Why the sales your ad platform counts are not your real orders
The conversion count reported by Google and Meta is almost always higher than the orders you actually received. Modelling, view-through conversions, ad blockers and double counting: where the gap comes from, and which number to base budget decisions on.
The ad report says 180 conversions; the panel shows 112 orders from ads that week. The gap is not a bug; the two numbers count different things. A store that raises its budget based on the platform's figure may be scaling a campaign that does not actually pay for itself. This article explains where the gap comes from and which number your budget decision should rest on.
Where does the gap come from?
View-through conversions. A customer who saw the ad but did not click, then came to the site on their own and bought, is counted as a conversion by the platform. For the store, that is an organic sale.
Modelling. Because of browser restrictions and privacy settings the platform cannot measure some conversions and adds "estimated" ones. That order may never have happened.
Double counting. If the same customer clicked both a Google ad and a Meta ad, both platforms claim the order. In your panel there is one order.
Ad blockers. A gap in the opposite direction: when the measurement code in the browser is blocked, a real order never reaches the platform. This time the platform undercounts and you see less than reality.
How do you get the real number?
Two things are needed. First, the source of every order has to be recorded on the store side: ad click, organic search, direct visit, e-mail or app. This record should be kept even if you never connect an ad account; it is the only way to see the channel breakdown. Second, server-side conversion sending: the purchase event is sent to the platform from the server, independently of the browser (Meta CAPI, GA4 Measurement Protocol). Even with an ad blocker the real order reaches the platform, and the platform's optimisation is fed with correct data.
On top of that, when you connect your Google Ads and Meta accounts you get the "what the platform says vs. what really happened" table: for each platform, spend, clicks, the platform's conversions and ROAS next to the store's real orders, real revenue and real ROAS. Orders from the mobile app land in the same ledger; if a customer came from an ad to the site and switched to the app, the campaign information travels with them.
Which number decides the budget?
The real-orders column. If real ROAS (revenue from ads divided by ad spend) is below 1, the campaign is not paying for itself, even if the platform shows 3. Cost per order is the only meaningful figure to compare with your product margin. Platform numbers are necessary for optimisation, not for budget decisions.
Three practical steps
1. Turn on server-side conversions; stop relying on browser measurement as the single source. 2. Use consistent UTM parameters on campaign links; read the channel breakdown weekly. 3. Bring ad-acquired customers back with segment and push campaigns; you pay no ad money for the second order.
In Ikost this measurement model is part of the panel: the order source is recorded on every order, server-side conversions are included in every plan and the Ad Performance report shows spend next to real orders. For the setup steps see the "Ad measurement" guide on the Guides page.
Share:
