Here's a pattern we see in almost every D2C account we audit. Google Ads reports a 6× ROAS. Meta reports a 3× ROAS. Shopify says total revenue is less than the two added together. Everyone knows something's off; nobody knows what to cut.
The answer is nearly always the same: Google is converting demand that Meta created, and reporting it as its own. Here's how to find out whether that's true for you, and what to do about it.
Why Google's number is inflated
Three things inside a typical ecommerce Google account are claiming credit they didn't earn.
- Brand search. Someone sees a Meta ad, googles the brand name the next day, clicks the paid ad above the organic result. Google records a conversion at a spectacular ROAS. They were coming anyway.
- Performance Max. Left as one campaign with the whole catalogue in it and brand terms included, PMax drifts towards the cheapest conversions it can find, which are brand searches and retargeting. The blended ROAS looks great; the new-customer share is tiny.
- Shopping on product-name searches. "Herd Mentality board game" is not a discovery search. It's someone who's already been told about the product, by a Meta ad, a friend, or a review.
A three-week test you can run yourself
- Split brand from non-brand. Move brand terms into their own campaign with their own budget, and exclude them from PMax. Report the two separately from now on. This alone usually halves the "real" Google ROAS.
- Split new from returning. Turn on new-customer reporting in Google (and in Shopify). Look at what share of Google's conversions are first-time buyers. Under 30% and Google is mostly harvesting.
- Read the overlap. In whatever attribution tool you have (GA4 at a minimum, a third-party platform if you can), look at paths that include both a Meta touch and a Google touch. If most Google conversions have a Meta ad earlier in the path, Meta made the sale and Google collected it.
If you want a harder answer, run a geo holdout: pause non-brand Google in a couple of regions for two weeks and watch whether total revenue in those regions actually drops. Often it barely moves.
What to do with the answer
None of this means turn Google off. It means run it as the second channel it is, with a job for each part:
| Campaign | Job | Judge it on |
|---|---|---|
| Brand search | Protection: stop competitors bidding on your name | Impression share, not ROAS |
| Non-brand search | Acquisition: high-intent category searches | New-customer CPA |
| Shopping / PMax by margin | Sell the hero products profitably; defend the rest | Contribution margin, split by product group |
And feed Google what it actually needs: a clean product feed. Titles that lead with what people search for, categories that are right, custom labels by margin so the structure survives Google's restructures. For Happy Camp3r, feed and structure work took Google to a 4.39× ROAS that we could stand behind, inside a combined Meta and Google account growing 118% year on year.
Why your agency won't tell you this
Because the inflated number is the one that renews the retainer. An agency paid to manage Google has no incentive to report that a third of its conversions belong to Meta. We run both channels and we're paid on verified revenue, so we'd rather know. Every account we take on gets an independent attribution platform alongside the platform reporting, and the weekly report shows platform-claimed and verified side by side. More on that on the attribution page.
Quick check: add your Meta-reported revenue to your Google-reported revenue. If the total is more than your Shopify revenue, you're double counting, and the difference is roughly how much Google is taking credit for.
Want to know what Google is really adding?
Send us access to both accounts and Shopify. We'll come back with the split before we speak.