Meta Ads Attribution · Windows, Models & Cross-Channel

Attribution answers the question: which ad should receive credit for this conversion? When a user clicks a Meta ad on Monday and purchases on Friday, Meta attributes that purchase to the Monday ad click — the ad interaction receives credit for the conversion that happened later.

Meta uses a last-touch attribution model for most reporting: if a user clicked a Meta ad and also clicked a Google ad before purchasing, Meta attributes the conversion to the Meta ad click and Google attributes it to the Google ad click. Both platforms claim the same conversion — which is why the sum of reported conversions across all ad platforms always exceeds actual conversion volume.

An attribution window defines how long after an ad interaction a conversion can be credited to that interaction. Meta offers two click-through window options and one view-through window:

The default Meta attribution window is 7-day click + 1-day view. This means a conversion counts if the user clicked a Meta ad within the last 7 days OR viewed (but did not click) a Meta ad within the last 1 day — whichever occurred most recently.

Impact of window choice on reported results

A wider attribution window always produces more reported conversions — not because more conversions occurred, but because more historical ad interactions qualify as the attributing touchpoint. Compare performance across windows to understand your typical conversion delay: if most conversions happen within 1 day of the ad click, 7-day and 1-day windows report similar numbers; if many conversions happen on day 5–7, the 7-day window captures significantly more than the 1-day window.

View-through attribution credits conversions to ad impressions — users who saw the ad but did not click it. A user who sees your Meta ad on Monday without clicking, then visits your website directly on Tuesday and purchases, would have that purchase attributed to the Monday ad view (under 1-day view attribution).

View-through attribution is the most contested form of attribution because it attributes conversion credit to the ad impression without demonstrated user response (the click). It conflates correlation (user was exposed to ad and later converted) with causation (the ad caused the conversion). Users who convert after a view impression may have converted anyway without the ad.

When to include/exclude view-through

• Include 1-day view for brand awareness campaigns where the goal is brand recall — view-through provides some signal of ad exposure contributing to downstream behaviour
• Exclude view-through (set attribution to 7-day click only) for performance campaigns where you want conservative conversion reporting — remove the view attribution to compare more cleanly against GA4 and other click-based attribution systems

Meta and GA4 will virtually never report identical conversion numbers for the same period. Several structural reasons:

• Attribution model difference. Meta uses last-touch within its window; GA4 uses data-driven attribution by default (distributes credit across multiple touchpoints). The same conversion is attributed differently.
• Cross-device. A user who sees a Meta ad on mobile but converts on desktop — Meta's cross-device tracking can attribute this if the user is logged in to both; GA4 may show the conversion as direct or from a different source depending on session tracking.
• iOS ATT signal loss. Meta models some conversions it cannot directly track — these appear in Meta's numbers but not GA4's (because GA4 only tracks actual sessions, not modelled ones).
• View-through conversions. Meta counts view-through conversions by default; GA4 does not attribute conversions to ad impressions (only clicks) — pure impression-attributable conversions exist in Meta but not GA4.
• Ad blocker users. GA4 JavaScript tracking is blocked by ad blockers; Meta's modelled data partially recovers these users.

The practical approach: use Meta-reported conversions as a directional indicator for optimisation decisions; use GA4 as your source of truth for business performance measurement. Track the ratio between Meta-reported and GA4-reported conversions over time — a stable ratio (e.g. Meta always reports 1.4× GA4) allows you to calibrate Meta numbers against your first-party data.

Attribution windows are set at the Ad Set level in Meta Ads Manager (under Conversion → Attribution Setting when creating or editing an Ad Set). Available combinations:

• 7-day click + 1-day view (default)
• 7-day click only
• 1-day click + 1-day view
• 1-day click only

Smart Bidding uses the selected attribution window as its optimisation signal — if you select 1-day click, Meta's algorithm optimises for users likely to convert within 1 day; 7-day click allows optimisation for users with a longer consideration cycle. Match the window to your typical conversion delay for the most accurate bidding signal.

Since iOS 14.5, Meta uses statistical modelling to estimate conversions that cannot be directly attributed due to ATT opt-outs. Meta identifies patterns from users who consented to tracking and uses those patterns to estimate the likely conversion rate among opted-out users with similar characteristics.

Modelled conversions appear in your Meta Ads Manager reports alongside directly-attributed conversions. The total reported conversions = directly attributed + modelled. Meta indicates when modelled data is included through tooltips in the interface.

Modelled attribution makes Meta-reported numbers more complete but also less auditable. You cannot verify modelled conversions against your order management system — they are estimates. Cross-referencing Meta totals against actual business outcomes (Shopify orders, CRM leads) remains essential for calibrating confidence in Meta's reported performance.

When running both Meta and Google Ads simultaneously, both platforms will claim credit for the same conversions. This is inherent to last-touch attribution across multiple channels — it does not indicate fraud or error, it is a structural feature of siloed platform reporting.

To understand cross-channel contribution more accurately:

• Use GA4 as a neutral arbiter. GA4 sees all channel touchpoints in a session sequence — its path analysis shows how Meta and Google interactions relate to each other in the customer journey.
• Run incrementality tests. Meta's Holdout Test measures the true incremental lift from Meta advertising — how many conversions would NOT have happened without the Meta exposure. This is more accurate than attribution-based measurement for understanding Meta's true contribution.
• Track new vs returning customer acquisition. Meta is often strongest at acquiring new customers; Google at converting existing searchers. Tracking this dimension helps each channel earn credit for what it genuinely does well.

Authentic Sources