Google Shopping Optimization and Product Feed Optimization

Google Shopping optimization is mostly feed work, not bid work. Shopping ads have no keywords, so Google matches shopper queries to your products using the data you send to Merchant Center: the product title above all, then product type, attributes, price and availability. Fix the titles so brand, product type and defining attribute sit in the first 70 characters, complete the attributes Google actually reads, then segment campaigns by margin instead of by store category. Merchants who restructure titles this way commonly report 15% to 30% more impressions and 10% to 20% higher click-through rate.

What is Google Shopping optimization?

Google Shopping optimization is the work of improving how your products are represented in Google Merchant Center and how budget is allocated across them in Google Ads. It has two halves that people often treat as separate jobs and should not. The first is the product feed, which is the targeting layer. The second is campaign structure and budget, which decides how much each product is allowed to spend.

The reason the feed comes first is structural. In a Search campaign you choose the keywords, so if you are appearing for the wrong queries you can go and change them. In Shopping you cannot. Google reads your product data and decides which queries you deserve. Your title is effectively your keyword list, your product type is your category signal, and your attributes are your qualifiers. A store with vague titles is not bidding badly, it is invisible for the searches that would have converted.

This is also why Shopping is where most ecommerce budget goes. Shopping ads take roughly 65% of ecommerce paid search clicks and average about $0.66 per click, against roughly $1.42 on ecommerce Search campaigns. Cheaper clicks with higher purchase intent is a good trade, and it is available to anyone willing to do unglamorous data work.

How do I optimize my Google Shopping feed?

Work in this order, because each step makes the next one more useful. Doing it in reverse is the most common reason a feed project produces nothing.

1. Clear every Merchant Center disapproval first. A disapproved product cannot serve at any bid. Check the Products page for both hard disapprovals and the softer "limited performance" warnings, which usually mean missing GTINs, a price mismatch against your landing page, or shipping and returns policies that are not configured.
2. Rewrite product titles. This is the single highest-return change in almost every account that has never done it. Structure is covered in the next section.
3. Complete the attributes Google reads. Not all fields are equal, and filling in everything is a waste of effort. The table below shows which ones move performance.
4. Fix availability and price accuracy. Spending on products you cannot ship is pure waste, and price mismatches between feed and landing page cause disapprovals. This depends on your stock data being right at the source rather than being patched in the feed, which is why stores with messy inventory records tend to have messy feeds too. If stock counts drift between your store, your warehouse and your suppliers, that is worth fixing with proper inventory tracking and forecasting before you blame the feed.
5. Add custom labels for segmentation. Tag each product with margin band, price band, bestseller status and seasonality. These do nothing on their own; they exist so you can split campaigns and budgets by them later.
6. Add supplemental feeds or feed rules. Use these to apply the changes above without touching your storefront. Your ecommerce platform keeps its own product names and Google gets search-optimized ones.

How should I structure a Google Shopping product title?

Put brand, product type and the defining attribute in the first 70 characters. Google accepts up to 150 characters in a title, but most Shopping placements truncate around 70, so anything after that helps matching but is not read by the shopper. Lead with what the buyer typed and follow with the qualifiers that separate this product from the others on the page.

The right order varies by category, because different things drive the purchase decision.

Two rules apply everywhere. Never put promotional language in a title: "Free Shipping", "Sale", "Best Price" and "Buy Now" all violate Merchant Center policy and will get products disapproved. And if you are generating titles with AI, submit them in the structured_title attribute rather than overwriting title, which is the supported path for automatically generated titles and keeps product data review clean.

Test rather than assume. Apply one title structure to half the catalog and a different one to the other half using feed rules, tag each group with a custom label, then compare click-through rate and return on ad spend by that label in Google Ads. It takes a few weeks and settles the argument with your own data instead of someone else's case study.

Which feed attributes actually affect performance?

Merchant Center accepts dozens of attributes. A much smaller set does the work.

How do I optimize Google Shopping campaigns?

Once the feed is sound, the campaign question is simply which products are allowed to spend how much. The default setup, one campaign containing everything, guarantees that your bestsellers subsidize a long tail of products that will never convert, and the account-level return on ad spend hides it.

