Does 7-day click count a sale that came back through Google?

Yes, if the purchase happened within 7 days of a link click on your Meta ad. Meta's window does not check which channel brought the buyer back. GA4's paid and organic last click model would give the same sale to the last channel clicked, such as a Google ad.

Is 7-day click accurate for products people research for weeks?

Less so. If your buyers often take longer than a week, a Meta link click that started the research drops out of the window before they pay. GA4's Key event attribution paths report shows how many days your own paths take, so check it before you trust the 7-day number.

Is 7-day click good enough to set budgets?

Not on its own. It shows which purchases followed a link click within a week, not which ones the ads caused. Use it to compare ads that share one setting, and set budgets from a Conversion Lift test or a regional holdout.

Is 7-day click attribution accurate?

Accurate as a record of timing: a purchase came within 7 days of a link click on your Meta ad. Not as a measure of cause. It credits buyers already on their way, misses slow ones and never looks at your other channels.

By Joris van Huët, Founder & CEOUpdated 2026-10-018 min read

Usually accurate as a record, not as a verdict. Meta's 7-day click counts purchases made within 7 days of a link click on your ad, and since March 2026 only link clicks qualify. It cannot say whether the ad caused the sale, and it never asks what else the buyer clicked.

The question hides two questions. Did these purchases follow a link click on your ad? Mostly yes. Meta fills some gaps with statistical modeling where data is missing, and says some accounts may still use the older click definition during the rollout.

Did the click cause the purchase? The window was never built to answer that.

What one store's data shows

The usual answer is a verdict: 7-day click is either honest or inflated. One store's GA4 export points to two errors that pull in opposite directions, and to where more of the revenue sits.

One store's anonymised GA4 export, 1 January 2024 to 21 August 2026. It holds shares of revenue only: no ad spend, no order counts.

What the export shows	Share of revenue	Source cell
Journeys with 1 touch (0.5 days to buy)	79.5%	Journeys sheet, 1 touch row
Journeys with 2 to 3 touches (12.5 days to buy)	12.2%	Journeys sheet, 2-3 touches row
Journeys with 4 to 9 touches (16.9 days to buy)	5.4%	Journeys sheet, 4-9 touches row
Journeys with 10 or more touches (16.0 days to buy)	3.0%	Journeys sheet, 10+ touches row
Direct, in last click, first click and touched views	57.7%	Channels sheet, Direct row

Error one is generosity. On the Journeys sheet, journeys with 1 touch hold 79.5% of revenue and take 0.5 days to buy. Any link click inside a journey that quick sits deep inside a 7-day window. The count is then right about timing and silent about cause, even for a shopper who was already heading to the checkout.

Error two is stinginess. The slower rows on the Journeys sheet take 12.5, 16.9 and 16.0 days to buy, all past the one-week mark. A Meta link click that opened one of those journeys would drop out of the window before the purchase. Together those rows hold 20.6% of revenue on the Journeys sheet (12.2% + 5.4% + 3.0%), about a fifth.

So in that store, far more revenue sits where 7-day click could be too generous than where it could be too stingy. No window length fixes the generous error. A 1-day click window still counts the shopper already on the way. A 28-day one adds slow buyers without asking about cause either.

Then the Direct row. On the Channels sheet, Direct holds 57.7% of revenue in last click, first click and touched views alike. GA4 files a visit as Direct when it comes from a saved link or a typed address. That is also where a shopper lands who tapped your ad on a phone and typed your address on a laptop later. Meta could count that sale as a 7-day click purchase while GA4 records one Direct visit.

What the export cannot show: whether that store ran Meta ads at all, what Meta counted, or any spend or orders. It is one store, not a benchmark, so read your own Journeys numbers the same way.

Is 7-day click more accurate than GA4?

On one point, it can be. Meta measures people across devices. Its help page on reporting differences says other tools "may not be able to measure cross-device conversions well due to cookie-based measurement". GA4 can join a phone and a laptop only through a User-ID you send for signed-in shoppers.

