Summarize this article with:
Most Google Ads specialists spend their time on campaign settings, bid strategies, and ad creative. Yet very few focus on the thing that determines whether any of those settings can actually work. The product feed.
Since 74 to 97% of Performance Max spend in ecommerce goes to Shopping placements, the feed is the primary creative asset in most accounts. Google uses product titles as keyword targeting signals. It uses GTINs to identify products and unlock premium placements. It uses images in Shopping auctions. When the feed is weak, every campaign built on top of it inherits that weakness. But better bids won’t fix generic titles. Higher budgets won’t fix missing GTINs. And more aggressive tROAS targets won’t fix a catalog where 15% of SKUs are disapproved.
I’ve rebuilt product feeds for 500+ brands over 12 years. Here’s the checklist.
What product feed optimization actually means
Product feed optimization is the process of improving the structured data your products send to Google Merchant Center so that Google can match them more accurately to search queries, approve them for more placements, and surface them in more relevant auctions.
First, the distinction that matters: a compliant feed meets Google’s minimum requirements. An optimized feed treats every attribute as a performance signal.
Google’s algorithm doesn’t read your product pages the way a customer does. It reads structured data fields. The title field is what Google primarily matches to search queries, not your product description, not your page content, not your site structure. A title that reads “Blue Hoodie” will match entirely different queries than “Nike Club Fleece Men’s Pullover Hoodie Blue Large.” The second title includes brand, product type, cut, material, gender, color, and size: seven attributes that dramatically improve match relevance. But since the default Shopify or WooCommerce export often produces titles like the first one, product feed optimization actually starts with understanding what those title fields are doing in an auction.
Beyond titles, the feed sends Google a series of signals it uses to determine placement eligibility, auction competitiveness, and relevance scoring: GTINs (which unlock premium placements and enable product matching across retailers), product categories (which determine which auctions your products are eligible for), custom labels (which control how your campaigns segment bids by margin, inventory, or performance), and image quality (which affects CTR in Shopping results).
Product feed optimization is not a one-time setup task. It’s a recurring discipline. Google’s Merchant Center diagnostics change as inventory changes, prices shift, and policies update. A feed that was clean 90 days ago may have accumulated disapprovals, price mismatches, and missing attributes if nobody is maintaining it.
Why most ecommerce brands have broken feeds and don’t know it
The most common scenario: a brand launches a Shopify store, installs a Shopping feed plugin, connects it to Merchant Center, and launches Performance Max. The plugin exports whatever data Shopify has. Shopify has whatever data the product team entered when building the catalog. Nobody reviewed it through the lens of how Google’s auction system works.
The result: titles optimized for the brand’s internal naming conventions, not search queries. And GTINs missing on 30 to 60% of SKUs because nobody went through the catalog to find and add them. Categories set to level one or two depth. Descriptions duplicated from product page copy. Custom labels empty. Meanwhile, Merchant Center diagnostics show 8 to 15% of the catalog disapproved for fixable reasons.
None of this shows up obviously on a campaign dashboard, though. Performance Max still looks like it’s spending. ROAS appears at some level. The brand assumes the feed is fine because products are showing up.
But the benchmark isn’t “showing up.” The benchmark is “matching the right queries at the right impression share with the right eligibility for premium placements.” A feed that scores on that benchmark looks dramatically different from a default export.
The second reason feeds stay broken: the fix requires cross-functional cooperation. Improving GTINs requires someone to research UPC codes. Title structure improvements need someone with authority to change product data. Better categories require someone who understands both the product catalog and Google’s taxonomy. In most ecommerce teams, this falls into a gap between the Google Ads manager, the ecommerce team, and whoever manages the data feed. Nobody owns it. So it doesn’t get done.
The 9-point product feed optimization checklist
Each item below is ordered by the impact it typically produces on account performance. Work through them in this sequence, since each one builds on the previous.