Split by economics, not by the category structure your website happens to use. Bestsellers with proven conversion history deserve their own campaign and their own target. High-margin products can afford a lower return on ad spend target than low-margin ones and still make more money per sale. Products with no conversions after meaningful click volume belong in a low-budget campaign where they can prove themselves cheaply or stay quiet.

Set targets from margin rather than from a benchmark. If you sell at a 40% gross margin you break even at 2.5x return on ad spend, so a 3x target is close to working for free. Most ecommerce accounts land between 3x and 8x, and top-quartile stores reach about 6x on Shopping, but those averages mean nothing until you have compared them to your own break-even. Our ROAS calculator does that arithmetic from your margin.

Then decide how Performance Max fits. PMax delivers roughly 15% to 20% better return on ad spend than Standard Shopping in most categories, but it will also happily spend on people searching your brand name and report those sales as its own. Apply brand exclusions account-wide and keep a dedicated brand Search campaign catching those queries far more cheaply. The Performance Max optimization guide covers the exclusions and asset group work in detail.

How long does Google Shopping optimization take to work?

Expect two to four weeks before feed changes show reliable results. Merchant Center needs to recrawl and reapprove products, which is usually 24 to 72 hours, and then the campaigns need enough impressions on the new titles for the comparison to mean anything. Judging a title change after five days is how good changes get reverted.

Campaign restructuring takes longer, roughly four to six weeks, because smart bidding re-enters a learning period whenever you split or merge campaigns. Resist the temptation to keep adjusting during that window. The one exception is disapprovals and availability errors, which should be fixed immediately and show up within days, because they are not optimization at all, they are products that cannot sell.

Should I do this myself or hire help?

A catalog under a few hundred products with stable inventory is genuinely manageable in-house once the initial title work is done. The upkeep is a couple of hours a month. Past that, the problem becomes frequency rather than difficulty: prices change, stock moves, competitors reprice, and a feed that was well optimized in March is stale by June.

That is the case for either an agency or software, and the split is straightforward. Agencies are worth paying for judgment about which products to sell and what the catalog should look like. Recurring execution, which is feed rules, negative keywords, product-level budget shifts and bid targets, runs better daily than monthly and is what ecommerce PPC services should really be measured on. Most agencies work an account weekly at best, which is the gap worth asking about before signing. If you are also running Meta and trying to decide where the next dollar goes, our comparison of Facebook ads versus Google ads for ecommerce covers how to judge the two together.

Common questions

Does Google Shopping optimization still matter with Performance Max?

It matters more. Performance Max removes keyword and placement control, which leaves the product feed as one of the few inputs you still own. PMax builds its Shopping inventory from the same Merchant Center data, so a weak feed limits a PMax campaign exactly as it limits a Standard Shopping one, with less reporting to tell you why.

How many characters should a Google Shopping title be?

Use the full 150 characters Google allows, but treat the first 70 as the part that must earn the click, because that is roughly where most Shopping placements truncate. Brand, product type and the defining attribute go first. The remainder still helps query matching even when the shopper never sees it.

Do I need GTINs for Google Shopping?

For branded products that carry them, yes. Missing GTINs restrict reach and can trigger limited-performance warnings, because Google uses them to match your listing to its own product catalog. Genuinely custom or handmade goods without a manufacturer barcode are exempt, and you declare that with the identifier_exists attribute set to false.

What is a good conversion rate for Google Shopping?

Shopping campaigns average around 1.9% conversion rate, while ecommerce Google Ads accounts overall average about 2.8%. Shopping sits lower because it captures shoppers earlier, when they are still comparing. Judge it on return on ad spend rather than conversion rate, since a 1.5% conversion rate on cheap clicks can outperform a 3% rate on expensive ones.

Why are my Shopping ads not showing?

Check in this order: product disapprovals in Merchant Center, then availability set to out of stock, then price or shipping mismatches against the landing page, then whether the product is actually included in an active campaign, and only then bids and budget. The great majority of "not showing" cases are data problems rather than auction problems.