subject to price-history requirements (US Shopping: 5 days in the past 30 OR 15 in the past 200). Not for loyalty or member prices — those go in loyalty_program |
loyalty_program |
Member or loyalty program pricing, points, and shipping benefits | Required for any member-only pricing as of the 2025 price-accuracy policy. Loyalty prices in price or sale_price create a price-accuracy mismatch → Preemptive Item Disapproval. See loyalty_program attribute documentation |
custom_label_0–4 |
Five internal campaign-segmentation labels (invisible to shoppers) | Used only in Google Ads Shopping, PMax, and Demand Gen for bid and budget segmentation. Max 100 characters each; up to 1,000 unique values per label account-wide. See the custom labels section below for the full treatment |
installment (with downpayment sub-attribute) |
Payment plan details for installment-priced products | The downpayment sub-attribute is required for any down-payment amount — encoding a down payment in the base price field violates price-accuracy policy |
auto_pricing_min_price |
Floor price for Google's automated discounts | Prevents automated promotions from discounting below a set price — critical for margin protection when running automated bidding strategies |
mpn (Manufacturer Part Number) |
The manufacturer's assigned part number | Required together with brand as the GTIN fallback when no GTIN was ever assigned. Neither mpn alone nor brand alone is sufficient as a fallback |
Source: Google Merchant Center Help, "Product data specification" — optional attributes; web-verified 2026-06-30. Google's full optional attribute list exceeds 35 attributes; this table foregrounds the highest-impact for a general ecommerce feed. Optimization of titles, descriptions, and images within the data source is covered in best practices for optimizing product feeds.
The Product Data Specification: What It Is and Why It Changes
The Google product data specification is the authoritative document governing every attribute Google accepts in a data source — its name, format, accepted values, character limits, and requirement tier. It is not a static reference. Google adds new attributes, deprecates old ones, and changes requirement tiers without publishing a changelog. A feed built in 2023 without an attribute audit since then is very likely submitting at least one deprecated attribute or violating at least one policy change from 2024–2025. The "unexplained disapprovals" pattern — products that were clean last month and are flagged this month without any feed change — is almost always a spec-change compliance problem.
The spec governs four categories of data:
- Basic product data — the core 7 required attributes (id, title, description, link, image_link, availability, price), brand, and the conditional attributes
- Price and availability — price accuracy rules, sale_price history requirements, loyalty_program for member pricing, installment and downpayment sub-attributes, preorder/backorder availability_date requirements, and the July 2025 US tax removal
- Product identifiers — GTIN formats, mpn, identifier_exists rules, and the Shopping Graph matching consequences of each
- Detailed product attributes — the optional and conditional attributes including apparel set, custom labels, google_product_category, product_type, additional_image_link, certification, and the 2025 EU energy efficiency changes
Feed optimization — how to write better titles, richer descriptions, and higher-quality images within the spec — is a separate layer covered in best practices for optimizing product feeds. Feed compliance and spec changes belong here. Disapprovals and the diagnostic workflow for fixing them are covered in debugging feed disapprovals and errors in Merchant Center.
Brand, GTIN, and Unique Identifiers: The +20% Click Benchmark
Unique identifiers are the highest-leverage data point in any product feed. Google uses GTINs — Global Trade Item Numbers — to match your product to the manufacturer's record in the Shopping Graph. That match unlocks better category classification, richer product descriptions pulled from the manufacturer's catalog, buyer review data from across the Shopping ecosystem, and more competitive placement signals. Google's own published benchmark: "Retailers who've added correct GTINs to their product data have seen a 20% increase in clicks on average" (Google Merchant Center, Tips to optimize your product data). That figure is not a projection — it is Google's observed average across retailers.
