Know which channels really drive your sales.

These are real campaign reports from stores we run ads for. Each one is a platform grading its own work: Google counts conversions under its model, Meta under its attribution setting, and a sale touched by both can appear in both. That is the problem ecommerce attribution solves, and why we never add platform numbers together.

We hold this site to the same ecommerce attribution standard. Its tracking respects consent mode, and when we changed the tracking script we tested the old and new versions side by side before switching, rather than trusting a single report.

Ecommerce attribution starts with the one number nobody argues about: orders and revenue in Shopify or WooCommerce. We pull last month’s orders and line them up with what GA4, Google Ads and Meta each report. If the platforms together claim far more revenue than the store made, that gap is the starting point, and its causes are usually in the list above.

Tracking comes before models. We place test orders and check that each purchase arrives once in every tool, with the right value, currency and order ID, and that consent mode works as it should. A model built on missing or doubled purchases is worse than no model. Our tracking and analytics service fixes these first if the audit finds them.

Then the settings. As of October 2026, GA4 offers three models in its attribution reports: data-driven, paid and organic last click, and Google paid channels last click, according to Google’s page on attribution settings, which also says first click, linear, time decay and position-based were removed in November 2023. The key event lookback window defaults to 90 days, with 30 or 60 as options. Google Ads uses data-driven or last click, per its page on attribution models. Meta credits purchases that happen within a set number of days after someone views or clicks an ad, under the attribution setting on each ad set, as its attribution system page explains. We document every setting so you know what each number means.

We don’t try to make the platforms agree, because they can’t. Each one sees only its own ads. Instead we give you one view for budget decisions: the marketing efficiency ratio (total revenue divided by total ad spend), new-customer revenue from the store, and each channel’s reported numbers beside it. When MER holds while one channel claims a jump, that claim gets a closer look.

For the biggest budget questions, we test. A holdout test, such as pausing a channel in some regions or for a set period while others carry on, shows incrementality: the sales that would not have happened without the ads. Meta and Google offer lift studies to eligible advertisers, and a simple geographic split works for smaller ones. One clean test answers more than a year of debating dashboards.

Last, the report. Each month you get orders and revenue from the store, MER, new-customer share, each platform’s claimed revenue with its model and window noted, and any test results. The comparison with Google Ads’ own model comparison view shows which campaigns last click undervalues, so budget moves are made with eyes open.