1. Title restructure by category. The standard title formula is Brand + Product Type + Key Attribute + Attribute 2 + Attribute 3. For apparel: brand, item type, gender, material, color, size. Electronics titles need: brand, product name, key spec, model number. For industrial/specialty products: brand, compatibility descriptor, part number. The first 70 characters receive the most weight in Google’s matching. Front-load the most important query terms. In accounts where I’ve rebuilt feed titles from generic exports to structured templates, impression share on high-intent queries typically improves by 20 to 40% within 30 days of the new feed being processed.
2. GTIN (Global Trade Item Number) coverage. GTINs are the UPC barcodes, EANs, and ISBNs that identify manufactured products. When GTINs are present and accurate, Google can match your product to the same item listed by other retailers, pool review data, and unlock premium Shopping placements. Indeed, retailers with correct GTINs see roughly 20% more clicks on average, according to Google Merchant Center’s own data. For Shopify stores, GTINs are entered in the product variant’s barcode field. For custom or private-label products without GTINs, the identifier_exists attribute should be set to “no” rather than left blank. Hustle Marketers’ GTIN guide for Google Shopping covers validation, common errors, and bulk update workflows.
3. Product category depth (3+ taxonomy levels). Google’s product taxonomy has over 5,000 categories organized in a hierarchy. Many feeds submit products at level 1 or 2 (Apparel, or Apparel > Clothing). Google prefers level 3 to 5 depth (Apparel > Clothing > Men’s Clothing > Tops & T-Shirts > T-Shirts). Deeper categories give Google more signal about what your product is, which affects placement eligibility, auction matching, and which Shopping features apply. Setting categories at appropriate depth using the google_product_category attribute is a 20-to-30 minute task per category type. It typically produces measurable impression share improvement within one crawl cycle.
4. Description enrichment for the Shopping tab. Product descriptions don’t appear in standard Shopping ads. But they feed Google’s relevance model for the Shopping tab, free listings, and increasingly, AI-powered Shopping surfaces. Descriptions should be 300 to 1,000 characters covering material, use case, key specifications, and differentiating features. Avoid marketing language (“best quality, lowest prices”). Focus on product facts that help Google categorize and match correctly. For feeds using descriptions directly exported from landing page content, the biggest improvement actually comes from rewriting them to be specification-dense rather than conversion-copy oriented.
5. Image quality and supplemental image angles. The main image must show the product on a white or plain background, since this is required for Shopping ads. Images below 250x250px are ineligible. Images below 800x800px receive lower placement priority. For high-competition categories, product images at 1000x1000px or larger with clean backgrounds meaningfully improve CTR. Supplemental images (additional_image_link) allow up to 10 additional angles: lifestyle images, detail shots, packaging. These appear in expanded Shopping ad formats and have measurable CTR impact in apparel and home categories.
6. Custom labels for campaign and bid segmentation. Custom labels (custom_label_0 through custom_label_4) are free-form fields that let you segment your catalog however your business logic requires. Common uses: profit margin tier (high/medium/low), inventory level (clearance/in-stock/pre-order), seasonality (evergreen/seasonal/holiday), and performance history (proven/testing/new). As a result, custom labels let your campaigns bid differently by segment without restructuring your whole campaign. Custom label setup is one of the first feed changes applied on new accounts because it unlocks margin-aware bidding from day one.
7. Price and availability accuracy. Price mismatches between the feed and the product landing page are one of the most common Merchant Center disapproval reasons. When the feed price doesn’t match the live landing page price, Google disapproves the product and removes it from Shopping placements. Feed refresh frequency also matters: default Shopify feed plugins often update daily or every few hours. For accounts with active promotions or frequent price changes, that’s not fast enough. Setting up the Content API (free from Google) or a feed tool that supports sub-hourly refresh cycles prevents disapprovals from price mismatches during high-traffic periods.