Understanding the five GTIN formats, knowing exactly when identifier_exists: no is valid, and knowing the correct fallback when no GTIN exists is the difference between a feed that earns maximum Shopping Graph coverage and one that is matched poorly. The table below decodes all accepted GTIN formats and the identifier_exists rules.
| Identifier / format | Digit count | Primary use | Notes |
|---|---|---|---|
| UPC (Universal Product Code) | 12 digits | US and Canadian retail products; the barcode on most North American product packaging | Submit as 12 digits (preserve leading zeros). A UPC-12 is a subset of EAN-13 — pad with a leading zero to convert |
| EAN (European Article Number) | 13 digits | European and international retail products; the most common format outside North America. Also called GTIN-13 | The UPC-12 is a subset of EAN-13. Three formats share 13 digits: EAN, JAN-13, ISBN-13 — the distinction is geographic and use-case, not digit count |
| JAN (Japanese Article Number) | 8 or 13 digits | Japanese retail products; technically a subset of EAN | GTIN-8 (8-digit) for small or compact packaging; GTIN-13 (13-digit) for standard Japanese retail items |
| ISBN-13 (International Standard Book Number) | 13 digits | Books. Required for books instead of brand |
ISBN-10 must be converted to ISBN-13 before submitting. Up to 10 GTINs per product are accepted |
| ITF-14 | 14 digits | Trade cases and wholesale packaging (multipacks). Less common in direct-to-consumer Shopping feeds but accepted | GTIN-14. Google accepts 0, 8, 12, 13, or 14 digit GTIN values (0-digit = identifier_exists: no declaration) |
| MPN (Manufacturer Part Number) | Alphanumeric (varies) | Fallback identifier when no GTIN is available. Submit with brand together — neither alone is sufficient |
Not a substitute for GTIN when a GTIN exists. Assigned by the manufacturer, not the retailer |
identifier_exists: no |
— | Valid ONLY for: custom-made goods, handmade items, vintage items, antiques, and books published before 1970 — products where no manufacturer-assigned GTIN was ever created | Invalid for branded or mass-produced products even if the GTIN is not on hand. Misuse: warning then disapproval. Find the GTIN on packaging, the manufacturer's website, or the GS1 Global Database |
★ Google's published benchmark: "Retailers who've added correct GTINs to their product data have seen a 20% increase in clicks on average" — source: Google Merchant Center, "Tips to optimize your product data" (answer/7380908). GTIN formats: GTIN attribute documentation (answer/6324461). identifier_exists rules: identifier_exists documentation (answer/6324478). Not MB Adv Agency data — Google's own published figure. [Verify at publish: confirm the +20% stat remains on answer/7380908 and that the identifier_exists valid-use list has not changed.]
The most expensive GTIN mistake: submitting identifier_exists: no for a branded product because the GTIN is not on hand. This is invalid. The GTIN is on the product packaging, the manufacturer's website, or the GS1 Global Database at gs1.org. The correct no-GTIN fallback when a GTIN genuinely does not exist: submit brand + mpn together and set identifier_exists: no only if the product falls into one of the valid categories above.
Custom Labels 0–4: Internal Campaign Segmentation
Custom labels are the most misunderstood data point in the product data spec. They look like product attributes — they go in the feed, next to title, description, and GTIN — but they do not behave like product attributes at all. Custom labels are invisible to shoppers and are not used by Google's matching algorithms. They have no effect on which searches trigger your products, no effect on Google's category assignment, and they never appear in Shopping ads or free listings. They are purely internal tools for segmenting products in Google Ads Shopping campaigns, Performance Max campaigns, and Demand Gen campaigns — so you can set different bids for different product tiers without building separate campaigns.
The five slots (custom_label_0 through custom_label_4) are free-form — Google assigns no required taxonomy. Each holds up to 100 characters and each label supports up to 1,000 unique values per account. Using them well requires alignment between the feed team (who sets the values) and the campaign team (who creates product groups based on those values). See Google Ads custom labels documentation for campaign-side setup.
