Google Shopping Ads services where the lever is your feed, not your bid
Complete attributes, accurate identifiers and properly constructed titles have been reported to expand impression share by 40 to 60% with no bid change at all. Most accounts spend their optimisation time adjusting bids, because bidding looks like advertising and data entry does not.
Feed quality, title construction, disapproval clearance and product level margin decisions, for paid Shopping and free listings together.
Blue Shirt
Northgate Oxford Cotton Shirt, Blue, Large
What are Google Shopping Ads services?
Google Shopping Ads services cover the product data behind the advertising: building and maintaining the Merchant Center feed, constructing titles and attributes so products are eligible for the right searches, clearing disapprovals and warnings, structuring campaigns around margin rather than around catalogue order, and monitoring the diagnostics that quietly remove products from visibility.
Campaign types, Performance Max structure and brand separation sit on our Google Ads management page.
This page is the feed, which is the input those campaigns are working from. Shopping is unusual in that the data matters more than the settings.
Why the feed decides almost everything
Shopping does not have keywords. Google works out what your product is from the data you send it, which makes that data your targeting.
Feed quality moves impression share more than bids do. Complete attributes, accurate identifiers and keyword-rich titles have been reported to expand impression share by 40 to 60% without touching a bid. That is not an optimisation, it is eligibility. You cannot bid on a search you are not eligible for.
Disapprovals remove products silently. Merchant Center diagnostics carry disapprovals, warnings and missing attribute notices that quietly erode visibility. Nobody receives an alarm when a product stops showing, and on a large catalogue it is entirely possible to lose a whole category without noticing for weeks.
Titles should be built, not written. On a catalogue of any size, feed rules can construct titles programmatically from fields you already hold: brand, product type, colour, size, material. That scales good titles across thousands of products without anybody editing them one at a time.
The same feed powers your free listings. Merchant Center Next consolidates product data across paid ads and free surfaces, so feed work pays twice. Improving it lifts what you pay for and what you do not, which is a rare thing in paid media.
Performance Max now carries the majority of Shopping impressions, and it performs well only when the feed underneath it is complete. Automation cannot infer an attribute you never sent.
What each attribute is actually doing
Worth knowing which fields earn visibility and which merely satisfy a validator.
| Attribute | What it decides | Common failure |
|---|---|---|
| Title | Most of what you are eligible to appear for | The store name for the product, with no attributes in it |
| GTIN and identifiers | Matching your product to the known catalogue | Missing or wrong, which limits reach and comparison placement |
| Product type | Your own categorisation, used for structure and bidding | Left blank, so campaigns cannot be segmented sensibly |
| Google product category | How Google classifies the item | Auto assigned and wrong, putting products in odd auctions |
| Colour, size, material | Eligibility for filtered and specific searches | Held in variant data but never sent in the feed |
| Availability and price | Whether the ad is allowed to run at all | Out of sync with the site, causing disapprovals |
| Images | Click through rate, which feeds everything downstream | Lifestyle shots where a clean product image performs better |
The colour, size and material row is the most common quiet loss. The data usually exists in the store, it simply never makes it into the feed, and every filtered search using those terms passes you by.
What our Google Shopping Ads services cover
Feed first, campaign second. That order is the entire method and it is the reverse of how most accounts are managed.
Feed build and maintenance
Merchant Center set up properly, connected to your store, with a feed that updates reliably rather than one somebody uploaded once as a spreadsheet.
Title construction
Built with feed rules from brand, type, colour, size and material, so thousands of products get good titles without anyone editing them individually.
Attribute completion
The fields that exist in your store but never reached the feed, which is where most of the recoverable impression share is sitting.
Disapproval clearance
Diagnostics worked through and kept clear, with monitoring so a category dropping out is noticed in days rather than at the next quarterly review.
Image review
Clean product images that comply and convert, since the image is doing most of the persuading before anybody reads a word.
Campaign structure by margin
Products segmented so high and low margin lines are not bid on identically, using product type and custom labels rather than one undifferentiated group.
Excluding what should not run
Loss making products, out of stock lines and items you cannot fulfil profitably, removed from spend rather than left in because they came with the catalogue.
Free listings
The same data improved once and used across paid and organic Shopping surfaces, which makes feed work unusually good value.
Product level reporting
Performance by product and margin band, because a Shopping account average hides both the winners funding everything and the losers quietly eating it.
Deciding what deserves spend needs product level margin. Our eCommerce profit margin calculator is a straightforward way to work out which lines can carry a click price at all.
