Can I set GA4's lookback window longer than 90 days?

Not in the attribution settings. For purchases and other key events, Google offers 30, 60 or 90 days, and 90 is already the default. If your buyers take longer, their earliest touches fall outside every option, so judge those channels with a holdout test instead.

How far back does GA4 attribution look?

Usually 90 days. For purchases and other key events, GA4 credits touchpoints from up to 90 days before the key event, unless someone chose 30 or 60. First visits and first app opens look back 30 days, or 7 if changed. Anything older gets no credit in any GA4 report.

By Joris van Huët, Founder & CEOUpdated 2026-10-016 min read

Run the numbers for your store: the free attribution window calculator.

Ninety days is the right default and an incomplete answer. GA4 runs more than one window, anyone with the right role can change them, and a change only counts from that day on. And the window reaches back from each purchase, not from the date range you picked.

What one store's data shows

One store's anonymised GA4 export, 1 January 2024 to 21 August 2026. It holds shares of revenue only: no ad spend, no order counts.

What the export shows	Share of revenue	Source cell
Journeys with 1 touch, 0.5 days to buy	79.5%	Journeys sheet, 1 touch row
Journeys with 2 to 3 touches, 12.5 days to buy	12.2%	Journeys sheet, 2-3 touches row
Journeys with 4 to 9 touches, 16.9 days to buy	5.4%	Journeys sheet, 4-9 touches row
Journeys with 10 or more touches, 16.0 days to buy	3.0%	Journeys sheet, 10+ touches row

Read the days column first. In one store's export, the slowest row is the 4 to 9 touches row, at 16.9 days to buy (Journeys sheet). Even journeys of 10 or more touches took 16.0 days (Journeys sheet). Each of those sits inside 30 days, the shortest purchase window GA4 offers.

Most of the money barely needs a window at all. Journeys with one touch hold 79.5% of revenue on the Journeys sheet, and those buyers took 0.5 days. With a single touch half a day before the sale, every window GA4 offers gives it to the same channel.

The slow fifth is where the setting bites. Journeys of two or more touches hold 20.6% of revenue in this one store (Journeys sheet: 12.2% + 5.4% + 3.0%). On the Journeys sheet they took 12.5 to 16.9 days to buy, and they are the only sales a window can split differently.

Two things the export cannot show. Each row is one figure for many paths, so a slow tail can hide inside it. And the paths were recorded under whatever window the property used. Google describes touchpoints as interactions within a specified timeframe leading up to a key event. A touch older than the window never made it into the file, so the export cannot say how many buyers started earlier.

Why does the usual answer mislead?

GA4 runs three windows, not one. For purchases and most key events, the default is 90 days, and 30 or 60 days are the alternatives. First visits and first app opens, which Google calls acquisition key events, get 30 days or 7. Engaged-view key events keep their own default of 3 days.

Someone can change it, and history keeps the old rule. Google says changes to the lookback window apply going forward. For example, a switch from 90 to 30 days in March leaves January and February on the old window. A year-on-year chart then compares two rulebooks. The reporting model works the other way: change it and GA4 rewrites past reports too.

It counts back from each purchase, not from your dates. The date range picks the key events, and the window reaches back from each one. GA4's attribution models report defaults to Event time, which credits touchpoints in the lookback window before the key events in your range. Google notes those touchpoints may happen before the range, so a March report can credit a click from January. Ad interaction time credits only the touches inside your dates.

Even Google counts the edge differently. GA4's help says a 30-day window credits a 30 January key event to touchpoints from 1 to 30 January. Google Ads' attribution help, for its own 30-day lookback, counts from 31 December to 30 January. One day will not change your strategy. It will change a reconciliation that refuses to balance.

It also decides what Direct inherits. Google says sessions that start with a direct entrance get the user's earlier campaign values. It also says the key event lookback window applies to session attribution. So a buyer who comes back by typing your address keeps the earlier source only inside the window.

What can the window not tell you?

It cannot reach a touch that GA4 never tied to the same person. GA4 joins visits into one journey through user ID, device ID or modelling. A buyer who switches phones or clears cookies can start a fresh path, and no window reaches into someone else's journey.

It is not your data retention setting either. Retention decides how long GA4 keeps user-level and event-level data for explorations and funnel reports: 2 or 14 months on a standard property. Standard reports are not affected by it. Two settings, two jobs, and one word that sounds like both.

And it cannot tell you how long your buyers really take, or what moved them. A window is a fence, not a measurement. A touch inside it is eligible for credit, which is not the same as having sold anything.

What to do this week

Read your windows. In Admin, under Data display, click Events, then Attribution settings, and note both key event lookback windows. Pass: purchases sit on the default, or on a shorter window someone chose on purpose. Fail: a short window that nobody can explain.
Find your slowest paths. In Advertising, open Attribution paths, select only purchase and sort the table by Days to key event. Pass: the slow paths that carry real revenue end well inside your window. Fail: they crowd its edge, so early touches are probably falling off.
Mark the date of any change. If the window changed, right-click that day on a line graph in Reports and click Add annotation. Pass: anyone reading a trend sees the break. Fail: a year of charts mixes two windows and nobody knows where.

Check the homework. Your GA4 Attribution paths export already holds the evidence. Causality Engine reads that one file and shows what each channel caused next to what last-click gave it, in 1 to 2 minutes, for €99 once (excluding VAT), refundable within 30 days. Check the homework

Frequently asked questions

Is GA4's lookback window the same as data retention?
No. The lookback window decides how far before a key event a touchpoint can earn credit. Data retention decides how long GA4 keeps user-level and event-level data for explorations and funnel reports, 2 or 14 months on a standard property. Standard reports are not affected by retention at all.
Can I set GA4's lookback window longer than 90 days?
Not in the attribution settings. For purchases and other key events, Google offers 30, 60 or 90 days, and 90 is already the default. If your buyers take longer, their earliest touches fall outside every option, so judge those channels with a holdout test instead.
Does the lookback window decide which channel gets a direct visit?
Partly. Google says a session that starts with a direct entrance gets the user's earlier campaign values. It also says the key event lookback window applies to session attribution. So a buyer who returns by typing your address keeps the earlier source only inside the window.

Go deeper: Causal attribution, explained.

Sixty-second versions of these ideas: Causality Engine on YouTube Shorts.

Keep reading

Terms in this article

Browse the full glossary