| Attribute | What it is | Common use case | Example values |
|---|---|---|---|
custom_label_0 |
One of 5 free-form text fields (max 100 characters each). Invisible to shoppers — never displayed in Shopping ads or free listings. Not used by Google's matching algorithms. Used only to create product groups in Google Ads Shopping, Performance Max, and Demand Gen campaigns for bid and budget segmentation. | Margin tier — drives bid strategy; high-margin products warrant higher bids, low-margin products warrant automated bidding or tighter caps | high-margin, low-margin, margin-tier-1 |
custom_label_1 |
Seasonality — flag products that sell well only in certain seasons so you can increase bids in-season and suppress budget out-of-season | summer, holiday, year-round, back-to-school |
|
custom_label_2 |
Sales performance — identify top-selling or high-conversion products to give them more aggressive bids or larger budget allocation | bestseller, top-seller, slow-mover, new-arrival |
|
custom_label_3 |
Price band — group products by price tier to adjust bids based on average order value impact | under-25, 25-100, over-100, premium |
|
custom_label_4 |
Promotional status — flag products in an active promotion or clearance event so you can create a matching campaign or adjust bids during the event window | clearance, black-friday, promo-q4, on-sale |
Source: Google Ads Help, "Use custom labels for Shopping ads" (answer/6275295) — custom labels are internal only, never shown to shoppers, 100-char max per value, 1,000 max unique values per label account-wide, used for product group segmentation in Shopping and PMax campaigns. Demand Gen support for custom labels added 2025. The use-case examples (margin, seasonality, performance, price band, promotional) are standard practitioner taxonomy — not MB Adv Agency client data. Note: custom labels are not the same as google_product_category (which influences Google's matching) or product_type (your internal taxonomy). Only google_product_category and product_type affect how Google categorizes and matches your products.
2025 Product-Data-Spec Changes: The Four Generating Disapprovals
A product feed that was fully compliant in 2023 is very likely violating at least one of these four changes from 2024–2025. Google does not send proactive notifications when the spec changes; feeds that haven't been audited recently accumulate silent compliance drift until it surfaces as a warning or disapproval. All four changes below are confirmed from Google's official documentation and are enforced as of the time of writing.
| Change | Effective | What changed | What to do |
|---|---|---|---|
| EU energy_efficiency_class deprecated | April 2025 | The energy_efficiency_class, min_energy_efficiency_class, and max_energy_efficiency_class attributes were deprecated for EU products — they no longer produce energy-label annotations in EU Shopping results |
Switch to the certification attribute with authority: EC and your product's EPREL registration number. EU energy-labeled products (appliances, lightbulbs, tyres) that still submit the old attributes lose the energy-label display annotation |
Member/loyalty pricing → loyalty_program |
2025 price-accuracy policy update | Member-only prices submitted in price or sale_price violate price-accuracy policy — the feed shows the member price while non-members see the regular price on the landing page, creating a mismatch → Preemptive Item Disapproval |
Submit member and loyalty prices exclusively via the loyalty_program attribute. Keep price and sale_price as the publicly-available prices that non-members see |
Installment down payments → downpayment sub-attribute |
2024–2025 spec update | Down-payment amounts must use the downpayment sub-attribute within the installment attribute — encoding a down payment in the base price field violates price-accuracy policy |
Merchants selling on installment plans: use the installment attribute with its sub-attributes (months, amount, downpayment). Do not modify the base price field to reflect a down-payment amount |
US tax attribute removed |
July 2025 | The tax attribute and all manual US sales-tax settings were removed from the product data spec and the Merchant Center UI. Google now auto-calculates US sales tax by shopper location at search time |
Remove the tax attribute from your US data source submissions. No manual tax setup is required or accepted for US. For EU and non-US markets: prices must still include VAT/GST in the price attribute |
Sources: Google Merchant Center Help, "Product data specification"; loyalty_program attribute documentation. Dates: energy_efficiency_class deprecated "April 2025" (no specific day on official docs); US tax removal confirmed "July 2025" from official sources; loyalty_program requirement confirmed as current enforced policy. The full diagnostic workflow for spec-change disapprovals is in debugging feed disapprovals and errors in Merchant Center. [Verify at publish: confirm the exact effective dates and enforcement behavior against the live spec — Google updates the spec without a changelog.]
Need a clean data source before your next campaign launch?
MB Adv Agency's feed team audits and manages Google Merchant Center data sources for ecommerce merchants — required attributes, GTIN coverage, conditional attribute completeness, 2025 spec compliance, and custom labels wired to your campaign structure.
