Attribution & Measurement
Why Meta and Google both claim the same Shopify sale
Meta and Google can both report a purchase as theirs, for the same order, at the same time — because each platform's attribution model only ever looks at itself. Add their reported revenue together and you're not measuring total ad-driven sales, you're counting some orders twice.
A week where the math doesn't add up
Say Shopify records 100 real orders in a week. Pull up Meta Ads Manager for the same week and it reports 70 attributed purchases. Pull up Google Ads and it reports 55 attributed purchases. Add those together and you get 125 — twenty-five more purchases than actually happened. This is a hypothetical example to illustrate the mechanic, not a claim about typical overlap on any real account; the actual split depends entirely on how much your customers cross paths with both platforms before buying.
Nobody double-charged a customer and nobody duplicated an order in Shopify. What happened is that some number of those 100 real orders — the ones where a shopper saw or clicked a Meta ad and a Google ad before buying — got counted by both platforms independently. Each platform's number is internally consistent. Adding them together is where the count breaks.
Why both platforms can honestly claim the same order
Overlapping attribution windows. Meta and Google each apply their own lookback window — commonly something like a week for clicks and a day for views on Meta, a comparable range on Google — to decide whether a purchase counts. Those windows aren't coordinated with each other. A purchase that falls inside both platforms' windows is fair game for both.
Both platforms crediting the same journey. A shopper sees a Meta ad on Instagram, doesn't click, searches the brand name on Google two days later, clicks a Google ad, and buys. Meta's model can credit the impression it served. Google's model can credit the click that immediately preceded the order. Both are describing something real about that journey — neither is lying — but only one order happened.
View-through credit inflating the count. An impression the customer never clicked can still earn attribution credit on both platforms if a purchase follows within the view-through window. Impressions are cheap and plentiful compared to clicks, so this is usually the single biggest driver of a platform's reported number running ahead of what click activity alone would explain. We cover this in more depth in view-through conversions explained.
Assisted-journey overlap. Multi-touch journeys are the norm, not the exception, for anything beyond an impulse purchase. The more touchpoints a purchase has, the more platforms have a plausible claim to it — and every platform's attribution model is built to take credit for the touchpoints it can see, not to defer to whichever channel actually mattered most.
This is normal platform behavior, not a bug
It's tempting to read a 125-order combined claim against 100 real orders as something broken — a tracking error, a duplicate-event problem, a misconfiguration worth chasing down. Usually it isn't. Meta's attribution model was never designed to know what Google saw, and Google's wasn't designed to know what Meta saw. Each platform is reporting an honest answer to a narrow question — “how much revenue can I defensibly attribute to my own ads, using only what I can observe?” — and neither platform is incentivized to net out the other's claim. The overlap is a structural feature of siloed, platform-level attribution, not evidence that either number is wrong on its own terms.
The fix isn't adding the platforms together
If summing Meta and Google's reported revenue overstates reality, the obvious next move — subtracting some estimated overlap percentage from the sum — just replaces one guess with another. The more reliable approach is to stop treating platform-reported revenue as something that should sum to a meaningful total in the first place.
Shopify's actual revenue for the period is ground truth — it's a record of what was purchased, not a model of what probably caused it. Compare that single number against total ad spend across every channel and you get a blended efficiency figure that can't be inflated by cross-platform double-claiming, because it never asks either platform to self-report in the first place. We walk through how blended ROAS and MER are calculated, and where each is the right tool, on Shopify ROAS tracking and MER vs ROAS for Shopify.
Platform-reported ROAS still has a job to do — comparing this week's Meta number to last week's Meta number is a reasonable way to judge whether a campaign is trending better or worse, because the same model bias applies consistently across both weeks. What it can't do is tell you total ad-driven revenue once you're running more than one platform, and it shouldn't be asked to.
What's actually worth watching
A combined platform claim of 125% of real orders isn't itself a red flag — the useful question is whether that ratio holds roughly steady from week to week. If Meta plus Google usually runs around 20–30% over Shopify's actual order count and one week it jumps to 60% over, something changed: an attribution window got widened, a new campaign is leaning heavily on view-through-eligible placements, or a duplicate-event issue is inflating one platform's count on its own. Chasing an exact match between the platforms and Shopify isn't realistic by design — watching the gap for sudden movement is.
Where Attribix fits
Attribix pulls Meta and Google's reported numbers alongside actual Shopify orders and calculates blended ROAS from the Shopify total, so the gap between what each platform claims and what the store actually made is something you can see directly instead of reconciling by hand. See how it connects orders to campaigns on Shopify attribution.
Common questions
Why do Meta and Google both show revenue for the same order?
Should I just add Meta's and Google's reported revenue together to see total ad-driven sales?
Is this a tracking bug I should try to fix?
Does turning off view-through attribution solve the overlap?
Want this checked on your own account?
Send an inquiry and we'll look at how much Meta and Google's reported revenue actually overlaps on your store, and what your real blended efficiency looks like.