Performance teams can spend hours reconciling Meta, Google, GA4, CRM and ecommerce data and still miss the only number the business cannot debate: money received.
Each system has a different purpose. Platform attribution helps an algorithm learn. Analytics helps us understand journeys. CRM records commercial progression. Finance records economic reality. The job is to connect them without pretending they are identical.
Expect disagreement
If Meta and Google totals add up perfectly to total company revenue, something is probably wrong. Platforms use different windows, identity methods and modeling assumptions. Their totals can overlap.
The target is not perfect equality. The target is a stable, explainable relationship between sources.
Build a reconciliation waterfall
For ecommerce, I like a weekly view that starts with gross orders, subtracts cancellations and returns, identifies new-customer revenue, then compares that truth to platform-attributed revenue and analytics-attributed revenue.
For lead generation, use spend, raw leads, qualified leads, opportunities, closed-won revenue and collected revenue. The funnel makes attribution inflation visible because each stage has a real denominator.
Track the attribution gap as its own metric
If a platform reports $500,000 of revenue while net new-customer revenue associated with the period is $300,000, the ratio itself is informative. Watch how it changes over time. A sudden expansion can signal retargeting saturation, tracking changes, consent loss or duplicate credit.
Use blended economics for executive allocation
Channel-level data is needed for optimization, but executive allocation should also track Blended CAC, marketing as a percentage of revenue, contribution margin after media and Payback Period.
Those metrics make it harder for a single platform to define success on its own terms.
What finance and marketing should agree on
- Revenue definition: ordered, shipped, collected or recognized.
- New-customer definition.
- Gross-margin assumptions.
- Which refunds and discounts are included.
- What date anchors the cohort.
- How offline sales are reconciled.
- Which attribution view is used for which decision.
Once those definitions are stable, dashboards become comparable across weeks and teams. Without them, every meeting can generate a new version of truth.