Why do we see discrepancies between Google Analytics and Atrilyx?

Some level of variance between Atrilyx and Google Analytics is expected, and in most cases it's explained by methodology rather than a tracking error. Topline numbers — total sessions, total conversions before attribution is applied — should sit close to each other. Differences appear once attribution methodology, conversion windows, and source assignment are factored in.

Where the differences come from

Attribution model

Atrilyx uses the Ads Preferred attribution model — a data-driven, last-paid-touch model that gives 100% of the conversion credit to the network of the last paid ad the user clicked through before converting. If there are no paid ad clicks anywhere in the path, the conversion is attributed to Organic. Google Analytics 4 (GA4) defaults to a data-driven model that distributes fractional credit across multiple touchpoints, including non-paid sources. A conversion that Atrilyx credits 100% to one paid network will often show up in GA4 split across paid and organic.

Lookback window

Atrilyx uses a 90-day lookback window for paid touch attribution. GA4 defaults to a shorter acquisition lookback for paid conversions (30 days by default, configurable). When a user clicks an ad, waits more than 30 days, and converts, GA4 may credit the conversion to a later non-paid touch while Atrilyx still resolves it back to the original paid click.

Source identification

Atrilyx ingests structured spend, click, and impression data directly from each ad network through validated API connectors, and matches paid clicks to user sessions through Atrilyx-specific URL parameters carried on every paid placement. GA4 relies primarily on UTM parameters and auto-tagging, which can be inconsistent across networks, missing on programmatic and connected-TV placements, and lost when redirects strip them. Network coverage and source fidelity differ as a result.

View-through conversions

Atrilyx separately reports view-through conversions (VTCs) — credit for conversions where a user was served an impression but did not click. GA4 does not measure these. Including or excluding VTCs alone will produce different totals between the two systems.

What this means in practice

Atrilyx and Google Analytics answer different questions. GA4 is built to describe how users behave across all sources of traffic, weighted across touchpoints. Atrilyx is built to measure the performance of paid media investment specifically — which campaigns, ad sets, keywords, and creatives are driving conversions and revenue. Both views are useful; they're not meant to match exactly. When evaluating paid media performance, Atrilyx is the authoritative system because it's built around paid attribution by design.

If a specific discrepancy is unexpected or significant, your Atrilyx Account Representative can walk through the conversion paths for the period in question and identify the exact drivers.