Marketing Attribution for Ecommerce: Why Meta and Google Both Claim the Same Sale

What is marketing attribution?

Marketing attribution is the process of assigning credit for a conversion, such as an ecommerce purchase, to the marketing interactions that helped produce it.

A customer might see a Meta ad on Instagram, return through an email campaign, search for the brand on Google, click a Google Shopping ad, and then complete a purchase.

Marketing attribution attempts to determine which of those interactions deserves credit for the sale.

Different attribution models answer that question in different ways. Last-click attribution gives the final interaction all the credit. First-click attribution emphasizes discovery. Multi-touch attribution distributes credit among several interactions.

No attribution model provides a completely objective version of the customer journey. Each model uses a set of assumptions to answer a particular question.

The challenge becomes greater when several advertising platforms measure the same journey independently.

Why Meta and Google Ads report different conversions

Meta and Google can both claim the same ecommerce sale because:

• Each platform sees activity within its own ecosystem.
• Their attribution windows may be different.
• They can count different types of interactions.
• Their conversion events may be configured differently.
• They may assign revenue to different dates.
• Their attribution models support their own optimization systems.

Consider a customer who discovers your brand through Meta. A few days later, the same person searches for your company on Google, clicks an ad, and makes a purchase.

Meta may claim the purchase because the customer previously viewed or clicked a Meta ad. Google may also claim it because the customer clicked a Google ad before converting.

Your ecommerce platform records one order.

All three reports can be technically correct because they are answering different questions.

Meta is asking whether a qualifying Meta interaction happened before the purchase.

Google is asking whether a qualifying Google interaction happened before the purchase.

Your store is recording how much revenue the business actually received.

These numbers are related, but they should not be treated as interchangeable.

The limits of attribution inside advertising platforms

Meta has detailed information about customer activity on Facebook and Instagram. Google has detailed information about activity across Search, Shopping, YouTube, and other Google properties.

Neither platform has a complete and neutral view of everything the customer did.

Google cannot fully observe every Meta impression or engagement. Meta cannot fully observe every Google search or shopping interaction. Each platform evaluates the purchase using the information available within its own environment.

This is the central limitation of walled-garden attribution. Each platform can see what happens inside its own system, while the customer and the business operate across several systems.

A more sophisticated attribution model can analyze the available data more intelligently, but it cannot remove the boundaries around that data.

Attribution windows create overlapping claims

An attribution window determines how long after an advertising interaction a platform can claim a conversion.

A platform might count a purchase if it occurs several days after an ad click. It may also count a purchase after someone views an ad without clicking it.

If Meta and Google use different attribution windows, their conversion reports are not directly comparable.

A longer window gives a platform more opportunities to claim later purchases. View-through attribution can create additional overlap because one platform may receive credit even when another channel produced the final click.

The appropriate attribution window depends on the customer's typical time to purchase. An inexpensive impulse product and a high-consideration purchase should not automatically be measured using the same settings.

The important thing is to understand the settings and keep them consistent enough to evaluate performance over time.

Click-through and view-through conversions are different

A click-through conversion happens after someone clicks an ad.

A view-through conversion can be attributed after someone sees an ad, does not click it, and later purchases through another route.

View-through attribution can reveal advertising influence that click-only reporting misses. It can also create more overlapping conversion claims.

For example, Meta might claim a view-through conversion after a shopper sees an Instagram ad. Google might claim the same sale after the shopper clicks a branded search ad.

Both interactions may have contributed to the purchase. However, they should not be added together as though they represent two separate orders.

Tracking and reporting differences also matter

Not every discrepancy is caused by attribution overlap. Tracking configuration can also produce inconsistent results.

Common causes include:

• Duplicate browser and server-side purchase events
• One platform counting purchases while another counts completed checkouts
• Missing or inconsistent transaction IDs
• Different account time zones
• Revenue being reported differently before or after discounts, taxes, and shipping
• Refunds appearing in store data but not in platform reports
• Test orders or subscription renewals being included in one source
• Different currencies or exchange-rate calculations

A conversion gap can contain several things at once. It may include legitimate attribution overlap, missing data, duplicate events, reporting delays, and inconsistent revenue definitions.

The gap is a signal that something needs to be investigated. It does not explain the cause by itself.

Are Meta and Google inflating their results?

It is tempting to conclude that Meta and Google are exaggerating their performance.

The reality is more complicated.

Each platform reports conversions that satisfy its attribution rules. Those reports are mainly designed to help the platform decide which campaigns, audiences, keywords, and ads should receive more budget.

The problem starts when marketers add self-attributed revenue from multiple platforms and treat the result as total business revenue.

If Meta claims $128,000 and Google claims $94,000, the combined claim is $222,000. If the store recorded $187,000, the difference is $35,000.

That does not prove that exactly $35,000 was double-counted. Tracking coverage, reporting dates, refunds, event definitions, and modeled conversions could also contribute to the difference.

What you have found is a measurement gap. More analysis is required to explain it.