8. Supplemental feed for rapid attribute updates. A supplemental feed overrides specific attributes in the primary feed without requiring a full rebuild. For title optimization, this means you can push improved titles for specific SKUs through a supplemental feed immediately, without waiting for the product team to update the main catalog. So supplemental feeds in Merchant Center are the fastest implementation path for product feed optimization changes on large catalogs. DataFeedWatch and Feedonomics both support supplemental feeds with rule-based transformation. Hustle Marketers’ AI feed optimization guide covers how AI-assisted supplemental feeds have changed what’s achievable at catalog scale.
9. Monthly Merchant Center diagnostics review. Merchant Center’s Diagnostics tab shows disapprovals, warnings, product-level errors, and attribute coverage gaps. Reviewing this monthly is non-negotiable, though. In catalogs over 500 SKUs, it’s common to find 5 to 15% of products disapproved for fixable issues after 60 to 90 days without review. Each disapproved product has zero Shopping impression eligibility: it’s simply invisible in auctions.
When feed changes break Performance Max
Significant feed changes, specifically title restructures and category changes on large portions of the catalog, can temporarily disrupt Performance Max performance. PMax learns from historical click and conversion data attached to specific products. When titles change, Google reprocesses products, which can reset some of that learning. In practice, this produces a 5 to 14 day period of elevated CPC and lower impression share while PMax recalibrates to the new data.
The right approach: stagger major feed changes rather than changing all titles simultaneously. Start with a supplemental feed targeting your top 20% of revenue-generating SKUs. Evaluate performance 14 days after processing. Then expand to the next segment once performance stabilizes.
Also worth noting: PMax campaigns don’t support negative keywords at the product level. Feed-level changes are how you control which queries products are matched to. Improving title specificity is the feed-level equivalent of adding negative keywords: it reduces irrelevant matches by making titles precise rather than broad.
Feed tools and when to use them
For catalogs under 500 SKUs: Merchant Center’s built-in feed rules and supplemental feed capabilities are sufficient. No paid tools needed.
Between 500 and 5,000 SKUs: DataFeedWatch ($49/month self-serve) provides rule-based title transformation, supplemental feed management, and a diagnostics dashboard that simplifies the monthly audit. Worth the cost at this catalog size.
For catalogs above 5,000 SKUs or multi-channel needs (Google + Meta + Microsoft + TikTok): Feedonomics ($500+/month full-service) or Akeneo (for PIM-level catalog management). At this scale, manual feed management is no longer viable and the tool cost returns its investment quickly through improved impression share and disapproval reduction.
What product feed optimization produced for real clients
P-REX Hobby, Shopify, hobby parts for Bin Chen. The account had clean tracking and reasonable campaign structure. Feed titles were the problem: generic product names that didn’t include the model-number compatibility descriptors Bin Chen’s customers searched. “Drive Shaft” instead of “Traxxas Rustler 4WD Compatible Drive Shaft 3.2mm Pin.” Rebuilding title logic to front-load brand, part type, and compatibility model number dramatically improved impression share on high-intent long-tail queries. ROAS hit 9x over 90 days. The P-REX Hobby case study documents the full feed process.
ArmorGarage, BigCommerce, garage floor coatings. Inherited account had 60% of SKUs missing GTINs. The products were branded and manufactured. GTINs existed but hadn’t been entered in Merchant Center. Adding GTINs unlocked premium Shopping placements and product-matching features across Google surfaces. Combined with title restructuring and a PMax rebuild using product-line asset groups, the account hit 1,500%+ ROAS within 90 days. The GTIN addition alone improved Shopping impression share by roughly 25% within two weeks of processing. The ArmorGarage case study covers the full account sequence.
ThePetsClub UAE, Shopify Plus, pet food and supplies. Product categories were set at level 1 to 2 depth across most of the catalog. After updating categories to level 3 to 4 depth (Animals & Pet Supplies > Pet Supplies > Dog Supplies > Dog Food), PMax impression share on pet food subcategory queries improved measurably within one processing cycle. Combined with Customer Match loading and meta campaign rebalancing, ROAS reached 14x over 90 days. Category depth was one of three feed changes applied in the first 30 days.
