Marketing Truth vs Marketing Claims: Your ROAS Is Not Your ROAS

Marketers often see very different ROAS numbers for the same campaigns across Meta, Google Analytics 4 (GA4), Shopify, and CRM systems. Meta may report a 5x ROAS while GA4 reports 3.8x and Shopify reports 4.2x. Which number is correct?

The answer is not always as simple as choosing the highest or lowest figure. Each platform can use different attribution methodologies, conversion definitions, attribution windows, identity signals, revenue sources, and reporting rules. Even the definition of advertising spend can differ.

ROAS is not a universal number. It is the output of a measurement methodology.

Understanding why these numbers differ—and establishing a consistent business-level measurement framework—is essential for making confident decisions about marketing budgets, campaign optimization, and sustainable growth.

ROAS is commonly calculated as:

ROAS = Attributed Revenue ÷ Advertising Spend

The formula is simple. The measurement behind it is not.

Two platforms can use the same formula and still report different ROAS because they may define revenue, conversions, eligible touchpoints, attribution windows, customer identity, and advertising spend differently.

For example, one system may calculate revenue using gross online sales, while another may exclude refunds and cancellations. One platform may consider only media spend, while a business-level calculation may include agency and other marketing costs.

Before asking which ROAS is correct, marketers should first ask:

Are both systems using the same revenue definition?

Are they using the same advertising spend?

Are they counting the same conversions?

Are they using the same attribution model?

Are they using the same attribution window?

Are they resolving customers and devices in the same way?

Are both datasets equally complete and finalized?

The real challenge isn't calculating ROAS. It's defining what the ROAS calculation actually represents.

Discrepancies in ROAS metrics do more than cause confusion—they directly impact how marketing budgets are allocated and campaigns optimized. When one platform signals underperformance while another indicates profitability, budget misdirection follows. This misalignment can stunt growth opportunities and reduce marketing efficiency. Without a consistent ROAS view, teams struggle to justify spend, forecast outcomes, or improve campaigns with confidence. Recognizing and addressing the root causes of these conflicts is vital to responsible marketing investment and maximizing return.

A customer rarely interacts with only one marketing touchpoint before making a purchase. A typical journey might include a paid social ad, a Google search, an email, a website visit, and finally a purchase.

Different analytics and advertising systems can assign credit to those touchpoints using different attribution methodologies. Google Analytics 4 currently uses data-driven attribution by default for paid and organic channel reporting, while other reporting configurations can use last-click attribution. Shopify also supports multiple attribution models depending on the report and configuration.

As a result, the same purchase can contribute different amounts of attributed revenue to different platforms.

The important point is not that one platform is necessarily wrong. Each platform may be answering a different attribution question using the signals available to it.

For business-level reporting, marketers should therefore define a consistent attribution methodology rather than assuming that every platform's reported ROAS is directly comparable.

Attribution windows determine how long after a marketing interaction a conversion can still be credited to that interaction.

A customer who purchases one day after clicking an advertisement may be attributed differently from a customer who purchases several weeks later. Longer windows can capture more delayed conversions, while shorter windows may exclude some of those interactions.

Attribution windows should therefore be evaluated alongside the customer journey and sales cycle. A short-window ecommerce purchase journey may require a different measurement approach from a B2B business where customers take weeks or months to convert.

Marketers should also distinguish between an attribution window and an attribution model. The window determines which interactions are eligible for credit; the model determines how credit is distributed among those interactions.

Comparing ROAS without considering both can lead to misleading conclusions.

The speed at which platforms update their reports varies, impacting ROAS accuracy in real-time comparisons. Meta’s ad data may refresh within hours, whereas CRM and Shopify data often lag due to offline purchase processing and payment reconciliation procedures. GA4 can also experience delays depending on its reporting pipeline. These latency differences mean ROAS numbers represent different data completeness and timeframes, complicating immediate decision-making.

Invalid traffic—such as fraudulent clicks and bot activity—distorts ROAS by inflating ad cost without genuine conversions. Platforms handle fraud detection variably; some apply robust filters, others do not. Failure to exclude fraudulent interactions leads to overstated ROAS and misinformed channel assessments, increasing the risk of inefficient spend.

Privacy restrictions and browser policies have reduced third-party cookie reliability, affecting how platforms track users and attribute conversions. Meta, GA4, and Shopify each manage signal loss differently, from integrating server-side tagging to modeling missing data. Such varying approaches introduce discrepancies in conversion tracking and subsequently in ROAS reporting.

