Your ROAS dropped from 4.0 to 2.0 overnight. Same campaign, same budget, same creative. You didn't touch a thing — except one dropdown: the attribution window. Here's what matters. Your actual revenue didn't move a baht. You changed the lens you're looking through, not the result you're measuring.
What an attribution window actually is
An attribution window is the time period Meta uses to credit a sale back to your ad after someone clicks or views it.
Set it to 7-day click and the rule is simple: someone clicks your ad Monday, buys Friday, and Meta counts that sale for your campaign. Set it to 1-day click and that same Friday buyer vanishes from the report — they bought more than 24 hours after the click.
Since iOS 14.5, the default is 7-day click plus 1-day view. View-through (people who saw the ad but never clicked) now only counts for one day.
The key point: the sales happened either way. The window only decides which ones show up in Ads Manager.
Why your ROAS seems to disappear
Most products aren't impulse buys. Someone sees your ad, checks reviews, asks a friend, waits for payday, then buys. That cycle takes 2-5 days for a lot of considered purchases. It's normal.
7-day click captures those buyers. 1-day click doesn't.
Put numbers on it. A campaign spends $300. The 7-day click report shows $1,200 in revenue, a ROAS of 4.0. Switch to 1-day click and the report shows $600, a ROAS of 2.0. Your backend still shows $1,200. The "missing" $600 is everyone who bought on days 2 through 7.
The window isn't just a report — it shapes optimization
This is where founders get burned. The attribution window you pick at campaign setup doesn't only format a number. It tells the algorithm what kind of conversion to chase.
Set 1-day click and you're telling Meta to find people who buy fast. The algorithm optimizes toward quick deciders. Set 7-day click and you're saying find people who buy within a week, which hands Meta a wider pool of signal to learn from.
For anything over a $15 price point, squeezing the window too tight starves the algorithm. Fewer conversions inside the window means a slower, shakier learning phase.
Which window should you use
| Your situation | Window to use |
|---|---|
| Low-price impulse product, under $10 | 1-day click |
| Standard product, $10-60 | 7-day click |
| High-price, long decision cycle | 7-day click, watch 1-day view too |
| Comparing ROAS across campaigns | Same window on every campaign |
Default to 7-day click for reporting. Use 1-day click as a secondary view when you want to see how much revenue comes from fast buyers specifically.
The trap nobody talks about
The real trap isn't picking the wrong window. It's comparing numbers across two different windows without realizing it.
Here's the common one. Last month you read the 7-day click report and saw ROAS 4.0. This month someone flipped the setting to 1-day click. You see ROAS 2.0, conclude the campaign got worse, and cut the budget — when performance never changed.
Another trap: comparing Ads Manager ROAS to your real backend revenue. Those two numbers will never match. Meta only counts conversions it can track inside the window. Your backend counts every order, including organic, walk-ins, and word of mouth. A 20-40% gap between them is normal.
Before you trust any number, check which window produced it. Top-right of Ads Manager there's a "Comparing attribution settings" option that shows windows side by side.
What to do next
Before you make a call on any campaign, open Ads Manager and check the current window. Then lock the same window across every campaign you plan to compare.
When you build a blueprint in AdBlueprint, look at the Optimization & Tracking section. The window recommendation adjusts to the product price and buying cycle you entered — so if your customers take a week to decide, it won't push 1-day click on you in the first place.