See Shopping Feed Management →The availability Attribute: Four Values, One Disapproval Trap
The availability attribute accepts exactly four values. No other values are accepted by the spec. The most frequent disapproval pattern comes from merchants who use preorder for any temporarily-unavailable product, rather than reserving it for genuinely unreleased items. preorder is for products that have not yet been released and are available for advance purchase only. Products that are temporarily out of stock but will return use backorder. Both preorder and backorder require availability_date; without it, the listing is disapproved.
| Value | Meaning | Requires availability_date? | Common mistake |
|---|---|---|---|
in_stock |
Product is available to order right now and ships within a normal timeframe. Must match the landing page in real time | No | Keeping this value when the product is actually out of stock — creates a price-accuracy mismatch → Preemptive Item Disapproval |
out_of_stock |
Product is not currently available for purchase and no return date is set or known | No | Using this when the product is temporarily backordered and expected to return — use backorder instead to preserve listing eligibility during the out-of-stock period |
preorder |
Product has not yet been released and is available for advance purchase before the release date. For unreleased products ONLY | Yes — required (ISO 8601 format) | Using preorder for any temporarily-unavailable product — this value is strictly for unreleased items not yet on sale anywhere |
backorder |
Product is temporarily out of stock but is expected to become available again on a known date | Yes — required (ISO 8601 format) | Omitting availability_date when set to backorder — the listing is disapproved without a valid return date |
Source: Google Merchant Center Help, "Product data specification — availability attribute" (support.google.com/merchants/answer/7052112). Exactly four values accepted — no others. Google's automatic item updates check price, availability, and condition on your landing pages continuously (on by default) and flag discrepancies. Keeping your data source in sync with your landing pages is a daily maintenance task, not a one-time setup.
How to Set Up and Audit a Product Data Source
Getting a data source from zero to compliant follows a clear seven-step sequence. The steps below also serve as an audit checklist for an existing feed that is generating unexplained warnings or underperforming in Shopping programs.
- Choose your input method. Match the method to your catalog size and update frequency. A catalog of under 500 SKUs with stable pricing starts well with Google Sheets. A CMS-generated catalog at Shopify or WooCommerce uses scheduled fetch. A large catalog with real-time pricing uses the Merchant API (v1, GA July 2025 — migrate off Content API before August 18, 2026). A catalog with clean schema.org markup on landing pages can start with automatic website crawl.
- Include all 7 required attributes. Every product needs: id (stable, max 50 chars), title (max 150 chars), description (max 5,000 chars, plain text), link (matching the landing page price and availability exactly), image_link (current minimum 100×100 px), availability (one of four values), and price (publicly-available price, excluding loyalty/member prices).
- Add conditional attributes where triggered. brand for all physical products; gtin when a manufacturer-assigned GTIN exists; item_group_id for variants (which also triggers the apparel set for clothing/footwear/accessories); availability_date for preorder and backorder products; certification with EPREL authority for EU energy-labeled products.
- Submit GTINs in the correct format. UPC-12 for North American products, EAN-13 for European/international, JAN-8 or JAN-13 for Japanese, ISBN-13 for books, ITF-14 for trade cases. If no GTIN was ever assigned for the product (custom-made, handmade, vintage, antique, pre-1970 book), submit brand + mpn together and set identifier_exists: no. Never set identifier_exists: no for a branded product because the GTIN is not on hand.
- Wire up custom labels 0–4. Set custom_label_0 through custom_label_4 to values that align with your campaign structure — margin tier, seasonality, sales performance, price band, promotional status. Coordinate these values with your Google Ads team before setting them, since the values in the feed determine which product groups are visible in the campaign structure.
- Audit for 2025 spec changes. Check for the four changes above: EU energy_efficiency_class → certification + EPREL; loyalty prices → loyalty_program; installment down payments → downpayment sub-attribute; remove US tax attribute. Any feed that has not been audited since 2023 is very likely violating at least one.
- Set your feed freshness schedule. Products expire in Merchant Center after 30 days without a refresh. Price and availability must match the landing page in real time. Update at minimum once daily; update multiple times daily during promotions or flash sales to prevent price-accuracy disapprovals from the automatic item updates system.
Getting the attribute layer right — required vs conditional vs optional, correct GTINs, compliant 2025 spec changes, custom labels wired to your campaign structure — is exactly the work our Google Shopping feed management team at MB Adv Agency handles day to day for ecommerce merchants. A well-structured data source is the foundation of a profitable Shopping program. The campaigns, bids, and budgets built on top of it are only as good as the data underneath. For the full ecommerce PPC strategy that runs on top of a clean data source, see our ecommerce PPC agency work. For supplemental data sources, attribute rules, AI automation tools, and platform integrations (Shopify, WooCommerce, BigCommerce), see managing and automating product feeds.
