The Pattern
An online store runs ads on several platforms and also watches Google Analytics and the store admin. Each shows a different number of sales for the same week. Meta claims one figure, Google Ads another, analytics a third, and the store records fewer than all of them. Budget decisions stall because nobody trusts the totals.
What Teams Usually Try First
They typically pick one dashboard as the truth, add a third-party attribution tool, or manually reconcile numbers in a spreadsheet each week.
Where It Stalls
The platforms use different attribution windows and credit rules, and several count the same order when a customer touched more than one of them. Tracking is only partly in place, consent choices hide some visitors, and order data from the store is not compared with platform data on a regular basis, so the expected gap is never defined.
Recommended Approach
- Treat the store's own order data as the source of truth for total sales and revenue, and use the platforms to explain where those orders came from
- Write down each platform's attribution window and counting rule side by side, so everyone understands why the numbers differ before anyone tries to force them to match
- Audit tracking end to end: purchase events firing once per order, correct values and currency, consent handling, and server-side or conversion API setups where used
- Create one weekly reconciliation view with store orders, platform-reported conversions and the expected gap, and investigate only when the gap moves outside its normal range
- Judge channels on cost per order against store data, and use holdout or incrementality tests where spend is large enough to make the question worth answering
What It Aims to Achieve
A shared understanding of why the numbers differ and a repeatable check that flags real tracking problems, so spend decisions rest on store-verified orders rather than whichever platform reports the highest figure.