Feed First
Titles rewritten for how people search rather than how the catalogue is organised. It is the highest-leverage change available in Shopping.
In ecommerce advertising the product feed does more work than the ad copy. Titles, attributes and availability decide which queries a product appears for at all. Ecommerce PPC starts there, then optimises at product level rather than campaign level — because a strong overall return can easily hide a catalogue where the profitable items are subsidising the rest.
Shopping campaigns have no keywords. What the feed says is what decides where products appear, and most feeds are whatever the platform exported.
Titles rewritten for how people search rather than how the catalogue is organised. It is the highest-leverage change available in Shopping.
Account-level return hides everything. The useful question is which products earn and which quietly consume the budget.
A return target set on revenue optimises toward volume. Where margins vary across the catalogue, that is optimising toward the wrong products.
They compete for the same queries. Managed separately, they bid against each other with your money.
Ecommerce demand is not flat, and budgets set once in January are wrong by October.
Four causes, and they compound.
Review My Account →A 5x return on a product with a 15% margin loses money. Optimising on revenue targets the products that sell, not the products that pay.
Product titles written for a catalogue page — "SKU-4471 Blue" — match almost nothing anyone searches.
Spend concentrates on whatever converts most, which is often a low-margin bestseller, while the rest of the catalogue never gets impressions.
Availability not syncing, so budget goes to products nobody can buy.
Feed, structure, measurement and the weekly work — with the catalogue treated as the variable it is.
Paid and organic overlap heavily on product and category terms — see ecommerce SEO.
The feed is fixed before the bidding is touched. Bidding on a bad feed optimises the wrong thing faster.
Every product checked for title quality, attribute completeness, category accuracy and availability sync.
Conversion values verified, and margin data passed where the store can provide it.
Products segmented by performance and margin so budget can be steered.
Product-level performance reviewed, negatives added, budget moved.
Budgets and bids adjusted ahead of the demand curve rather than during it.
Shopping campaigns have no keywords. Google matches queries to products using what the feed says, which makes the feed the closest thing to keyword targeting you have.
The words someone would actually type, in roughly the order they would type them. Brand, product type, and the attributes that distinguish it — size, colour, material, model — rather than an internal SKU and a catalogue label.
Titles are truncated in the ad, so the important words belong at the front. "Waterproof Hiking Boots — Mens, Size 10, Brown" outperforms "Product 4471 — Brown, Waterproof, Mens Hiking Boots Size 10" for no reason other than word order.
This is unglamorous work and it is usually the single largest available improvement in a Shopping account.
A return-on-ad-spend target treats every pound of revenue as equivalent. If your catalogue has products at 60% margin and products at 12%, it does not — and the bidding will happily push spend toward whichever converts most easily.
Where the store can pass margin or profit as the conversion value instead of revenue, the automated bidding optimises toward what actually pays. Where it cannot, segmenting campaigns by margin band achieves a rougher version of the same thing.
Either way, the account stops rewarding the products that sell well and start rewarding the products worth selling.







Ecommerce PPC is paid search for online stores, run mostly through Shopping campaigns where the product feed rather than a keyword list determines which queries a product can appear for.
Shopping campaigns have no keywords. Google decides which searches a product is relevant to using the feed — its title, description, attributes and category. A feed exported straight from the store usually describes products the way the catalogue does, not the way people search.
Usually both, coordinated. Shopping covers product searches with images and prices; Search covers category and brand queries and lets you control messaging. Run independently they bid against each other, which is why they belong in one plan.
There is no universal figure — it depends entirely on your margins. A store with 60% margins can sustain a far lower return than one at 15%. Anyone quoting a target ROAS without asking about your margins is quoting a number, not advice.
They should stop being advertised, which requires availability syncing correctly between the store and Merchant Center. Broken sync is a common and entirely avoidable way to spend budget on products nobody can buy.
Still deciding if ecommerce ppc is right for you?
Talk to UsMost ecommerce accounts are managed as though every product deserves the same treatment. Budget is set at campaign level, performance is read at account level, and the conclusion is that the account is doing well or badly.
Underneath that number, catalogues are almost always uneven. A minority of products generate most of the profitable revenue, a larger group breaks roughly even, and a tail consumes budget while returning little. All three are averaged into the same figure.
Reading performance at product level tends to change what people want to do next. It is the same account and the same data — just disaggregated enough to act on.
Give us read-only access to your Google Ads and Merchant Center. We will audit the feed, check availability sync, and show you performance broken out by product rather than by campaign.
