PRODUCT FEED OPTIMISATION: THE ATTRIBUTES GOOGLE WEIGHS

When it comes to product feed optimisation on Google Shopping, the most common question is: where do you start? The feed contains dozens of attributes, title, description, GTIN, images, price, availability, colour, size, category, custom labels, brand, condition, and many more. Not all of them carry the same weight on ranking. Some determine whether the product appears for a given query. Others influence CTR. Others still are purely operational and have no bearing on positioning.

Product feed quality in the Google Merchant Center feed is not a binary concept. It is a hierarchy where some attributes have a direct and immediate impact on ranking, others have an indirect and delayed impact, and others have no impact on positioning at all but are still useful for the operational management of the campaign.

This article analyses the hierarchy of Google Shopping feed attributes 2026 in order of priority: starting from the official Google Merchant Center documentation and flagging where direct experience on real accounts leads to more precise conclusions than those documented.

The title: the attribute with the greatest weight

According to the official Google Merchant Center documentation on the title attribute, the title is the attribute Google uses as the primary signal to determine a product's relevance to a search query. If the user's query contains words not present in the title, the product is not considered relevant for that query, and does not appear in the carousel for that search, regardless of the bid.

This makes Shopping product title optimisation the highest-leverage intervention among all feed optimisation work. A poorly structured title cannot be recovered by any other attribute or by any bid increase. Conversely, an optimised title can expand coverage across hundreds of relevant queries without touching the budget.

The optimal title structure in 2026

The structure that produces the greatest coverage of relevant queries, confirmed by both Google documentation and experience on e-commerce accounts across various categories, is:

Brand + Product name + Variant + Key attribute

• Clothing: "Nike Air Max 90 Men's White Size 9", not "Nike Running Shoe"
• Electronics: "Samsung Galaxy S25 128GB Black", not "Samsung Smartphone"
• Furniture: "Ikea Billy Bookcase White 80x202cm", not "Wood Bookcase"
• Food: "Lavazza Qualità Oro Ground Coffee 250g", not "Lavazza Coffee"

The variant and key attribute in the title cover the specific queries that generate the highest conversion rates: someone searching "men's running shoes white size 9" has a much more precise and commercial intent than someone searching "running shoes". Including the variant and attribute in the title is the most direct way to intercept high-conversion queries.

What to avoid in the title: promotional text ("OFFER", "DISCOUNT", "BEST"), unnecessary special characters, and titles that are too short to cover the relevant attributes for the category.

Inference on the title: on accounts where titles were systematically rewritten following the brand + product + variant + attribute structure, average impressions per product increased by 35–60% in the first 3–4 weeks, without any bid changes. The impact is more pronounced in categories with specific queries (clothing, electronics) than in generic categories. It is the single intervention with the best time/impact ratio on any Shopping account.

Optimal title length

Google accepts titles up to 150 characters, but typically displays only the first 70–80 characters in the carousel. The correct strategy is to place the most relevant information for ranking in the first 70 characters (brand + product + main variant) and use the remaining characters for secondary attributes that cover more specific queries.

GTIN, category and technical attributes: relevance and enrichment

Google Shopping GTIN, the product's EAN, UPC or ISBN code, is the attribute with the second-highest impact on product feed optimisation. The GTIN is not a direct relevance signal. It is a quality and identity signal that unlocks significant benefits across the Google ecosystem.

According to the official Google Merchant Center documentation on GTIN, providing the correct GTIN allows Google to match the product against its internal catalogue with all associated information (aggregated reviews, technical specifications, variants, price history), enable advanced features such as product reviews in the carousel, and improve visibility on specific queries that include the EAN code or standardised technical specifications.

Inference on GTIN: on accounts where GTIN was added to products that lacked it (but had an official EAN), average CTR improved by 10–20% in the first 4 weeks. The improvement is attributable to automatic data enrichment , in particular to product reviews that appear in the carousel only for products with a verified GTIN.

The Google Product Taxonomy category

The category (the google_product_category attribute) is the second technical attribute with significant impact on optimise Merchant Center feed work. Google uses the category to determine which additional attributes are relevant for that type of product and to filter products in category-based searches in the Shopping tab. The correct practice is to use the most specific category available in Google's taxonomy, not a generic category.

Colour, size and material

For categories where these attributes are relevant, colour and size in Google's standardised format are attributes with a direct impact on the coverage of specific queries. A product without colour in the feed does not appear for "white running shoes". A product with colour in a non-standard format may be interpreted correctly but this is not guaranteed.

Images in the feed: indirect but measurable impact

Images do not directly influence Ad Rank in the Google Merchant Center feed in the strict sense. But they influence CTR, and CTR is one of the three components of the Quality Score that feeds into Ad Rank. The impact is real and measurable, but indirect and delayed compared to title optimisation.

According to the Google Merchant Center documentation on image requirements, the minimum technical requirements are: at least 100×100 pixels (250×250 for clothing), JPG, PNG or GIF format, no overlaid text, no brand logo on the main image. The recommended resolution is 800×800 pixels or higher.

But minimum requirements are not the same as images that maximise CTR. From a product feed optimisation perspective, images should be optimised based on category:

• Pure white background: optimal for electronics, tools, technical products where the user wants to see the product clearly.
• Lifestyle or in-context images: perform better for sportswear, furniture, accessories. The user sees the product in use and can better assess how it fits their style.
• Product on mannequin vs flat lay: for clothing, the image worn by a model tends to generate higher CTR than flat lay because it gives proportions and shows how the garment falls.

Inference on images: in A/B tests on identical products with different images, the higher-CTR variant almost always matched the visual expectations of its specific category. For sportswear and furniture, lifestyle images showed CTR 15–25% higher than white background. For electronics and technical products, the white-background image with the product clearly visible performed equivalently or better.

Custom labels, description and other attributes: what they actually do

Custom labels: management tools, not ranking factors

Shopping feed custom labels are among the most misunderstood attributes in the Google Shopping ecosystem. Many advertisers treat them as if they contributed to product ranking or visibility. They do not.

Custom labels (0 to 4) are internal feed attributes that Google uses exclusively to allow the advertiser to segment products into groups within the campaign. They are not visible to the user in the carousel, do not influence a product's relevance to any query, and do not factor into any component of Ad Rank.

Their value is purely operational. Examples of correct use:

• custom_label_0 = "high_margin" → higher bid for products with greater profitability
• custom_label_1 = "season_summer" → activate or deactivate product groups by season
• custom_label_2 = "bestseller" → aggressive bid on products with a high conversion history
• custom_label_3 = "low_stock" → reduced bid or pause for products nearing sell-out

Segmentation via custom labels is one of the most effective tools for managing large catalogues, not for improving the ranking of individual products, but for allocating budget intelligently across segments with different priorities.

The product description: secondary relevance signal

The Google feed product description has a lower impact on ranking than the title, but it is not irrelevant. Google uses it as a secondary relevance signal for long-tail queries not covered by the title. A description with specific technical attributes, material, dimensions, compatibility, can expand coverage to queries the title does not reach.

The optimal length is between 500 and 1,000 characters. Duplicate descriptions across multiple products lower the perceived product feed quality in Merchant Center and should be avoided.

Price and availability: implicit quality attributes

Price and availability do not directly influence ranking as explicit signals, but they affect the implicit quality of the feed in two ways. Products with a price that is inconsistent between the feed and the landing page are penalised or disapproved by Merchant Center. Products with "out of stock" availability in the feed are automatically excluded from the carousel. Keeping price and availability synchronised between feed and site in real time is the minimum prerequisite for a functioning Shopping campaign.