A Single Source of Truth for Marketing Attribution - Is It Actually Possible?
A Single Source of Truth for Marketing Attribution - Is It Actually Possible?. Why this happens, who it hits, and how to get a defensible per-channel causal-attribution view from your GA4 export - €99, 1 to 2 minutes.
By Joris van Huët, Founder & CEOUpdated 5 min read
Run the numbers for your store: the free UTM builder.
A Single Source of Truth for Marketing Attribution - Is It Actually Possible?
Quick answer: No single dashboard can be your source of truth, but one number can: the revenue in your Shopify orders. Meta Ads Manager, Google Ads, TikTok, GA4 and tools like Triple Whale each apply a different credit rule to a different slice of data, so in the same week Meta might claim €61,950 while Shopify records €52,140. Reconcile every report to Shopify revenue, then decide what each channel caused with a holdout test, not with another dashboard.
The dashboards don't disagree because one of them is broken. Each was designed to answer a different question, with a different rule for assigning credit, on a different slice of data, and people who work with these tools say so openly: see Free To Grow CFO's warning on Shopify + Triple Whale data and Measured on the dangers of multi-touch attribution. This article walks through why every dashboard tells you a different number, why no marketing attribution dashboard can settle which channel actually drove the sale, and what to use instead.
Why none of these dashboards agree
There are three separate reasons your dashboards disagree, and they all stack on top of each other.
First, self-attribution bias. Every ad platform optimises and reports on its own conversions. Northbeam's own KB on Facebook mismatch explains Northbeam's own admission of this: Meta inflates Meta's contribution because Meta is the one counting. Same on Google. Same on TikTok. Angler AI on Northbeam vs Meta gaps quantifies typical gaps.
Second, definitional differences. Triple Whale's own KB on Shopify conversion-rate mismatch and Triple Whale's own KB on Shopify customer-count mismatch - Triple Whale's own knowledge-base articles - openly catalogue how their session definition, customer-classification logic, and conversion-rate denominator differ from Shopify's. None of it is wrong. It is just different. Triple Whale's own KB on GA4 mismatch extends the same point to GA4: each browser tab counts as a session in Triple Whale; GA4 consolidates them.
Third, the deeper problem: correlation, not causation. Measured on the dangers of multi-touch attribution lays this out cleanly. Multi-touch attribution measures which touchpoints were present in a journey, not which one caused the conversion. Two tools applying different MTA rules to the same data will return different credit splits. Neither is closer to the truth.
What this looks like in your data
A typical week for an ecommerce brand running Meta + Google + TikTok + email:
| Source | What it says it drove | What is actually true |
|---|---|---|
| Meta Ads Manager | €60,000 | Meta is counting any sale where the buyer saw any Meta ad. Inflated. |
| Google Ads | €25,000 | Google is counting last-click on Google, including post-purchase brand searches. Inflated for branded search. |
| TikTok Ads | €18,000 | TikTok is counting view-through. Heavily inflated for view conversions. |
| GA4 (last-click) | €52,000 total | Closer to honest, but blind to view-through and dependent on UTMs. |
| Triple Whale / Northbeam | €54,000 total, distributed | Distributed differently each, depending on their MTA rule. |
| Shopify (truth) | €52,000 in revenue | The total exists. |
Why no MTA tool can fix this
The fundamental issue is category. MTA tools answer the question "which touchpoints touched the journey?" with a rule. Causal inference answers a different question: "would this revenue have happened without this channel?" The two are not the same. Two MTA tools can both be technically correct and still disagree, because the question they each answer is different from the question you actually want answered.
This is why Measured on the dangers of multi-touch attribution concludes that "many companies still struggle to realize MTA's benefits, and most organisations would not recommend their current MTA providers."
What a causal model does differently
A causal approach does not apply a credit-splitting rule. It estimates the counterfactual: how much revenue would have happened without each channel. The simplest version needs no model at all. Pause or cut one channel's spend in a few regions for two weeks, keep everything else the same, and compare revenue there with the regions you left alone. The revenue you lose is what that channel caused.
That replaces competing dashboards with one answer per channel, measured in the one number they all share: your Shopify revenue. Test one channel at a time, and repeat a test before you move a large budget.
FAQ
Will the causal view agree with Meta Ads Manager? Self-attribution bias is real and it inflates Meta's self-reported contribution.
Will the causal view agree with GA4 last-click? Sometimes yes, sometimes no. Channels that sit mid-funnel (newsletter, organic social, certain influencers) are systematically under-credited by last-click. The causal model surfaces those.
What if my UTMs are inconsistent? Fix them first where you can: inconsistent UTMs send sales to the wrong channel in every report, GA4's included. A holdout test doesn't depend on UTMs at all, because it compares revenue by region.
Can I just upload my GA4 export? For a holdout test you don't need to upload anything: your ad accounts and your Shopify orders are enough. Later, if you want every channel estimated at once, a causal attribution tool like Causality Engine reads what each channel caused from a GA4 export, next to what last-click gave it.
Go deeper: Incrementality testing, explained.
Sixty-second versions of these ideas: Causality Engine on YouTube Shorts.
Have an audience of ecommerce operators? Earn 20% recurring commission by recommending the €99 read.
Keep reading
Terms in this article
- AttributionAttribution identifies user actions that contribute to a desired outcome and assigns value to each. It reveals which marketing touchpoints drive conversions.
- Causal AttributionCausal Attribution uses causal inference to determine which marketing touchpoints genuinely cause conversions, not just correlate with them.
- Causal InferenceCausal inference is the set of methods for estimating how much an outcome changes because of a cause, rather than alongside it. It works by comparing what happened with an estimate of what would have happened without the cause.
- Causal ModelA Causal Model is a mathematical representation describing the causal relationships between variables, used to reason about and estimate intervention effects.
- CounterfactualA counterfactual is what would have happened under conditions that did not occur, such as a customer's outcome had they not received the campaign they did receive. It is never observed, so every effect estimate has to stand in for it with a control group, comparable units or a forecast.
- Holdout TestA holdout test is an experiment that keeps a random part of an audience from seeing a campaign and compares its results with those of the part that could see it. The held-out part is the control group, so the gap in outcomes estimates what the campaign added.
- Marketing AttributionMarketing attribution assigns credit to marketing touchpoints that contribute to a conversion or sale. Causal inference enhances attribution models by identifying true cause-effect relationships.
- Multi-Touch AttributionMulti-Touch Attribution assigns credit to multiple marketing touchpoints across the customer journey. It provides a complete view of channel impact on conversions.