Skip to content
DATANOSTRO ACADEMY

What is an attribution window

An attribution window is the time within which a conversion still gets credited to an ad click or view. Why it matters and why it makes numbers differ between platforms.

5 min Read Intermediate Updated 7.6.2026

When a customer clicks an ad today and buys a week later, should that conversion be credited to the ad? The answer is set by the attribution window — the time within which a platform still attributes a conversion to a prior interaction with the ad.

How it works

An attribution window is a time limit. If the conversion happens within the window after the click (or view), the platform credits it to the ad; outside the window it doesn't. There's a click-through window (after a click) and a view-through window (after a view without a click).

Why it differs between platforms

Each platform has its own default windows and settings. Google Ads typically uses a longer click window (configurable), Meta shorter windows (e.g. 7 days after click / 1 day after view). So the same purchase may be counted by one platform and not another — and the numbers "don't match". Context is in why GA4 and Google Ads don't match.

Why it matters

  • A longer window = more credited conversions (including ones where the ad was just "around").
  • A shorter window = stricter attribution, but undervalues longer buying journeys.
  • When comparing channels, compare the same windows, or you're comparing apples to oranges.

Relation to attribution and identification

The attribution window works with the attribution model (how credit is split) — see first-touch vs last-touch. And for a conversion to fall within the window at all, identification must survive long enough — which server-side handles for longer journeys.

Summary

An attribution window is the time within which a conversion still gets credited to an ad. It differs between platforms, which is why numbers don't match — that's a feature, not a bug. When comparing, always compare the same windows.

Did this article help you?
✓ Thank you for the feedback