Why Meta, Google, Shopify and Klaviyo report different revenue

Which revenue number should you use?

Use your reconciled order or accounting records to establish what the business sold. Use attribution reports to investigate how marketing may have contributed. Those are different questions, so the numbers do not need to be identical.

A customer can see a Meta ad, search your brand on Google, open an email and then order. More than one platform can credit that same order under its own rules. Adding their attributed revenue creates a total the business never actually earned. A third-party attribution platform offers another consistent view of credit; it does not turn estimated contribution into proven causation.

Start by writing the decision you need to make: whether to increase acquisition spend, change an email programme or fix missing purchases. Our measurement stack guide explains the role of each data source.

Align definitions before comparing dashboards

Choose one completed date range, reporting timezone and currency. Record whether each source uses order time or ad-interaction time. Then check whether revenue includes tax, shipping, discounts, refunds, cancelled orders, wholesale sales or subscription renewals. A gross sales report and a net sales report can both be correct while disagreeing.

Next, write down each attribution model, click window and any view or email-open credit. Record the settings actually in use, rather than assuming a default. Identify whether the report is about all customers or new customers. Keep unknown settings visibly unknown.

Shopify documents several reasons analytics systems disagree. Klaviyo separately documents how message attribution windows determine credit. Use those definitions to interpret the gap before treating it as a tracking fault.

Trace a small set of real orders

Select a mixture of new customers, returning customers, refunded orders and purchases with more than one marketing interaction. Use internal order IDs to compare the order system, analytics event and available platform records. Keep customer details in the authorised systems.

Check that each purchase fires once, carries the correct value and currency, and uses a stable transaction identifier. Where browser and server events are both sent, inspect the integration's deduplication behaviour. An accepted event is evidence of delivery, not proof that every report will assign it the same credit.

Classify each difference as a definition difference, attribution difference, collection problem or unresolved gap. Our attribution gap diagnostic helps organise the investigation. Avoid inventing a correction multiplier to make the charts agree.

Use a weekly reconciliation process

1. Finance or ecommerce owner: confirm actual net sales and any excluded revenue.
2. Media owner: reconcile channel spend over the same window.
3. Analytics owner: investigate missing or duplicate orders and record open issues.
4. Team: read each attribution view separately, then choose one budget decision.
5. Decision owner: log the change, expected effect and review date.

For an illustrative store with $100,000 of defined net sales and $20,000 of included marketing spend, blended revenue divided by spend is 5.0. If Meta credits $70,000 and Google credits $50,000, that does not increase actual sales to $120,000. The blended ratio also includes demand from returning customers and other sources; it is not a causal return on advertising.

When does an attribution platform help?

Consider one when an established ecommerce business repeatedly makes meaningful cross-channel budget decisions and the existing reporting cannot explain enough of the customer journey. First confirm the core order data, product costs and channel connections are reliable. Otherwise, the new dashboard inherits the same weak inputs.

Compare Triple Whale and Northbeam against a written use case: new-customer acquisition, cross-channel allocation, product contribution or reconciliation time. Ask about supported stores, returns, currencies, attribution definitions, history limits, exports, API access and the total contract. Product capabilities and prices change, so verify the current Triple Whale and Northbeam packages.

Run a reporting pilot with an agreed success criterion before making it an agency standard. Avoid changing bidding, tracking providers and attribution software at once. For service businesses, connecting qualified leads and won jobs through the offline conversion process may be the more useful first investment.

Keep attribution and incrementality separate

Attribution distributes credit among observed interactions. Incrementality asks what would have happened without the marketing. A campaign can have strong attributed revenue while capturing customers who would have purchased anyway. It can also help create demand that another channel receives credit for.

Use suitable controlled experiments to test major causal questions where the business has enough volume and a viable design. Record exclusions, conversion lag and uncertainty. A change in the dashboard after a budget adjustment is useful evidence, but seasonality, promotions and other changes can also explain it.

Questions to settle before the next report

Should email revenue match GA4?

No. Check both systems' attribution rules and event collection first. A difference alone does not show that either channel is worthless.

Can we add Meta and Google revenue?

Not as a business sales total. Their credited orders can overlap.

Will another pixel recover every missing sale?

No. Collection depends on the integration, identity, consent and customer journey. Validate the actual setup and label what remains unobserved.

Download the reconciliation worksheet below and use it in your next weekly reporting review.