Meta, GA4, Shopify, and CRM systems are designed around different data sources and measurement objectives.

Advertising platforms primarily measure interactions with their advertising environments. Analytics platforms collect customer and event activity across websites and applications. Commerce platforms maintain transaction and order information, while CRMs can connect marketing activity with customer, sales, and offline revenue data.

These systems can therefore differ in:

Event collection

Identity resolution

Conversion definitions

Attribution logic

Revenue sources

Reporting windows

Data processing and reconciliation

Even when multiple systems observe the same customer journey, these differences can result in different ROAS calculations.

Attribution isn't the only reason ROAS differs. The definition of revenue can also change the calculation.

Consider a business that generates ₹10 lakh in gross sales. After discounts, cancellations, refunds, and other adjustments, the revenue used by finance may be significantly lower.

One system might report gross ecommerce revenue, while another uses adjusted revenue. A CRM may also include offline purchases or repeat customer transactions that aren't available to an advertising platform.

Before comparing ROAS, marketers should establish whether revenue means:

Gross sales

Net sales

Revenue after discounts

Revenue after refunds and cancellations

Online revenue only

Online plus offline revenue

New-customer revenue

Total customer revenue

Different revenue definitions can produce different ROAS even when attribution is identical.

The denominator of ROAS can also vary.

An advertising platform may calculate ROAS using only the media spend recorded within that platform. A business-level measurement framework may include additional marketing costs such as agency fees, creative production, influencers, or other channel expenses.

For example, suppose a campaign generates ₹10 lakh in attributed revenue.

If the platform reports ₹2 lakh in advertising spend:

Platform ROAS = ₹10 lakh ÷ ₹2 lakh = 5x

But if the business incurred another ₹1 lakh in related marketing costs:

Business-level ROAS = ₹10 lakh ÷ ₹3 lakh = 3.33x

Neither calculation is necessarily wrong. They simply answer different questions.

Always define what is included in marketing spend before comparing ROAS across systems.

A conversion can mean different things to different systems.

An advertising platform may record a purchase event. A commerce platform may record an order. A CRM may count only a confirmed transaction, while finance may recognize revenue after refunds, cancellations, or other adjustments.

These differences matter because ROAS depends on which conversions are included in attributed revenue.

Before reconciling ROAS, define:

What event represents a conversion?

Are cancelled orders removed?

Are refunded orders removed?

Are duplicate events removed?

Are offline conversions included?

Are repeat purchases included?

If platforms aren't counting the same conversions, their ROAS numbers should not be expected to match.

Consider a customer who sees a Meta advertisement, later searches for the brand on Google, visits the website through an email campaign, and finally makes a ₹10,000 purchase.

Different systems may assign different amounts of revenue to the journey:

| Measurement system | Revenue credited | Advertising spend | Reported ROAS |

|---|---:|---:|---:|

| Meta | ₹7,000 | ₹2,000 | 3.5x |

| GA4 | ₹5,000 | ₹2,000 | 2.5x |

| Shopify | ₹6,000 | ₹2,000 | 3.0x |

| Unified business model | ₹4,500 | ₹2,000 | 2.25x |

The numbers are different because the systems may use different attribution rules, eligible touchpoints, identity signals, and measurement definitions.

The goal isn't necessarily to force all four systems to report the same number.

The goal is to understand why they differ and determine which measurement should drive business-level decisions.

Not all ROAS metrics answer the same business question.

Platform ROAS asks:

*How much revenue did this advertising platform attribute to its activity?*

Channel ROAS asks:

*How much revenue did this channel generate under our chosen attribution framework?*

Blended ROAS or MER asks:

*How much total revenue did the business generate relative to total marketing spend?*

Incremental ROAS (iROAS) asks:

*How much additional revenue was actually caused by the marketing investment?*

These metrics can all be useful, but they should not be treated as interchangeable.

Platform ROAS is useful for platform optimization. Business-level measurement is needed for budget allocation and strategic investment decisions.

A campaign reporting 6x ROAS may look like an obvious candidate for additional budget. But attributed ROAS does not necessarily measure incremental or profitable revenue.

Suppose some customers would have purchased without the advertisement. If the platform receives credit for those purchases, reported ROAS can remain high even though the campaign generated relatively little incremental revenue.

The calculation can also be affected by:

• Repeat purchases
• Refunds and cancellations
• Discounts
• Existing customer demand
• Organic traffic
• Unaccounted marketing costs
• Attribution overlap between channels

This is why marketers should distinguish between attributed ROAS, profitability, and incremental impact.