What I’d audit first in any ecommerce feed
First, start with Merchant Center diagnostics. How many products are disapproved? What percentage of the active catalog has zero Shopping impression eligibility right now? This is a 5-minute check that often reveals a 10 to 20% impression share problem with a fixable cause.
Then pull 20 product titles at random from the highest-revenue SKUs. How much of each title is brand name versus product attributes? Does the title match the actual search terms shown in the Shopping search terms report? If there’s a significant mismatch between title terms and query terms, feed restructuring will lift performance faster than any campaign setting change.
Then, check GTIN coverage. What percentage of the catalog has GTIN values? For branded products, any SKU without a GTIN is leaving potential premium placement eligibility on the table.
Fourth, check product category depth. What percentage of the catalog is categorized at level 3 or deeper?
Finally, check the feed refresh rate. How often is the primary feed updating Merchant Center? If there are promotions, flash sales, or frequent price changes, a daily or every-few-hours refresh may not be fast enough to prevent price mismatch disapprovals.
What product feed optimization costs and takes
Title restructure for top 200 SKUs by revenue: 4 to 8 hours of catalog research and supplemental feed setup, one-time. This is the highest-impact work in the checklist, so it’s worth doing well.
GTIN research and entry for branded products: 2 to 6 hours for catalogs under 500 SKUs, depending on how many products require manual research. Most GTINs can be found via the GS1 US database (gs1us.org) or the product’s manufacturer.
Category depth update for a catalog of 200 to 500 products: 2 to 4 hours using Merchant Center’s bulk upload workflow.
Custom label setup (5 labels across the full catalog): 2 to 3 hours if margin and inventory data exists in the ecommerce platform.
Feed tool subscription: $49/month (DataFeedWatch self-serve) to $500+/month (Feedonomics full-service).
Total first-pass feed optimization investment for a catalog of 200 to 1,000 SKUs: $800 to $2,500 in agency or specialist time, one-time. The return materializes in the first 30 to 60 days through improved impression share, reduced disapprovals, and better query matching.
Why work with Ishant Sharma on product feed optimization
Twelve years. 500+ ecommerce brands. $780M+ in trackable client revenue. Google Partner and Meta Business Partner. Upwork Top Rated Plus with a 99% Job Success Score and a 5.0/5.0 rating. Clutch Award Winner 2024.
Every new ecommerce engagement starts with a feed audit before campaign changes. P-REX Hobby, ArmorGarage, and ThePetsClub all produced their documented ROAS improvements through feed-level changes applied before bid strategy or creative adjustments. The feed determines what the campaigns can do. Optimizing campaigns on top of a weak feed produces marginal improvements. Optimizing the feed first creates the conditions for every other optimization to compound.
Hustle Marketers manages feed optimization as part of our standard ecommerce PPC engagements. Our ecommerce PPC agency page covers how those engagements are structured and what the first 90 days typically produces.
What to take from this
Product feed optimization is the work that makes every other Google Ads optimization more effective. Titles determine query matching. GTINs determine placement eligibility. Categories determine auction access. Custom labels determine bidding precision. Together, these attributes determine whether your campaigns can reach the right queries at the right impression share with the right product information.
In fact, the brands that outperform on Google Shopping aren’t running more sophisticated campaigns. They’re running better-maintained feeds. Their titles match how customers actually search. GTINs are complete. Categories are specific. And Merchant Center diagnostics show near-zero disapprovals. Everything else (Smart Bidding, PMax asset groups, audience signals) works better because the data underneath it is accurate and complete.
About Ishant Sharma
Ishant Sharma is a Google Ads specialist and Founder of Hustle Marketers, a Google Partner and Meta Business Partner agency working with e-commerce and lead-gen brands across the US, UK, UAE, and Australia. 12+ years in performance marketing. Trackable client revenue across his work has crossed $780 million. Upwork Top Rated Plus with a 99% Job Success Score and a 5.0/5.0 rating. Clutch Award Winner 2024. Based in Chandigarh, India.