What Is a Marketing Attribution Window? A Plain-English Guide
by Om Rathod
|
6 min read
Aug 24, 2026
Ask your ad platform why ROAS dropped 40% overnight with zero change in spend or sales, and you'll usually get a shrug. The real answer is almost always attribution windows. So what is a marketing attribution window? It's the setting quietly deciding which sales your ads get credit for, and which ones they don't.
What Is a Marketing Attribution Window, Exactly?
An attribution window is the fixed stretch of time after someone clicks (or sees) your ad during which a conversion still counts as "caused by" that ad.
Say you're running a 7-day click window. Someone clicks your ad on Monday and buys the following Monday: that's day 7, so the sale gets credited. They buy on Tuesday instead, one day later, day 8: the platform records zero credit for that ad, even though the click still happened.
Most platforms split this into two pieces. There's the click-through window (how long after a click a sale still counts) and the view-through window (how long after someone merely sees an ad, without clicking, a sale still counts).
None of this is universal. Meta's default is a 7-day click, 1-day view window. Google Ads often runs 30 to 90 day windows depending on the conversion action. Same customer, same purchase, wildly different attribution outcome depending on which platform's clock is running.
Why Attribution Windows Exist in the First Place
Real purchase journeys are messy. Someone clicks a Meta ad on their lunch break, forgets about it, gets a retargeting email nine days later, and finally buys two weeks after that first click. Which touchpoint gets the credit?
Without a cutoff, every ad you've ever run could theoretically claim partial credit for every sale, forever. That's not useful to anyone. So platforms draw a line and say: if the conversion happens outside this window, we stop counting it.
It's a real tradeoff, though. Set the window too short, and you undercount purchases that take real consideration time, think a $200 skincare bundle someone researches for two weeks. Set it too long, and you inflate ROAS by crediting ads for sales that would've happened anyway, or that another channel actually closed.
Click-Through vs View-Through Windows
Click-through window
What it measures: Conversions from someone who clicked the ad and then purchased within the set period
Typical length: 1, 7, or 28 days depending on platform and campaign type
View-through window
What it measures: Conversions from someone who simply saw the ad (no click) and purchased shortly after
Typical length: Usually 1 day, sometimes disabled entirely
View-through attribution is the more contested one. It's crediting an ad for a purchase someone made without ever clicking it, and often without consciously remembering they saw it. Critics argue it's closer to correlation than causation. Platforms argue impressions still shift behavior even without a click. Both are a little right.
Practical tip: before you compare ROAS across Meta and Google Ads, check whether each platform's reported number is blending click-through and view-through conversions. If one includes view-through and the other doesn't, you're not comparing the same thing, no matter how similar the dashboards look.
How Attribution Window Length Changes Your Reported ROAS
Here's the part that catches people off guard. Same spend, same actual sales, three different reported ROAS numbers, just by changing the window.
Imagine $10,000 in ad spend for the week. Under a 1-day click window, the platform only counts sales from people who clicked and bought within 24 hours: say that's $15,000 in attributed revenue, a 1.5x ROAS. Widen it to 7 days, and now sales from people who clicked and came back three or four days later also count: attributed revenue climbs to $28,000, a 2.8x ROAS. Stretch it to 30 days and you might see $35,000, a 3.5x ROAS. Nothing about the business changed. Only the lens did.
This is exactly why brands that got pushed from Meta's old 28-day click window down to the 7-day (and briefly 1-day) default after iOS 14 saw reported ROAS drop 30 to 50% almost overnight. The sales didn't disappear. The credit did.
It's also why comparing platform-reported ROAS to your own Shopify or GA4 revenue numbers so often ends in confusion. The platform is grading itself on its own attribution window. Your store's order data isn't using any window at all, it's just recording what actually got bought. Apples and oranges, dressed up to look like the same fruit.
Choosing the Right Attribution Window for Your Business
There's no universally "correct" window. There's only the window that matches how your customers actually buy.
Impulse-buy CPG products, cheap accessories, anything under $30 with low consideration: a 1 to 7 day click window is usually honest enough. Considered purchases like furniture, mattresses, or a higher-ticket skincare set: you probably need 14 to 30 days to capture what's really happening, because people research before they buy.
Don't guess this. Pull actual time-to-purchase data from GA4 or your order timestamps and see how long it typically takes someone to go from first click to checkout. That's your real answer, not a default some platform picked for you.
One warning: resist the urge to pick a longer window purely because it makes ROAS look better. That's optimizing the report, not the business, and it'll bite you the moment someone asks why performance "changed" after you touched a setting. Whatever window you land on, keep it consistent. Switching windows mid-quarter makes every period-over-period comparison meaningless.
Attribution Windows vs Attribution Models: Don't Confuse Them
These two get mixed up constantly, and they're not the same thing.
The window is how far back you look. The model is how you split credit across every touchpoint inside that window, last-click, linear, position-based, data-driven, and so on.
Same customer journey, two different questions: the window asks "does a purchase 10 days after the click still count at all?" The model asks "if it counts, does the first Meta ad get 100% of the credit, or does it split with the retargeting email and the Google search ad that came after?"
Here's the part platforms don't advertise: their reported numbers use in-platform models that can only see their own touchpoints. Meta doesn't know about your Google ad. Google doesn't know about your email flow. Each one is grading its own homework, which is a big part of why marketing leaders and performance marketers eventually stop trusting any single platform's number in isolation and start looking for a unified view instead.
Getting Past the Attribution Window Problem Altogether
No attribution window is "correct." Each one is a lens, and every platform's lens happens to be pointed in a way that flatters itself.
The way out isn't finding the perfect window setting. It's stepping outside platform-reported numbers entirely: pulling raw order data and ad spend into one place, so you're measuring what actually happened rather than what one platform's clock decided to count.
That's the whole idea behind BI reporting built on unified order and ad data: blend Meta, Google, and Shopify numbers against your actual revenue, without any single platform's window quietly setting the terms of the debate.
If you're tired of explaining to your CFO why ROAS "changed" without any real change in the business, it might be time to stop relying on platform dashboards to grade themselves. Book a walkthrough and see what blended, window-free reporting actually looks like.
Revenue growth leader and co-founder driving Trivas's commercial strategy. Om has led the product vision and execution from scratch. With a strong background in SaaS sales and GTM strategy, Om bridges product innovation with real-world customer needs.
Continue Reading
explore more insights
Platform-Specific Real-Time Tracking Strategies
3 min read
Conclusion: Powering Up Shopify Profit Analytics For Ecommerce Success
.d53b12e5.png&w=3840&q=75 "Trivas.ai")
