Why do Meta, Google and your store show different sales?

Your store counts orders: every completed order is in the back end, that is your store’s admin, wherever the person came from. Google Ads and Meta count something else, namely purchases they attribute to one of their own ads. This assignment is called attribution, and each platform has its own rules for it.

On top of that, Google and Meta only see their own touchpoints. If someone first saw a Meta ad and later clicked a Google ad, both platforms can count the same purchase for themselves. The sum of the ad accounts therefore never replaces the number in your store.

§ 25 TDDDG, a German law on privacy in digital services, allows information to be stored on users’ devices, or read from them, only with their consent. There are exceptions, among others for what is strictly necessary for a service the user has expressly asked for to work⁠¹. So stores ask for consent to advertising and statistics cookies through the cookie banner; your lawyer decides what counts as strictly necessary in your store.

For your numbers this means: if someone declines, the order still happens in the store. But the ad accounts do not see this purchase, or see it only partly. The gap grows with the share of people who decline, not with the quality of your ads.

For this, Google offers Consent Mode, a consent setting for its measurement scripts. In the basic version, Google’s scripts send nothing at all without consent; in the advanced version, they send signals without cookies, from which Google estimates missing values⁠⁷. Estimated values are not counted orders. So when you compare, note which version runs in your store.

An example with round numbers, not from a real store: the store reports 100 orders in a week; Google Ads and Meta together report 84 purchases. That sounds like a gap of 16. But the real question is how many orders the ad accounts actually attribute to an ad click.

Whether purchases after merely viewing an ad count is a choice, not an error. What matters is that everyone involved uses the same definition. Duplicate reports, on the other hand, are a setup error, and that can be fixed.

The list brings every source onto the same period and the same definition. Work through it once in full; after that, the weekly comparison with the template below is enough.

Fill in the template on the same day every week, at the earliest 48 hours after the last day of the week. Google recommends waiting for this processing time before you compare final results⁠².

After four weeks you will see whether the difference is stable. If it suddenly jumps, look first for changes made that week: a new theme, a new cookie banner, a new app in the checkout.