Which marketing attribution number should you trust?

No single metric can answer every marketing performance question.

Business question	Best starting point
How much revenue did the business generate?	Ecommerce platform or financial records
How efficiently did total ad spend produce revenue?	Blended ROAS or marketing efficiency ratio
Which Meta campaigns perform best inside Meta?	Meta Ads reporting
Which Google campaigns perform best inside Google?	Google Ads reporting
How should budget be allocated between platforms?	Cross-channel analysis using consistent business metrics
Did advertising generate additional sales?	Controlled experiments or incrementality analysis

Your ecommerce platform or financial system should be the source of truth for actual orders and revenue.

Platform reporting remains useful for comparing campaigns within the same platform. It becomes misleading when self-attributed revenue is treated as a neutral record of total business performance.

How to build a more reliable cross-channel view

Perfect marketing attribution may not be achievable, but you can create a much better foundation for decision-making.

1. Define the business outcome

Decide exactly what counts as a purchase and how revenue should be calculated.

Clarify whether your revenue includes shipping, taxes, discounts, refunds, cancellations, subscription renewals, and returning-customer orders.

Use the same definition across your reports wherever possible.

2. Review your attribution settings

Document the attribution window and eligible interactions used by each platform.

Check whether reports include click-through, view-through, or engaged-view conversions. Do not compare two ROAS figures before checking what each one includes.

Avoid changing these settings frequently. Consistency makes performance trends easier to interpret.

3. Validate your conversion tracking

Compare several real orders across your ecommerce platform, Meta, Google Ads, and your analytics system.

Look for duplicate events, missing transaction IDs, inconsistent purchase values, and time-zone differences.

This helps you separate expected attribution overlap from a genuine tracking problem.

4. Create a unified performance table

Bring the following information together for each day or week:

• Meta spend
• Google spend
• Total advertising spend
• Meta-attributed revenue
• Google-attributed revenue
• Actual store revenue
• Orders
• New customers
• New-customer revenue
• Refunds
• Contribution margin, when available

This cross-platform view will not automatically determine which channel caused each sale. It creates a consistent foundation for identifying discrepancies and making decisions.

5. Calculate blended ROAS

Blended ROAS measures total revenue relative to total advertising spend.

Blended ROAS = Total revenue ÷ Total advertising spend

If the store generated $187,000 and spent $55,000 across Meta and Google:

$187,000 ÷ $55,000 = 3.4x blended ROAS

Blended ROAS avoids the problem of adding overlapping platform claims. It tells you whether advertising is efficient at the business level.

However, it cannot determine which platform or campaign deserves credit. It should be used as a business-level guardrail alongside platform and cross-channel analysis.

6. Separate new and returning customers

Total revenue can make customer acquisition look healthier than it really is.

An established ecommerce brand may generate significant repeat revenue even when its current campaigns are attracting fewer profitable new customers.

Where possible, track:

• Cost per new customer
• New-customer revenue
• New-customer ROAS
• First-order contribution margin
• Customer lifetime value
• Repeat-purchase rate by acquisition source

This helps distinguish campaigns that capture existing demand from campaigns that create valuable new demand.

Where attribution models fit

Different marketing attribution models are useful for different decisions.

First-click attribution helps you understand how customers initially discover the brand.

Last-click attribution identifies the final measurable interaction before purchase. It can give too much credit to branded search and other channels that capture existing demand.

Multi-touch attribution distributes credit among multiple interactions. Its usefulness depends on how successfully those interactions can be observed and connected.

Data-driven attribution uses available conversion data to estimate the contribution of different interactions. Its results still depend on the scope and quality of the available data.

Incrementality testing asks a different question. It tries to estimate how many purchases would not have happened without the advertising.

A strong measurement system does not force every decision through one attribution model. It uses the appropriate method for the question being asked.

What should marketing attribution software actually do?

Marketing attribution software should provide more than another dashboard filled with platform-reported metrics.

For ecommerce teams operating across Google, Meta, and other channels, a useful system should help them:

• Bring cross-platform advertising data into one consistent view
• Compare platform claims with actual business results
• Apply consistent metric definitions
• Identify meaningful performance changes
• Evaluate customer quality as well as conversion volume
• Support budget decisions across platforms
• Explain how recommendations were produced
• Keep marketers in control of important changes

The objective is not to create a perfect number that ends every attribution debate. It is to give marketers enough reliable context to make better decisions.

The real problem is fragmented decision-making

Marketing attribution discussions often focus on finding one final source of truth.

In practice, the more important question is how the business should respond when every platform presents a different version of performance.

Answering that question requires a layer above the ad platforms. It needs to compare their data, connect it to business outcomes, identify meaningful changes, and help marketers decide what to do next.

That is the problem Coretas is building to solve.

Coretas brings cross-platform advertising data into one performance marketing intelligence layer. It helps ecommerce teams plan, analyze, and optimize campaigns while keeping marketers in control of the final decisions.

The objective is to help the business perform better across Meta, Google, and every other connected advertising platform.