For a complete overview of all programs that draw from your data source — free product listings, Shopping ads, Local Inventory Ads, promotions, and product ratings — see overview of Merchant Center programs.
Frequently Asked Questions: Product Feeds and Data Sources in Google Merchant Center
What is a product feed in Google Merchant Center?
In Google Merchant Center Next, a product feed is called a data source — a structured connection between your product catalog and Google submitted via one of five input methods: file upload (SFTP or Google Cloud Storage), Google Sheets, scheduled fetch, the Merchant API (v1, GA July 2025), or automatic website crawl. Google migrated all retailers to Merchant Center Next by September 2024, renaming "feeds" to "data sources" and the Feeds tab to Products → Data sources. The term "product feed" is still widely used in third-party guides because it is the search term merchants recognize; the current platform term is "data source." Both are in use.
What are the required attributes for Google Merchant Center?
Google's product data specification defines seven required attributes for all products: id (your unique stable product identifier), title (product name, max 150 characters), description (max 5,000 characters, plain text), link (the landing-page URL), image_link (main product image URL), availability (one of four values: in_stock, out_of_stock, preorder, or backorder), and price (the publicly-available selling price matching the landing page exactly). In addition, brand is conditionally required for virtually all new physical products. These seven attributes are the minimum for any product to be served in Shopping ads or free listings. Missing any one of them disqualifies the product from all serving.
Does Google Merchant Center require a GTIN?
Google requires a GTIN when a manufacturer-assigned one exists for your product. Google's own published benchmark: retailers who add correct GTINs see a 20% average increase in clicks (source: Google Merchant Center, Tips to optimize your product data, answer/7380908). If a GTIN exists for your product — and for any branded or mass-produced retail product, one does — submit it. If no GTIN was ever assigned (custom-made, handmade, vintage, antique, pre-1970 book), submit brand and mpn together instead. Do not set identifier_exists: no for a branded product just because the GTIN is not immediately on hand. Find it on the packaging, the manufacturer's website, or at gs1.org.
When can I use identifier_exists: no in Google Merchant Center?
identifier_exists: no is valid only for products where no manufacturer-assigned GTIN, MPN, or brand was ever created: custom goods (handmade, one-of-a-kind), vintage items, antiques, and books published before 1970. It is not a valid shortcut for any branded or mass-produced product for which a GTIN exists but is not on hand. Misusing identifier_exists: no on a branded product results in a warning, then disapproval, and prevents Google from matching your product to the Shopping Graph's manufacturer record — which is exactly what costs you the 20% click uplift. The correct fallback for a genuinely identifier-less product: submit brand plus mpn together and set identifier_exists: no. Neither brand alone nor mpn alone is sufficient as a fallback.
What are custom labels in Google Merchant Center?
Custom labels (custom_label_0 through custom_label_4) are five free-form text attributes that are invisible to shoppers and not used by Google's matching algorithms. They do not influence which searches trigger your products and do not appear in any shopper-facing surface. Custom labels are purely internal tools for segmenting products in Google Ads Shopping campaigns, Performance Max campaigns, and Demand Gen campaigns — so you can apply different bids to high-margin versus low-margin products, exclude seasonal products outside their season, or give bestsellers separate budget allocation. Each label holds up to 100 characters; each label supports up to 1,000 unique values per account. They differ from google_product_category and product_type, which do influence Google's categorization and search matching.
What changed in the 2025 Google Merchant Center product data spec?
Four changes in 2024–2025 are generating disapprovals in feeds not updated since: (1) the EU energy_efficiency_class attribute was deprecated in April 2025 — EU merchants must now use the certification attribute with authority: EC and the product's EPREL registration number instead; (2) member and loyalty prices must use the loyalty_program attribute, not price or sale_price — putting a member-only price in sale_price creates a price-accuracy mismatch against the public price on the landing page; (3) installment down payments require the downpayment sub-attribute within the installment attribute — encoding a down payment in the base price field violates price-accuracy policy; (4) the US tax attribute was removed in July 2025 — Google now auto-calculates US sales tax and the tax attribute no longer exists in the spec for US products. Remove it from US feed submissions.
