When Meta, Shopify and GA4 Disagree, Which Number Should You Trust?
This is a familiar situation for D2C founders.
Meta says one thing.
Shopify says another.
GA4 gives you a third number.
The agency reports one ROAS.
The finance team reports another.
Now nobody wants to increase spend because nobody is sure which dashboard represents reality.
The problem isn't necessarily broken tracking.
It may be different attribution systems answering different questions.
Why the Same Purchase Can Get Different Channel Credit
Imagine a customer sees your Instagram ad on Monday.
They don't click.
On Wednesday they search your brand on Google.
They click.
They browse your website.
On Friday they return directly.
Then they purchase.
Which channel gets the sale?
Meta may have a claim based on its own measurement.
Google Analytics may assign credit according to its configured attribution model.
Shopify may present marketing attribution using its own models and dimensions.
There is no universal dashboard that magically knows the one true cause of a purchase.
Google defines attribution as the process of assigning credit for a conversion across interactions in the customer's path, and GA4 currently allows marketers to compare data-driven attribution with last-click approaches.
Shopify likewise provides multiple attribution models in its marketing reports, including last click, last non-direct click, first click, any click and linear approaches.
So different numbers are not automatically a tracking failure.
They can be a consequence of different measurement logic.
Which Number Is “Correct”?
Answer: The number that answers the specific business question you're asking.
That's the part many founders miss.
If you're asking:
“How did my Meta campaigns perform according to Meta's measurement?”
Look at Meta.
If you're asking:
“What revenue did my Shopify store actually record?”
Look at Shopify.
If you're asking:
“How does Google Analytics distribute credit across channels?”
Look at GA4.
If you're asking:
“Is the entire business acquiring customers profitably?”
You need a blended view.
That can include:
Total revenue
Total marketing spend
Blended CAC
Contribution margin
New customers
Repeat customers
Now you're asking a business question rather than a platform question.
Why Shopify Attribution Can Differ From Other Dashboards
Shopify explicitly notes that sales attributed to marketing can differ from other sales figures because its report includes sales that can be directly attributed to trackable marketing efforts. It also allows different attribution models to change how credit is assigned.
That means a founder should not simply compare:
Shopify attributed revenue
with:
Meta attributed revenue
and conclude that one is “wrong.”
Start by identifying:
What attribution window?
What attribution model?
What traffic source rules?
What conversion definition?
What date range?
Without that context, the comparison can be misleading.
How GA4 Changes the Picture
GA4 currently offers data-driven attribution as well as paid-and-organic last-click and Google-paid-channels last-click in its attribution reporting. It also allows users to compare attribution models side by side.
Data-driven attribution is particularly different from a simple last-click approach because Google says it uses account data to estimate the contribution of interactions along the conversion path.
That means the same customer journey can be valued differently depending on the model.
So when your agency says:
“Meta generated ₹20 lakh.”
and finance says:
“We only see ₹16 lakh attributed to Meta.”
the productive next conversation is not:
“Someone is lying.”
It's:
“Which attribution methodology produced each number?”
That's a much better starting point.
What Should a D2C Founder Use as the Business Scoreboard?
Answer: Your ecommerce backend plus a blended marketing view should anchor business decisions, while platform-specific attribution should be used for channel optimisation.
The distinction matters.
Meta is excellent for optimising Meta.
It shows you:
Creative performance
Spend
CPM
CTR
Clicks
Purchases
Platform-attributed revenue
That information is useful.
But a founder should also know:
Total store revenue
Total paid media spend
Blended CAC
Contribution margin
AOV
New customer revenue
Repeat revenue
Returns
Those numbers tell you whether the business is actually getting healthier.
That's the difference between performance reporting and business reporting.
Why Platform ROAS Should Not Be Used as the Sole Scaling Decision
Imagine Meta says:
4.8X ROAS
But:
Shopify revenue is flat.
Blended CAC is rising.
Google revenue is falling.
Organic traffic is declining.
Repeat purchases are weakening.
Would you immediately increase Meta spend?
Probably not.
Now imagine Meta reports:
3.2X ROAS
but:
Total revenue is growing.
New customers are increasing.
Blended CAC is stable.
Contribution margin is improving.
Would you automatically turn Meta off?
Again, probably not.
That's why attribution should inform decisions, not replace business judgement.
For Arlox, this sits directly inside the idea of scientific advertising.
We want the team to know:
What happened?
Why did it happen?
What evidence supports that conclusion?
What should we test next?
What Should You Do When Dashboards Disagree?
Create one internal reporting framework.
Start with:
Business layer
Actual Shopify revenue.
Actual orders.
Actual refunds.
Actual returns.
Contribution margin.
Acquisition layer
Meta spend.
Google spend.
Creator spend.
Other marketing spend.
Platform layer
Meta-attributed revenue.
Google-attributed revenue.
Other platform-reported conversions.
Blended layer
Total marketing spend ÷ total revenue.
New-customer CAC.
Revenue per new customer.
Contribution after acquisition.
This prevents teams from arguing over individual numbers without context.
It also gives founders a consistent scoreboard for India, UAE, UK and US markets.
Can Better Attribution Fix a Bad Business Model?
No.
Better measurement tells you what's happening more clearly.
It doesn't make an unprofitable product profitable.
That's why attribution should ultimately connect to unit economics.
A campaign reporting 5X ROAS is not automatically valuable.
The question is:
What did those customers contribute after product cost, shipping, payment fees, returns and other variable costs?
That's why the BROAS Ecommerce Profit Intelligence tool is useful alongside platform reporting. The current tool is built around inputs such as AOV, product costs, shipping, fulfilment, payment fees and marketing budget, and provides projected profit, break-even ROAS and scaling ROAS outputs.
Attribution tells you where credit went.
Unit economics tell you whether the result was worth buying.
What Does a Good Reporting System Actually Look Like?
Not a dashboard with 70 metrics.
A useful weekly founder report can fit on one page.
Revenue
Orders
Marketing spend
Blended CAC
Meta CAC
Contribution margin
Meta-attributed revenue
New customer percentage
Return rate
Top product contribution
Then underneath:
What changed?
Why?
What are we testing next?
That's enough to make much better decisions.
The aim isn't perfect attribution.
Perfect attribution does not exist.
The aim is consistent measurement that leads to better decisions.
If your Meta, Shopify and GA4 reports are constantly telling different stories, book a strategy call with Arlox.io. You can also use BROAS to bring the economics behind those numbers into the conversation.
Written by -
Evyan Kumar is Head of Marketing & Brand Growth at Arlox.io — a scientific advertising agency helping D2C fashion brands scale profitably on Meta. Based in Gurugram, India.
Arlox is a performance marketing agency for D2C fashion brands, built on scientific angle testing — structured hypothesis testing across Meta, Google, TikTok and WhatsApp, measured against contribution margin rather than platform ROAS. 450+ brands worked with, with hundreds of on-camera founder interviews on record. Founded by Varinder Singh Gakhar (Vann Laniakea).