Google Shopping ROAS Calculator

What the formula says

Shopping campaign ROAS is calculated identically to search, revenue ÷ spend, but the mechanics differ. There are no keywords; the product feed and the product itself determine which queries the ad appears for.

That makes feed quality the primary lever, where in search it would be keyword and ad copy.

Break-even ROAS is still 1 ÷ contribution margin, but it should be calculated per product rather than per account because Shopping bids at product level.

The numbers, worked through

A Shopping campaign spending $12,000 and returning $30,000 is a 2.5 ROAS overall.

Broken down by product: the top ten products return 3.4 on $7,000 of spend, the middle group returns 2.1 on $3,500, and the tail returns 0.9 on $1,500.

The tail is spending $1,500 to generate $1,350 of revenue and roughly $743 of contribution, a $757 loss hidden inside a healthy-looking account average.

Excluding or bidding down that tail lifts the account from 2.5 to 2.7 and adds $757 of profit without any change in how the other products are managed.

What the number leaves out

An account-level Shopping ROAS almost always conceals a wide distribution at product level. The average is arithmetically true and operationally useless.

Product margins also vary, so a single break-even threshold across a mixed catalogue mismanages both ends of it.

Turning it into a decision

Segment Shopping campaigns by margin band or by performance tier so different products can carry different targets. It is the single highest-return structural change in most Shopping accounts.

Then work on the feed. Titles, product types, images and attributes determine which queries a product matches, and feed improvements change what traffic arrives rather than merely what it costs.

Why the feed is the real lever

In Shopping there is no ad copy to write and no keyword to add. The only things determining whether a product shows for a valuable query are the feed attributes and the landing page.

A title that leads with the brand rather than the product type, or omits the size, colour or model, matches fewer relevant queries and more irrelevant ones. Rewriting titles to lead with the terms buyers actually search is often worth more than months of bid work.

That work is unglamorous, sits with whoever owns the product data rather than with the media buyer, and is consistently the highest-return activity in a Shopping account. Accounts that plateau on bidding almost always have feed work left undone.

Separating best-sellers into their own campaign with their own budget prevents the long tail from consuming spend that the proven products could use.

Because Shopping campaigns bid at product level within a shared budget, a poorly performing tail competes directly with the products that work, and splitting them is usually worth more than any bid adjustment within a single campaign.

The same logic applies to seasonal products, which deserve their own campaign rather than competing with year-round lines for shared budget.

Custom labels in the feed are the mechanism for that segmentation, allowing products to be grouped by margin, season or performance tier without restructuring the catalogue.