A high ROAS is useful—but it is not automatically proof that increasing spend will generate proportional incremental revenue.

Conflicting ROAS metrics become a business problem when teams use different numbers to make budget and growth decisions.

If one platform reports strong performance while another shows weak performance, marketers may increase investment in one channel while reducing another without understanding the underlying measurement differences.

This can affect:

Budget allocation

Campaign optimization

Revenue forecasting

Executive reporting

Channel investment

Growth planning

The objective is therefore not to eliminate every numerical difference between platforms. It is to establish a consistent measurement framework that allows marketers to interpret those differences and make better decisions.

Start by listing the attribution models each platform uses by default and alternatives available. Understand differences among last-click, first-click, linear, time decay, and data-driven approaches to interpret ROAS divergences meaningfully.

Assess each platform’s data freshness, fraud detection robustness, and latency effects. Adjust your analysis to account for delayed reporting or invalid traffic that could skew ROAS metrics.

Select an attribution model and conversion window aligned with your business objectives. Apply these consistently across platforms for integration or comparison to minimize discrepancies and improve decision clarity.

Focus on metrics that reflect your sales cycle and customer journey. For complex or longer sales cycles, prioritize multi-touch attribution and extended windows to capture revenue velocity accurately.

ROAS measurement requires ongoing auditing. Regularly review data consistency, attribution effectiveness, and fraud detection to adapt as your channels, products, or privacy constraints evolve.

ROAS reconciliation does not mean forcing Meta, GA4, Shopify, and CRM systems to report identical numbers.

Each platform has its own measurement environment, data availability, attribution methodology, and reporting purpose.

Instead, marketers should establish a business-level measurement framework and use platform-specific ROAS for what it does best: optimizing activity within that platform.

The important question isn't:

"Which platform has the correct ROAS?"

It is:

"Why do these numbers differ, what does each number represent, and which measurement should guide our business decision?"

Relying only on last-click or one platform’s metrics can obscure the full customer journey, undervaluing upper-funnel or assisted touchpoints critical to conversions. Use multi-touch attribution and multiple data sources for a complete picture.

Making decisions based on partial or delayed data skews optimization. Incorporate real-time event tracking solutions wherever possible for timely, accurate insights.

Ignoring invalid traffic inflates spend and distorts ROAS metrics. Implement strong fraud detection practices and consistently filter suspicious activity to maintain trustworthy data.

Key metrics include:

• ROAS — attributed revenue relative to advertising spend
• MER — total revenue relative to total marketing spend
• CAC — customer acquisition cost
• CPA — cost per acquisition
• CLV/LTV — customer lifetime value
• Contribution margin — revenue after variable costs
• New-customer ROAS — efficiency of acquiring new customers
• Payback period — time required to recover acquisition costs
• Incremental ROAS — additional revenue generated because of marketing activity

> ROAS measures attributed efficiency. It does not by itself measure incrementality, profitability, or long-term customer value.

Marketing measurement is not a one-time implementation. Customer journeys, privacy environments, advertising platforms, tracking infrastructure, and business models continuously change.

Regularly monitor:

• Attribution differences
• Identity match rates
• Event completeness
• Conversion discrepancies
• Revenue reconciliation
• Refund and cancellation rates
• Fraud signals
• Reporting latency
• Changes in attribution methodology

Periodic reconciliation helps ensure that marketing decisions continue to rely on consistent and trustworthy measurement.

DriveMetaData brings together customer, advertising, web, mobile, CRM, commerce, and offline interactions into a unified customer data layer.

By connecting these interactions through identity resolution, marketers can build a more complete view of the customer journey and apply a consistent attribution framework across channels.

This provides a common measurement foundation for comparing platform-reported performance with business-level attribution.

With AI-driven fraud detection, DriveMetaData identifies and filters invalid traffic and anomalies, improving the accuracy and reliability of attributed conversions and ROAS metrics.

The platform’s real-time event processing reduces reporting delays and provides marketers with fresh, harmonized campaign insights, enabling faster and more informed budget decisions.

Your advertising platforms tell you what they can attribute. Your business needs to understand what actually drives revenue.

DriveMetaData helps unify customer identity, marketing events, attribution, revenue, and fraud signals into a consistent measurement layer—giving marketing teams a clearer view of channel performance and a stronger foundation for budget decisions.

Move from platform-reported ROAS to decision-grade marketing measurement with DriveMetaData.