So one buyer can be a Meta click-through sale and a GA4 Direct visit, and both reports follow their own rules. The device gap gets its own answer in why Meta ads go missing in GA4.

On another point, 7-day click is blind. Meta gives a purchase on your site to your ad if it happens within a set number of days of a view or click. Nothing in that rule looks at your other channels. A buyer who tapped your ad on Monday, then came back through a Google ad and an email on Thursday, is a full Meta purchase.

GA4 reads the same journey differently. Under its paid and organic last click model, a path of social, paid search and then email gives the whole sale to email. Add every platform's own count together and you can claim more orders than you took.

So neither report is the accurate one. Meta follows the person but only sees its own ads. GA4 sees every tagged channel but loses buyers who switch devices.

What can a 7-day click count not tell you?

Whether the sale would have happened anyway. Meta itself frames the real question as what an ad caused that "would not have happened otherwise". Its own answer is an experiment such as Conversion Lift, which it calls the gold standard.

In between sits Meta's incremental attribution column, which keeps only the conversions its model counts as caused by the ad. In Meta's own example, a campaign with 100 conversions shows 70 under incremental attribution. Meta recommends comparing within one model rather than incremental against standard. Use it to rank campaigns, not to shrink one number by another.

A lift test is the stronger check, and it comes with small print. Meta's lift page says a test can show whether your attribution model lines up with a controlled experiment. Its page on lift results says those results "are not meant to be compared with ad campaign results in Meta Ads Manager". Read the pair this way: let the test set the budget, and keep 7-day click for comparing ads under one setting.

What to do this week

Measure the slow buyers 7-day click misses. In Ads Manager, open Columns: Performance > Performance > Compare attribution settings, tick 7-day click and 28-day click, and click Apply. Pick a campaign at least four weeks old, or Meta shows partial 28-day data. Pass: 28-day click adds little, so few buyers pay more than a week after a link click. Fail: it adds a lot, so judge those campaigns on more than the 7-day number before you cut them.
Count last week's first-time customers. In Shopify, go to Analytics > Reports, set the Category filter to Customers and open New vs returning customers, grouped by week. Pass: Meta's 7-day click purchases for that week stay below your first-time customers. Fail: Meta claims more, so part of its credit sits with people who had already bought from you.
Rank your campaigns twice. In Ads Manager, open the Columns: Performance dropdown, choose Compare attribution settings and pick 7-day click. Then choose Compare attribution models, select Incremental and click Apply, with dates from 1 April 2025 on. Pass: the same campaigns lead both rankings. Fail: the 7-day click leader slides down the incremental ranking, so test it before it gets more budget.

Check the homework. Your GA4 Attribution paths export already holds the evidence. Causality Engine reads that one file and shows what each channel caused next to what last-click gave it, in 1 to 2 minutes, for €99 once (excluding VAT), refundable within 30 days. Check the homework

Frequently asked questions

Does 7-day click count a sale that came back through Google?
Yes, if the purchase happened within 7 days of a link click on your Meta ad. Meta's window does not check which channel brought the buyer back. GA4's paid and organic last click model would give the same sale to the last channel clicked, such as a Google ad.
Is 7-day click accurate for products people research for weeks?
Less so. If your buyers often take longer than a week, a Meta link click that started the research drops out of the window before they pay. GA4's Key event attribution paths report shows how many days your own paths take, so check it before you trust the 7-day number.
Is 7-day click good enough to set budgets?
Not on its own. It shows which purchases followed a link click within a week, not which ones the ads caused. Use it to compare ads that share one setting, and set budgets from a Conversion Lift test or a regional holdout.

Go deeper: Incrementality testing, explained.

Sixty-second versions of these ideas: Causality Engine on YouTube Shorts.

Keep reading

Terms in this article

Browse the full glossary