Definition
An attribution window is the period after an ad interaction during which a later purchase is still credited to that interaction. Google Ads calls it the conversion window, Meta calls it the attribution setting, and GA4 calls it the lookback window. Each platform sets its own defaults and counts clicks and views differently, which is the main reason two dashboards can report different revenue for the same set of orders.
Where the window is set on the platforms a store uses
- Google Ads click-through window
- Configured per conversion action, commonly 30 days, adjustable within a range that reaches 90 days
- Google Ads view-through window
- Set separately and much shorter, applying to impressions on Display and video that were never clicked
- GA4 key event lookback
- Chosen in the attribution settings, with a longer window available for purchase events than for acquisition events
- Meta attribution setting
- Selected per campaign, with a 7-day click and 1-day view combination as the common default
- Store order data
- No window at all, since an order is recorded when it is paid and belongs to that day
- Effect of a change
- Widening a window credits more historical orders to the same spend, so ROAS rises without anything about the campaign changing
Settings and defaults are configured per account and per conversion action, so check yours rather than assuming. The point of the table is that four systems are counting the same orders against four different rules.
Why it matters
Attribution windows matter because they quietly decide the numbers a store makes decisions on. A campaign selling a considered purchase that people research for two weeks will look weak inside a short window and healthy inside a long one, with no difference in what actually happened. Smart bidding is affected too: the platform learns from conversions it can see inside the window, so a window shorter than the real purchase cycle teaches the algorithm to favour fast, low-consideration sales. Setting the window to match how customers actually buy, then leaving it alone, is what makes period-to-period comparison meaningful and keeps bid strategies pointed at the right outcome.
Where it goes wrong
- Changing the window mid-quarter and reading the jump as improvement, when the same orders have simply been reassigned to a different set of days
- Setting one window for a catalogue that spans impulse items and considered purchases, which under-credits the slow products and over-credits the fast ones
- Summing revenue across platforms whose windows overlap, so the same order is counted by Google and by Meta and the total exceeds what the store banked
- Comparing a view-through window with a click window as though they measured the same commitment, since an unclicked impression and a deliberate visit are not equivalent evidence
- Forgetting that a window is a maximum, not a claim: an order two days after a click is inside a 30 day window and does not become more credible because the window is wide
Questions about attribution window
How do you choose the right attribution window for an online store?
Look at your own time from first visit to purchase. Analytics reports the distribution of days between the first session and the order, and the window should cover the bulk of it rather than the long tail. A store selling everyday consumables usually finds nearly everything closes within a few days; furniture, jewellery and technical equipment need considerably longer.
Does changing the attribution window change past reports?
In Google Ads, changing the conversion window can reprocess conversions for recent dates, so figures you have already reported may shift. GA4 applies the lookback setting when reports are generated rather than storing a fixed answer. Either way, note the date of the change on your reporting so that a step in the chart is explained by the setting rather than mistaken for performance.
What is the difference between an attribution window and an attribution model?
The window decides which interactions are eligible for credit. The model decides how credit is divided among the eligible ones. Google Ads and GA4 now work with data-driven and last-click models, and either of them behaves differently under a short window than a long one, so the two settings have to be read together.