
Attribution Window: How to Choose the Right One
The most common attribution window mistake is leaving it on the platform default. Your attribution window, which Google Ads calls a conversion window and GA4 calls a lookback window, is the number of days after a click or view during which a purchase still counts toward the ad. This should always match how long your customers take to buy, because a wider window raises reported ROAS, whether or not the ads earned it.
Conversion lag = Date of the purchase − Date of the ad click or view
Why does a wider attribution window always raise ROAS?
You spend $10,000 on a campaign and every purchase is worth $100. The ads and the spend stay the same in every row, only the window changes.
Attribution window | Purchases credited | Share of all purchases | ROAS |
1 day | 120 | 43% | 1.2 |
7 days | 210 | 75% | 2.1 |
30 days | 250 | 89% | 2.5 |
90 days | 280 | 100% | 2.8 |
ROAS more than doubles between the one-day window and the ninety-day window, but nothing about the campaign changed. A wider window will claim more purchases, so the temptation is to keep widening it.
However, some of the late purchases would have happened without the ad. If you leave the window open forever,every purchase from anyone who has ever seen your brand attributes to a campaign, which looks great in a report but says nothing about how the campaign actually performed.
How long do your customers take to buy?
Picture this: three months ago you saw a magazine ad for a major soft drink brand, and today you bought that soft drink. Three months ago you also saw an ad in an enthusiast magazine for a replacement gasket kit for a 1978 Italian lever espresso machine, and today you bought that kit.
The gap is three months in both cases, but you wouldn't credit the soft drink ad because you buy soft drinks often and see the brand everywhere. The gasket kit is a rare purchase from a small audience, and that ad may be the only place you learned the kit exists.
The right window depends on your product's buying cycle and on how much your buyers see the brand anyway. Both should be derived from your own data.
What are the default attribution windows on Meta, Google Ads, GA4, and LinkedIn?
Every platform sets its own default.
Platform | Click window default | View window default | Options |
Meta Ads | 7 days | 1 day, plus 1-day engage-through | 1 or 7 days click; 1 day view or none |
Google Ads | 30 days | 1 day | 1 to 90 days click |
GA4 key events | 90 days | Not applicable | 30, 60, or 90 days; 7 or 30 for first visits |
LinkedIn Ads | 30 days | 7 days | Configurable per conversion |
Meta counts purchases up to seven days after a click, while Google Ads counts up to thirty days, so a Google Ads ROAS that looks stronger may just have a longer window behind it. One purchase can also count on Meta, Google Ads, and LinkedIn simultaneously, because each platform credits itself inside its own window. This same overlap is covered in double counting conversions across platforms.
Before you touch a setting, see how far apart your platforms already are. Ask Summer: "Compare purchases reported by Meta Ads, Google Ads, and LinkedIn Ads with GA4 purchases for last month." The gap between the platform total and GA4 is where overlapping windows show up first.
How do you size an attribution window with data?
Pull days to conversion for one key event. In GA4, the Attribution paths report under “Advertising” shows “Days to key event”. Select a single key event, because the metric blends events together when several are selected.
Group the purchases by how many days they took at one, seven, thirty, and ninety days. Note the cumulative share at each point, as seen in the table above.
Find where the curve flattens. In the example, the first week adds 90 purchases, the next twenty-three days add 40, and the last sixty days add 30. A thirty-day window covers eighty-nine percent of purchases, and each day after it adds about one purchase every two days.
Pull what else the late buyers saw before they bought. A buyer who converted sixty days after a Meta click but opened three emails, searched your brand name, and bought twice before doesn't belong to that click. Check the late buyers against email, organic search, and repeat purchases, and compare purchase volume in weeks when the campaign was paused. Purchases that happen at the same rate without the campaign don't belong in its window.
Set the window per conversion action and match it across platforms where the options allow. Meta's ad set click window is one or seven days, with twenty-eight days available for comparison only, so when Google Ads runs thirty, note the difference next to every cross-platform ROAS in your report. A new window applies from the day you change it, so past reports keep the old one.
Re-check the curve when your price, product, or audience changes, because each one changes how long people take to buy.
How do you compare attribution windows across platforms?
Although Meta Ads Manager, Google Ads, and GA4 each report attribution clearly for their own campaigns, the comparison usually happens in a spreadsheet rebuilt every reporting cycle. AI connected to each platform through MCP reads all of them live, but connecting every API still leaves five definitions of a conversion and three attribution windows, so the model must decide on the fly what Meta's number means next to Google's.
Summer connects GA4, Google Ads, Meta Ads, LinkedIn Ads, Klaviyo, and Shopify into one normalized layer, so you can ask live marketing data anything to get verified answers, client-ready reports, and next steps. Ask any campaign question in plain English, get a verified answer back, and every number shows its receipt, a per-answer audit trail down to the rows behind it.
When a brand asks why Meta's ROAS is lower than Google's, you can show which source and which rows produced each figure. Connect once, ask anything.
Frequently asked questions
What is an attribution window?
An attribution window is the number of days after an ad click or view during which a purchase is credited to that ad. A purchase after the window closes doesn't count toward the ad.
Is a lookback window the same as an attribution window?
Yes. Google Ads calls it a conversion window and GA4 calls it a lookback window, and all three describe the same setting. The platform name differs, and the defaults differ more.
What's the difference between click-through and view-through attribution?
Click-through windows credit purchases after someone clicks the ad, while view-through windows credit purchases after someone only sees it. View windows are shorter by default on Meta, Google Ads, and LinkedIn because a view is a weaker signal than a click.
Can I still use a 28-day attribution window on Meta?
Not in the ad set, where the click options are one or seven days. The twenty-eight-day click window is available only in Ads Manager's compare attribution settings, for reporting.
What is the Google Ads conversion window?
Google Ads applies a thirty-day click-through window to new Search and Display conversion actions, and you can set it anywhere from one to ninety days. View-through conversions default to one day.
What attribution window should I use?
Use the shortest window that covers most of your days to conversion, after removing purchases that happen at the same rate without the campaign. Fast, frequent purchases need a short window, and rare, considered purchases need a longer one.
Why don't my Meta and Google Ads conversions add up to my GA4 total?
Each platform credits purchases inside its own window, so one purchase can count on several platforms while GA4 counts it once. The windows also differ, with Meta at seven days after a click and Google Ads at thirty by default.

Hot Mike
CTO of Summer / Host of Hot Mike Cool Data