How the Shopping work runs
Clear the diagnostics first
Disapprovals and warnings resolved before anything else, because products that cannot serve make every other improvement theoretical.
Fill the attributes you already have
Colour, size, material, identifiers and product type pushed into the feed from data your store already holds. Cheapest visibility available anywhere in paid media.
Rebuild titles with rules
Constructed programmatically rather than written by hand, so the improvement covers the whole catalogue and survives new products being added.
Then structure the spend
Segmented by margin, with loss making lines excluded and reporting at product level. Bids come last because by this point they are moving something worth moving.
What feed work will not do
Three things worth saying, because feed optimisation gets sold as a cure for problems it cannot touch.
It cannot make an uncompetitive price competitive. Shopping puts you next to your competitors with the price visible. If you are meaningfully more expensive with no visible reason, better data simply gets you seen losing more often.
It cannot fix a product page that does not convert. Shopping is unusually efficient at delivering high intent traffic. If those visitors arrive and leave, the money is being spent well and lost on arrival, and the fix belongs on the page.
It cannot rescue products with no margin. Some lines cannot carry a click price at any level of optimisation. The correct answer is to exclude them from spend, not to advertise them more efficiently into a loss.
Where it does earn its money is reliable: catalogues with incomplete data, stores that have never had disapprovals cleared, and any account where somebody has been adjusting bids for months while the feed stayed exactly as the platform exported it.
How much do Google Shopping Ads services cost?
A flat monthly fee, with the initial feed work priced as a project because it is finite and front loaded.
Feed audit
Attribute completeness, disapprovals, title quality and which products are eating budget without earning it. Yours to act on with or without us.
Feed rebuild
Diagnostics cleared, attributes filled, titles rebuilt with rules and campaign structure aligned to margin. A one off, scoped after the audit.
Ongoing management
Feed monitoring, new product coverage, disapproval clearance and product level reporting. Flat monthly fee, cancellable.
Judge the work on impression share and on profit by product. A rising account average frequently just means the cheap products sold more, which is not the same as the account improving.
Shopping alongside the rest of the account
Google Ads management
Campaign types, Performance Max structure and keeping brand traffic out of the campaign that would otherwise absorb it.
See Google Ads →eCommerce SEO services
Product and category pages in organic search. The same product data problems tend to show up in both places.
See eCommerce SEO →Product listing services
Listing creation and product data across marketplaces, for sellers publishing the same catalogue in several places.
See listing services →Common questions about Google Shopping Ads
Check Merchant Center diagnostics first. Disapprovals, warnings and missing required attributes remove products from serving without any alarm being raised, so it is entirely possible to lose a category quietly. After that, the usual cause is eligibility rather than budget: if your data does not describe the product properly, you are not entering the auctions you think you are.
Fix the feed. Complete attributes, accurate identifiers and properly constructed titles have been reported to expand impression share by 40 to 60% with no bid change at all. Shopping has no keywords, so your product data is your targeting, and improving it changes what you are eligible to appear for rather than how much you pay for it.
The attributes a person would actually type, in a sensible order: brand, product type, then distinguishing details like colour, size or material. On any real catalogue these should be constructed with feed rules from fields you already hold rather than written individually, so the improvement covers everything and survives new products being added.
Where they exist for your products, yes. Identifiers are how Google matches your item to the known catalogue, which affects reach and comparison placements. Missing or incorrect identifiers limit visibility, and incorrect ones can also trigger disapprovals, which is the worse of the two outcomes.
Yes, and it is one of the better arguments for spending money on data rather than bids. Merchant Center consolidates product information across paid ads and free surfaces, so improving the feed lifts both what you pay for and what you do not. Very little else in paid media works twice like that.
No, and this is where a lot of budget goes quietly. Products with margin that cannot support a click price should be excluded rather than optimised, and out of stock or unfulfillable lines should not be running at all. Segmenting by margin so the whole catalogue is not bid on identically is usually the second biggest improvement after the feed itself.
Be careful here. Shopping displays your price directly next to alternatives, so better data mainly gets you seen more often in a comparison you may be losing. If there is a visible reason for the premium, such as bundle contents, warranty or delivery speed, that can work. If there is not, the honest answer is that feed optimisation will not solve a pricing problem.
Find out what your feed is leaving on the table
We will check your attribute completeness, clear the diagnostics you are currently failing, review your titles against what people actually search, and tell you which products should not be running at all. On most catalogues the recoverable visibility is larger than anyone expects. No obligation, no sales pitch.
Get my free feed audit- Attribute completeness
- Disapprovals and warnings
- Products to exclude