A clean data source is the foundation — what's built on top of it matters too
If your feed has been live for a while without an attribute audit, the 2025 spec changes (energy_efficiency_class, loyalty_program, downpayment) introduce silent disapprovals before you notice the traffic drop. Our ecommerce PPC audit catches these before they compound. For end-to-end Shopping strategy built on top of a clean feed, see our ecommerce PPC agency work.
Methodology
All attribute names, requirement statuses, GTIN formats, spec-change dates, and the +20% GTIN click benchmark in this guide trace to Google's official product data specification and associated help pages (support.google.com/merchants). The +20% GTIN stat is from Google Merchant Center Help, "Tips to optimize your product data" (answer/7380908), confirmed live 2026-06-30. The GTIN format digit counts are from the GTIN attribute documentation (answer/6324461). The identifier_exists rules are from the identifier_exists attribute documentation (answer/6324478). The Merchant API v1 GA date (July 2025) and Content API for Shopping sunset date (August 18, 2026) are from the Merchant API latest updates changelog (developers.google.com/merchant/api/latest-updates). The 2025 spec changes (energy_efficiency_class EU deprecation April 2025; loyalty_program requirement; downpayment sub-attribute; US tax removal July 2025) are from the product data specification and confirmed from the Google Ads Community and official policy pages.
No fabricated MB Adv Agency client metrics. All MB Adv Agency POV in this guide is qualitative — reflecting feed management experience without invented client benchmark data. No figures are attributed to mbadv that are not in Google's official documentation. No WordStream or LocaliQ figures are used anywhere in this guide — they publish zero Google Shopping attribute or feed data. Last updated: 2026-06-30. [Verify at publish: the product data specification is updated without a changelog; re-verify all attribute requirements, dates, and the GTIN stat against the live spec before publishing.]
In the competitive landscape of e-commerce, visibility and organization are crucial for success. One of the valuable tools that online retailers can leverage is the product feed within Google Merchant Center. This specialized type of data feed plays a significant role in showcasing products on Google, aiding merchants in reaching their target audiences. In this article, we will explore what a product feed is, its definition and purpose, its importance for e-commerce, and the steps required to create a feed effectively.
What is a Product Feed in Google Merchant Center?
A product feed in Google Merchant Center is essentially a structured file that contains all the crucial information about the products being sold by an online retailer. This file typically includes details such as product titles, descriptions, images, prices, availability, and more. By uploading this feed to Google Merchant Center, merchants can make their products eligible for display in Google’s shopping features, including Google Shopping Ads and the shopping tab.
These feeds are generally formatted in either XML or TSV (tab-separated values) format, allowing for seamless interaction with Google’s services. The structure of the feed enables Google to understand not only what products are being offered but also the specifics that might appeal to potential buyers. For instance, including high-quality images and detailed descriptions can significantly enhance the visibility and attractiveness of products in search results, leading to higher click-through rates and conversions.
Types of Product Feeds
Within Google Merchant Center, there are various types of product feeds, including:
- Standard Product Feeds: These are the most common and include all essential product data to ensure accurate representation in Google shopping.
- Supplemental Feeds: These are used to provide additional data for products that may already exist in the standard feed, allowing for further customization and detail.
- Data Feeds: These are often used for specific marketing campaigns or promotions, helping merchants target particular audiences more effectively.
In addition to these primary feed types, merchants can also utilize dynamic remarketing feeds, which allow for personalized ads that showcase products users have previously viewed. This capability not only enhances user experience but also increases the likelihood of conversion by reminding potential customers of items they may have forgotten. Furthermore, optimizing product feeds with relevant keywords and attributes can improve the chances of products appearing in relevant searches, thereby driving more traffic to the retailer's website.
Moreover, keeping product feeds updated is crucial for maintaining accurate listings. Google requires that product information be refreshed regularly to reflect changes in inventory, pricing, and other essential details. Retailers who neglect this aspect may find their products disapproved or underperforming in search results. Therefore, implementing automated solutions or regular audits can help ensure that product feeds remain current and effective in capturing consumer interest.
Definition and Purpose
To understand the importance of a product feed, we first need to define its purpose. The primary goal of a product feed is to convey important product information to Google, allowing it to display these products effectively to users searching for relevant items.
