Shopping reporting defaults to ROAS at campaign level, which hides almost everything worth knowing. Revenue is unevenly distributed across products, the conversion data underneath it is more fragile than it looks, and a campaign average quietly conceals the products losing money inside a profitable campaign.
ROAS is the metric everyone reports and the one most likely to mislead, because it treats revenue as if margin were uniform across the catalogue. These cover how to calculate it properly and what to use once revenue alone stops being enough.
Before any analysis is trustworthy, the definitions have to be. Shopping carries metrics that sound interchangeable and are not, reported at grains that do not always line up.
Account-level numbers are an average of things that behave nothing alike. Splitting by brand and by query intent usually reveals that a modest overall ROAS is one strong segment carrying a weak one.
Every decision above rests on conversion data, and that data is more fragile than it looks. Attribution windows and conversion lag distort what you see; consent requirements and browser restrictions decide whether conversions are recorded at all.
Why last-click ROAS systematically misleads in Shopping, and what the alternatives change
How long conversions take to land, why recent data always looks weak, and how long to wait before judging
Recover conversions lost to consent and cookie restrictions, and move collection server-side
Most Shopping accounts leak budget through products that take clicks and never convert, and search terms nobody audits. None of it registers as a problem in campaign reporting — the averages stay respectable while the tail drains money every day.
A good ROAS varies by industry and margin. Generally, 300-400% (3-4x) is considered healthy for most e-commerce. High-margin products can be profitable at 200%, while low-margin products may need 500%+ to be viable. Calculate your break-even ROAS based on your profit margins.
Conversion lag is the delay between a click and the attributed conversion. Google Ads attributes conversions back to the original click date, which can take 1-30 days depending on your conversion window. This means recent data is incomplete and will improve over time. Always account for conversion lag when analyzing performance.
In Google Ads, navigate to Products > Product groups to see performance by product. You can also use the shopping_performance_view in the Google Ads API for programmatic access. Third-party tools like SKU Analyzer combine this with Merchant Center data for complete product analytics.
Essential metrics include: ROAS (return on ad spend), conversion rate, cost per conversion (CPA), click-through rate (CTR), impression share, and average CPC. At the product level, track cost, revenue, conversions, and profitability to identify winners and losers.