How Google Shopping bidding actually works inside the carousel

Shopping ads are not search ads

If you have spent time managing Google Search campaigns, you will have a mental model of how auctions work: you bid on keywords, Google ranks you based on bid and Quality Score, and you pay the minimum needed to maintain your position. That model is roughly correct for text ads, but Shopping works differently in several important ways.

Google Shopping ads are not keyword-targeted. You do not bid on search terms. Instead, Google matches your product feed data to user queries and decides which of your products to show, in which position, and at what cost. The auction still exists, but the inputs are different — and understanding those inputs is essential if you want to compete effectively.

The product feed is your foundation

In standard Search campaigns, your ad copy and keyword selection determine relevance. In Shopping, your product feed does that job. Google uses your feed data — titles, descriptions, product categories, GTINs, images, pricing, and availability — to determine which queries your products are eligible to appear for.

This means feed quality is not a nice-to-have. It is the single most important factor in determining your Shopping visibility, before you even get to bidding.

What Google looks at in your feed

• Product titles: These carry the most weight. A title that includes the brand, product type, key attributes (colour, size, material), and relevant modifiers will match a far wider range of queries than a vague or truncated title.
• Product descriptions: Less heavily weighted than titles, but still used for query matching. Detailed, accurate descriptions improve your eligibility for long-tail searches.
• Google Product Category: Mapping your products to the correct Google taxonomy helps the system understand what you sell and match it appropriately.
• GTINs and brand identifiers: Products with valid GTINs can be matched to Google's product graph, which improves your chances of appearing in relevant results and enables features like price comparisons.
• Images: Google assesses image quality and can demote products with poor images, watermarks, or misleading visuals.
• Price and availability: Products that are out of stock or priced significantly above competitors may be suppressed or shown less frequently.

How the auction actually runs

When a user searches for something that triggers Shopping results, Google runs an auction among all eligible product listing ads. Here is what happens, step by step.

Step 1: Query matching

Google scans its index of product feeds and identifies all products that are relevant to the user's query. This is not keyword matching — it is semantic matching. Google determines relevance based on your feed data, using natural language processing to understand what the user is looking for and which products fit.

A search for "men's navy wool overcoat" will match products whose titles and descriptions contain those terms or close variants. But it may also match products described as "dark blue wool coat for men" if Google's systems judge the semantic overlap to be sufficient.

Step 2: Bid evaluation

For each eligible product, Google evaluates the advertiser's bid. This is either a manual CPC bid or a target generated by an automated bidding strategy (Target ROAS, Maximise Clicks, etc.).

Critically, the bid alone does not determine ranking. Google combines the bid with its own assessment of expected click-through rate and relevance to produce an ad rank — similar in concept to the Ad Rank used in Search, though the specific formula is not identical.

Step 3: Ranking and position

Products are ranked by their combined score — bid multiplied by expected performance and relevance. The highest-ranked products appear first in the carousel or in the most prominent positions on the page.

This means a product with a lower bid but excellent feed data and strong historical click-through rate can outrank a product with a higher bid but weaker relevance signals. Bidding high is not sufficient. You need your product data to work as hard as your budget.

Step 4: Pricing (second-price mechanics)

Google Shopping uses a variant of the second-price auction. You do not pay your maximum bid. You pay the minimum amount needed to maintain your position above the next-highest-ranked advertiser. In practice, this means your actual CPC is often significantly below your maximum bid.

This is why aggressive bidding in Shopping is less risky than many retailers assume. Setting a higher maximum bid improves your chances of winning favourable positions, but you will usually pay substantially less than that maximum.

The "Quality Score" equivalent

Standard Google Search campaigns have an explicit Quality Score metric — you can see it in your account, and it directly affects your ad rank and CPC. Shopping does not have a visible Quality Score. But make no mistake: an equivalent system exists.

Google assesses your products on several quality dimensions:

• Expected click-through rate (CTR): Based on historical performance of your products and similar products. If your ads consistently earn clicks when shown, Google rewards you with better rankings.
• Feed quality and completeness: Products with comprehensive, accurate feed data are favoured. Missing attributes, incorrect categories, or thin descriptions reduce your quality signals.
• Landing page experience: Google evaluates the page users land on after clicking. Page speed, mobile-friendliness, relevance to the product shown, and ease of purchase all contribute.
• Price competitiveness: Google knows what your competitors charge for the same or similar products. Significantly higher prices can reduce your impression share, even if your bid is competitive.
• Merchant reputation: Seller ratings, delivery speed, return policies, and Google Merchant Centre account health all feed into the system.

The combined effect of these factors is substantial. Two retailers bidding the same amount on an identical product can see wildly different results based purely on these quality signals.

Shopping ads versus Search ads: key differences

Understanding the distinctions between Shopping and Search auctions helps you avoid costly mistakes.

No keyword targeting

In Search, you choose which queries to bid on. In Shopping, Google decides which queries your products match. You can add negative keywords to exclude irrelevant queries, but you cannot proactively target specific terms. This makes feed optimisation — not keyword research — the primary lever for controlling your visibility.

Multiple products per advertiser

In Search, you typically show one ad per query. In Shopping, Google can show multiple products from the same merchant in a single carousel. This means a retailer with a broad, well-optimised product catalogue can dominate the Shopping results for certain queries.

Visual format changes behaviour

Shopping ads include an image, price, merchant name, and (where applicable) ratings and promotions. Users make a significant portion of their click decision based on the image and price before they even notice the merchant. This means your creative assets — specifically your product images and pricing — have an outsized impact on CTR relative to Search ads, where the headline and description carry more weight.

Position matters differently

In Search, the top organic or paid result gets disproportionate attention. In the Shopping carousel, users tend to scroll horizontally, and the drop-off between positions is less steep. Being in position three or four in the carousel is not dramatically worse than position one, particularly on mobile where the carousel is the dominant format. This has implications for bid strategy: you do not always need to win the top spot to achieve strong returns.

Practical implications for bid strategy

If you take one thing from this piece, it should be this: Google Shopping is not a pure auction where the highest bidder wins. It is a system that rewards the combination of competitive bids and high-quality product data.

Invest in feed quality before increasing bids

Improving your product titles, adding missing attributes, and ensuring accurate categorisation will often deliver better results than simply raising your bids. Many retailers overspend on CPCs to compensate for poor feed quality — a strategy that is both expensive and self-limiting.

Use negative keywords aggressively

Since you cannot choose which queries trigger your Shopping ads, negative keywords are your primary tool for eliminating wasted spend. Review your search term reports regularly and exclude queries that are irrelevant or unprofitable.

Monitor your pricing position

If your products are consistently more expensive than competitors showing in the same carousel, your CTR will suffer regardless of your bid. Price monitoring tools can help you understand where you sit relative to the market and adjust your strategy accordingly.

Consider the CSS advantage

Everything described above operates within the auction mechanics. But recall that the bid entering the auction depends on which CSS places it. A third-party CSS means your full bid competes, rather than a discounted version. This does not change the auction mechanics, but it changes where you sit within them — and that difference compounds across every product and every click.

The bottom line

Google Shopping is a sophisticated auction where success depends on far more than bid price. Product data quality, historical performance, landing page experience, and pricing all play material roles in determining your visibility and cost per click. Retailers who treat Shopping as a feed-quality challenge first and a bidding challenge second consistently outperform those who take the opposite approach.

Understanding the mechanics is the first step. The second is ensuring your Shopping setup — from feed to bidding to CSS choice — is working as hard as it can.