by Om RathodAttribution windows sound like a technical setting you configure once and forget. They're not. The window you pick decides which sales get credited to which channel, and that decision quietly reshapes your entire marketing budget. Get ecommerce attribution window best practices wrong and you'll scale the wrong campaigns for months before anyone notices the ROAS was fake.
This post covers how the major platforms handle windows by default, how to match a window to your actual purchase cycle, and how to test your assumptions instead of trusting whatever Meta or Google shipped as a default three years ago.
An attribution window is the amount of time between someone clicking (or viewing) an ad and a conversion that platform is still willing to credit to that ad. Seven days, 28 days, 90 days, whatever the setting says.
Here's the part people miss: changing the window doesn't change how many sales happened. It changes which channel gets to claim them.
Take a single Meta campaign. Run it with a 7-day click window and you might see a 2.5x ROAS. Pull the same spend and the same sales through a 28-day click window, and that number can jump to 5x or higher, because now you're capturing purchases that happened three weeks after the click, some of which probably would have happened anyway.
Same ad spend. Same sales. Two completely different stories depending on the setting.
This is also why your Meta dashboard, your Google Ads dashboard, and your GA4 reports rarely agree on "the same" period. They're not lying to each other. They're just counting on different clocks.
Every platform ships with its own default, and almost nobody changes it.
Meta defaults to 1-day view, 7-day click as the modern standard, a big shift from the old 28-day click default that was common before iOS 14. Apple's tracking changes made longer windows harder to substantiate with real data, so Meta pulled back. If you're still comparing current performance to a 2020 benchmark, you're comparing two different measurement systems, not two different years of results. See how this shows up in Meta reporting specifically, since it's the platform where this shift hit hardest.
Google Ads defaults to a 30-day click-through window for most standard conversion actions, though some advertisers configure it up to 90 days depending on the conversion action settings. That's already 4x longer than Meta's current default, which alone explains a chunk of the "Google always looks better than Meta" narrative in a lot of dashboards. Worth checking your actual Google Ads conversion action settings rather than assuming.
GA4 uses a 90-day default lookback for conversion paths, which is a totally different measurement philosophy than platform-reported, last-click-biased attribution. GA4 is trying to reconstruct a path; the ad platforms are trying to claim credit. Both can be "correct" and still disagree by a wide margin. If you haven't looked at how GA4 attributes differently from Meta and Google's own dashboards, that gap is usually bigger than people expect.
TikTok and other newer channels often default to shorter windows still. Compare TikTok's short-window numbers directly against Meta's or Google's longer windows and TikTok will look like it's underperforming, when really you're just measuring it on a stricter clock.
There's no universal correct window. There's only the window that matches how long your customers actually take to decide.
Impulse and low-AOV products (under $50, single SKU, low commitment) fit a 1-7 day click window. The decision happens fast or it doesn't happen. Stretching the window here just lets unrelated purchases sneak in and inflate ROAS.
Considered purchases ($100-300 AOV, apparel bundles, multi-SKU carts) usually need 14-30 days. People browse, close the tab, come back after payday or after checking a review site.
High-consideration categories (furniture, electronics, subscriptions, anything over a few hundred dollars) can need 30-90 days, and even then you might be missing research that happened across devices or in-store. This is where a post-purchase survey ("how did you hear about us?") fills gaps that click-based windows structurally can't see.
Don't guess at this. Pull actual time-to-purchase data from GA4's path reports or your own order history: median days between first touch and purchase, broken out by category if you sell more than one type of product. That number should set your window, not a platform default someone picked in 2019.
Mismatched windows across platforms. Running Meta on a 7-day click window and Google on 30-day, then putting both ROAS numbers on the same slide, isn't a comparison. It's two different rulers pretending to measure the same thing.
Set-and-forget defaults. CAC creeps up, purchase behavior shifts, your product mix changes, and the attribution window nobody's touched since launch quietly stops reflecting reality.
Ignoring view-through conversions entirely. Turn those off and you'll systematically undercount upper-funnel, awareness-driving channels like TikTok and Meta prospecting. They don't get the click, but they start the journey. Zero view-through credit makes prospecting campaigns look like they're failing when they're actually doing exactly their job.
Over-crediting branded search with a long window. Give branded search 30 or 90 days while everything else runs shorter, and it'll hoover up credit for demand that other channels actually generated. This is one of the sneakier ways blended CAC ends up looking better than it really is; the "efficient" channel is just skimming credit from everything upstream of it.
Don't pick a window and move on. Test it.
Run a holdout or geo-lift test: turn ads off in a subset of markets or for a set period, and watch what actually happens to sales. Compare that real-world drop (or lack of one) against what your attribution window claims those ads were driving. If the platform says a campaign drove $50k and sales barely move when you pause it, the window is overcrediting.
Cross-check platform numbers against GA4 and actual order data from Shopify for the same period. Three numbers, same week, and if they're wildly apart you've got a window mismatch, not a data problem.
Move the window incrementally. Go from 7 to 14 days, watch how ROAS and CAC shift, then try 14 to 30. You're looking for the point where the numbers stop moving much, that's usually close to your real purchase cycle. Big jumps at every step mean you haven't found stability yet.
Once you land on a number, write it down along with the reasoning: why this window, based on what data. Otherwise six months from now someone changes it back to a platform default and nobody notices until CAC numbers stop making sense.
The real problem isn't picking one good window. It's that Meta, Google, TikTok, and GA4 all apply their own logic by default, and stitching those native numbers into one dashboard means you're mixing measurement systems without realizing it.
A unified view needs one consistent attribution model applied across every channel, not four platform-native ones bolted together after the fact. This is the actual argument for Redshift-based reporting: instead of importing each platform's self-reported, self-favoring numbers, you pull raw data and apply one attribution logic across all of it. Trivas builds its dashboards this way specifically because platform-native attribution is built to make each platform look good, not to tell you the truth about your marketing.
This is also usually the point where blended CAC and real incrementality conversations start making sense. Once every channel is measured on the same clock, you can actually see which ones are additive and which are just claiming credit for demand that existed anyway.
Match your window to your purchase cycle, not to whatever the platform defaulted to. Keep it consistent across Meta, Google, TikTok, and GA4 so ROAS comparisons mean something. Validate with a holdout test instead of trusting the number on the screen.
None of this is a one-time setup. Purchase behavior shifts, so revisit the window every couple of quarters, especially after a pricing change or a new product line.
If you're tired of reconciling four platforms' worth of self-reported numbers by hand, see how Trivas centralizes attribution across Amazon, Shopify, and your ad platforms into one dashboard. Worth a look before your next budget review.
explore more insights
3 min read
3 min read
3 min read
.d53b12e5.png&w=3840&q=75 "Trivas.ai")