By accurately defining and organizing product details, merchants can enhance their online visibility, ensuring that potential customers encounter their offerings in the crowded marketplace. This clarification is essential for driving traffic and increasing conversions, as users are more likely to click on organized and detailed listings. Furthermore, a well-optimized product feed can lead to improved search rankings, as search engines prioritize high-quality, relevant content. This means that investing time and resources into creating a comprehensive product feed can yield significant returns in terms of customer engagement and sales.
In addition to boosting visibility, product feeds also play a crucial role in maintaining competitive advantage. As e-commerce continues to grow, businesses that leverage accurate and timely product information are better positioned to respond to market trends and consumer demands. By regularly updating their product feeds with the latest information, including seasonal promotions or new arrivals, merchants can ensure that they remain relevant and appealing to their target audience. This proactive approach not only fosters customer loyalty but also helps establish a brand's reputation as a reliable source for quality products.
Key Components of a Product Feed
A well-structured product feed contains several key components:
- Product ID: A unique identifier for each product.
- Title: A clear and concise name that describes the product.
- Description: Detailed information that helps the user understand the product's features.
- Link: The URL leading to the product page on the merchant's site.
- Image Link: A link to the image that represents the product.
- Price: The selling price of the product.
- Availability: Information on whether the product is in stock.
- Brand: The name of the brand associated with the product.
Each of these components serves a specific purpose in the overall effectiveness of the product feed. For instance, the product ID not only helps in tracking inventory but also aids in preventing duplicate listings, which can confuse potential buyers. The title and description work hand-in-hand to capture user interest and provide essential details that can influence purchasing decisions. In addition, high-quality images are vital; they not only enhance the visual appeal of the listing but also help customers make informed choices by providing a clear view of what they are buying. Moreover, including accurate pricing and availability information is critical, as it directly impacts customer satisfaction and trust. If a product is listed as available but is out of stock, it can lead to frustration and lost sales, highlighting the importance of maintaining an up-to-date product feed.
Importance for E-commerce
In today's digital shopping era, an effective product feed is invaluable for e-commerce businesses. First and foremost, it enhances product visibility on major platforms like Google, exposing items to a broader audience than traditional marketing methods. This increased exposure is essential for standing out in a crowded marketplace where consumers are bombarded with choices. By utilizing a well-structured product feed, businesses can ensure that their offerings are not only seen but also attract the right customers who are most likely to make a purchase.
Increased visibility leads to higher chances of clicks and conversions, ultimately resulting in more sales. Additionally, a product feed allows merchants to leverage Google's sophisticated algorithms that relate to user search behavior. When a feed is optimized correctly, products can be shown to users actively searching for similar items, thus facilitating a highly targeted marketing strategy. This targeted approach not only improves conversion rates but also enhances the overall shopping experience for consumers, as they are presented with relevant products that meet their needs.
Boosting SEO and Engagement
Another element of importance is that an accurately maintained product feed can significantly boost a merchant's search engine optimization (SEO) efforts. While the feed is designed primarily for Google Shopping, the additional information can be used in other areas, such as optimizing the product pages on the retailer's website. This cross-utilization of data ensures that all aspects of a merchant's online presence are aligned, creating a cohesive brand image that resonates with potential buyers. Furthermore, the integration of structured data can enhance visibility in organic search results, driving even more traffic to the site.
Moreover, features like Google’s Product Ratings and User Reviews can be enhanced through a well-structured feed. Customer feedback on products can build credibility and trust, making it more likely that users will convert into buyers. Positive reviews not only serve as social proof but can also influence purchasing decisions, as consumers often rely on the experiences of others when making choices. Additionally, incorporating user-generated content into the product feed can foster a sense of community around the brand, encouraging repeat business and customer loyalty. By actively engaging with customers and showcasing their feedback, e-commerce businesses can create a dynamic online environment that promotes interaction and drives sales.
How to Create a Feed
Creating a product feed for Google Merchant Center may seem overwhelming initially, but by following a systematic approach, it can be achieved effectively. Below are the steps involved in creating a successful product feed.
Step 1: Gather Product Information
The first step in creating a feed is to collect all relevant product information. This includes not just the basic details like titles and prices, but also specifications, dimensions, and any other attributes that would be important for potential buyers.
