Every marketing attribution model overcounts — because every channel claims credit for purchases that multiple channels influenced, and the sum of all channel-reported revenue typically exceeds total Shopify revenue by 40 to 120 percent. Advize is an AI-powered performance marketing agency that uses three attribution-neutral measurement methods alongside channel-level data for every client, because the decision a brand makes based on inflated channel attribution — increasing spend on the channel with the best reported ROAS — is frequently the decision that most damages real ROI.
Why does every marketing attribution model overcount channel performance?
Every marketing attribution model overcounts because it assigns credit to a channel based on its role in the customer's journey — but multiple channels can all claim to be the decisive touchpoint for the same purchase. A customer who saw a Meta ad, clicked a Google ad, received an email, and then purchased through direct traffic has generated four claimed conversions in four different channel reports — but only one purchase in Shopify. This double-counting is not a bug in any specific platform — it is a structural feature of any multi-channel attribution approach.
How much does attribution overcounting typically inflate reported ROAS?
Attribution overcounting inflates reported ROAS by 40 to 120 percent for most multi-channel DTC brands, according to Advize's cross-channel attribution audit data from 30 DTC accounts. The specific inflation level depends on the number of active channels and the attribution windows used. A brand with Meta (7-day click, 1-day view), Google (30-day click), email, and organic all running simultaneously will see the highest overcounting — each channel claims the same purchase through different attribution windows. A brand running a single paid channel with only click-through attribution will see much lower overcounting.
What are the 3 attribution-neutral measurement approaches that measure real marketing performance?
Three measurement approaches produce more accurate performance data than any single attribution model.
1. Marketing efficiency ratio (MER): total Shopify revenue divided by total ad spend across all channels. This number cannot be inflated by attribution because it uses only two inputs — one from Shopify, one from your ad accounts. MER of 3.5 means $3.50 of Shopify revenue for every $1 of total ad spend, regardless of which channel claims credit.
2. Shopify UTM-based last-click attribution: pull revenue by UTM source in Shopify Analytics. This understates channel contributions (it only credits the last click) but provides a consistent, non-inflatable channel comparison that identifies relative performance without double-counting.
3. Incrementality testing: pause a channel's spend by 50 to 100 percent for 14 days and measure the impact on total Shopify revenue. The revenue that disappears when the channel is paused is the channel's true incremental contribution.
How do you calculate MER and use it to make better budget allocation decisions?
Calculate MER weekly: take total Shopify revenue for the week and divide by total paid ad spend across all channels for the same week. Track MER as a time series. When MER is rising, the marketing programme is becoming more efficient. When MER is falling, the programme is becoming less efficient — and the cause needs to be identified by looking at channel-level trends, not by trusting individual channel ROAS numbers. Use MER targets to set total budget allocation: if the MER target is 3.0, the maximum total ad spend for a week is total Shopify revenue divided by 3.
Conclusion
Attribution is a solved problem in theory and an unsolvable problem in practice — because human purchasing behaviour involves multiple touchpoints, multiple channels, and multiple time periods that no single attribution model can accurately reconstruct. The practical solution is not to find a better attribution model. It is to use attribution data for directional channel comparison, use MER for total programme evaluation, and use incrementality testing to measure the real contribution of individual channels when budget allocation decisions require it.
