An attribution window (also called a lookback window) is the maximum time between a marketing touchpoint and a conversion for which that touchpoint is still allowed to receive credit. A 30-day click-through window means any conversion within 30 days of a click gets attributed to that click; a conversion on day 31 gets attributed to whatever touchpoint is still inside its own window, or to no paid channel at all. This setting is configured once, buried in platform settings, and then silently shapes every reported conversion number until someone notices two platforms disagree and goes looking for why.

How window length mechanically changes reported conversions

Lengthening the window doesn't just capture more conversions in absolute terms — it reallocates credit away from channels with short consideration cycles toward channels that get clicked early in a longer journey. A 1-day window on a paid social ad will only capture same-day impulse conversions; a 30-day window on the same ad captures conversions where the user clicked, researched for three weeks, and converted later — inflating that channel's reported contribution relative to a 1-day measurement, even though nothing about the ad's actual effectiveness changed. This is why switching window length is never a neutral technical setting: it's a measurement choice that structurally favors certain channel types.

window_length_reallocation.txt

text

Same user journey, two window settings:

Day 0:  Paid Social click (consideration-stage ad)
Day 12: Organic search visit
Day 14: Branded search click
Day 14: Purchase

1-day window on Paid Social:
  -> Paid Social gets NO credit (outside window)
  -> Branded search click gets the credit

30-day window on Paid Social:
  -> Paid Social is back inside the window
  -> Credit now splits or shifts toward Paid Social
     depending on the attribution model layered on top

Nothing about the ad changed. Only the window did.

Window length and attribution model are two separate decisions

It's easy to conflate "which touchpoints are even eligible" (the window) with "how credit is split among eligible touchpoints" (the model — last-click, linear, data-driven, etc.). A long window with a last-click model still gives all the credit to one touchpoint, just a touchpoint that can now be much further back in time. Changing the window changes the eligible pool; changing the model changes how that pool is divided. Diagnosing a reporting discrepancy means checking both independently.

Why ad platforms and GA4 disagree by default

Google Ads' default attribution window is commonly 30 days for clicks and 1 day for engaged-views on video, Meta Ads has historically defaulted to 7-day click / 1-day view (configurable up to 28-day click), and GA4's default lookback for its data-driven attribution model is 30 days for acquisition events and 90 days for other conversion events — three different defaults, each tuned by its own platform to flatter that platform's own reported performance, since a platform showing more conversions inside its own longer window looks more effective in isolation.

This is the single most common cause of "the numbers don't match" conversations between a paid media team looking at platform dashboards and an analytics team looking at GA4: they are frequently, without anyone configuring it intentionally, using different window lengths for the same underlying conversions.

Typical default attribution windows by platform

Platform	Click-through default	View-through default	Configurable range
Google Ads	30 days	1 day (engaged-view)	1-90 days (click), fixed/limited (view)
Meta Ads Manager	7 days	1 day	1, 7, or 28 days (click); 1 day (view)
GA4 (data-driven)	30 days (acquisition)	N/A — view-through not modeled	30-90 days depending on event type
LinkedIn Ads	30 days	7 days	Limited; largely fixed by campaign objective
TikTok Ads	7 days	1 day	1 or 7 days (click); 1 day (view)

Reconciling platform numbers requires matching windows, not just matching math

Normalizing attribution *model* across platforms (e.g. forcing everything to last-click) still won't reconcile the numbers if the underlying *windows* differ. A true apples-to-apples comparison across Google Ads, Meta, and GA4 requires explicitly setting each platform's window to the same length — most teams standardize on a single org-wide window (commonly 7 or 30 days) and configure it everywhere, accepting that each platform's own "native" number will look slightly different once normalized.

Choosing a window length deliberately

The right window length tracks the actual consideration cycle of what's being sold, not a platform default. A low-cost impulse purchase (a $20 DTC product) justifies a short window — a conversion 25 days after a click probably wasn't caused by that click. A high-consideration B2B purchase with a multi-month sales cycle justifies a much longer window, sometimes 90 days or more, because crediting only clicks within 7 days of a close would systematically undercount every top-of-funnel touchpoint that actually mattered. The practical approach: look at your own median and 90th-percentile time-to-conversion in GA4 or your CRM before picking a number, rather than inheriting whatever a platform defaulted to.

What's next

Window length determines which touchpoints are eligible for credit at all; view-through attribution is a related but distinct question about whether an impression someone never clicked should be eligible in the first place.

Next: View-Through vs. Click-Through Attribution →

I build these systems professionally.

Whether it's a RAG pipeline, analytics migration, or AI workflow — let's talk.

Work With Me →

Need production AI or MarTech expertise? I build custom LLM workflows, server-side GTM tagging, and analytics infrastructures professionally.

Need custom AI or MarTech setup? Let's build together.

Let's talk →