Organizing this information in a spreadsheet can help in formatting it correctly for the feed later on.
Step 2: Choose the Right Format
Next, you'll need to choose the appropriate file format for the feed. Google recommends using either XML or TSV format. For retailers who are less technically inclined, using a spreadsheet program like Google Sheets or Excel to create a TSV file can be highly effective.
Step 3: Upload the Feed to Google Merchant Center
Once the feed is prepared, the next step is uploading it to Google Merchant Center. Navigate to the Feeds section and follow the prompt to input the necessary information about your feed, including the frequency of updates.
Step 4: Regularly Update and Optimize
Finally, to maintain performance and relevance, it is crucial to regularly update the feed. Products may go out of stock or change in price, and keeping your feed current is essential for preventing issues with your listings on Google. Additionally, continuous optimization based on performance metrics will help improve how products perform over time.
In summary, understanding and effectively utilizing a product feed in Google Merchant Center is a cornerstone of a successful e-commerce strategy. By defining its components, recognizing its importance, and following the appropriate steps to create and manage it, retailers can greatly enhance their presence in the online marketplace.
As a Google Ads expert, I bring proven expertise in optimizing advertising campaigns to maximize ROI.
I specialize in sharing advanced strategies and targeted tips to refine Google Ads campaign management.
Committed to staying ahead of the latest trends and algorithms, I ensure that my clients receive cutting-edge solutions.
My passion for digital marketing and my ability to interpret data for strategic insights enable me to offer high-level consulting that aims to exceed expectations.
Google Partner Agency
We're a certified Google Partner Agency, which means we don’t guess — we optimize withGoogle’s full toolkit and insider support.
Your campaigns get pro-level execution, backed by real expertise (not theory).
4.9 out of 5 from 670+ reviews on Fiverr.
That’s not luck, that’s performance.
Highly recommend Matteo to set up your server side tracking. He has a deep understanding of e-commerce tracking and will go above and beyond to make sure everything is set up correctly and working 100%. If you are scaling your store this set up is non-negotiable in my opinion and there isn't many people who have this much knowledge or put the effort in to get it right. Thanks again!
avorodesign.com
I can only recommend Matteo! He was very patient, professional and very knowledgeable about GA4, Consent Mode v2, and GDPR compliance. Communication was clear, and the setup was done professionally and efficiently. Highly recommend him for anyone needing reliable tracking implementation.
www.natureiki.life
Matteo shines in the realm of online professionals. His work is not only deep in data but also complemented by his proactive communication and cooperation, setting a new standard for freelancers. If you want someone who truly exceeds expectations, look no further. Highly recommended!
www.omanbeverlysmyth.com
Exceptional Service Beyond Expectations - Outstanding Service Impeccable depth, flawless delivery, and exceptional language fluency—this service exceeded all expectations. Highly recommended. Matteo truly ROCKS!!!
ium-paris.com
Top-notch, always highly value working with Matteo. An absolute Google Ads Genius. This is approximately the 8th time I have hired him and he's helped us get 6-7 ROAS. We are excited in continuing to improve our lead flow. Hire this guy if you need Google Ads help. Thanks Matteo!
www.dleeventgroup.com
I finally found the guy who can setup server side tracking and all the ecosystem properly. I definitely recommend Matteo. He is very responsive, kind and wants to dig into things. He configured GA4, Meta, Google Ads, Outbrain and google consent v2 with Cookiebot. Thanks Matteo.
inomega.fr
MB Adv delivered exceptional work with outstanding professionalism and lots of patience, taking time to see effects of changes made and not just do the work and submit it. The proactive communication and video summaries of the work completed made working with Matteo a pleasure, as he consistently went above and beyond. Highly recommended for web analytics projects! We are already working on another project.
www.withnellsensors.co.uk
Working with Matteo on my Google Ads was a game-changer. He's not just a strategist, he's a true partner. He understood my goals and tailored a campaign that perfectly reached my target audience. I'm grateful for his expertise and dedication.
dccargo.com
Click-driven mind
with plastic-brick obsession.
We build Google Ads campaigns with the same mindset we use to build tiny brick worlds: strategy, patience, and zero tolerance for wasted pieces.
Data is our blueprint. Growth is the only acceptable outcome.
