What is the difference between All conversions and First conversion on Meta?

All conversions counts every purchase that happened after an ad click or view inside the window. First conversion counts only the first one. A big gap between them means people are buying more than once after the same click.

Can I compare a Conversion Lift result with my 7-day click numbers?

Meta advises against it. A lift test counts all conversions in its test and holdout groups between the test's start and end dates. Ads Manager counts purchases inside attribution windows. Use the lift result to judge the campaign, not to correct the window.

Does 7-day click include purchases from people who already bought from me?

Yes. Meta's window does not check whether the buyer is new to your store. Shopify's New vs returning customers report shows how many of last week's buyers had ordered before, so you can set that number next to Meta's.

How to test 7-day click attribution, step by step

Count repeat purchases with First and All conversions, and see how much of the count needs days 2 to 7. Rank campaigns under Meta's incremental model and count returning buyers in Shopify. Then book a lift test for the campaign whose profit depends on the window.

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

Run the numbers for your store: the free repeat purchase rate calculator.

To test 7-day click attribution, run four checks. Count repeat purchases on the same click, and see how much of the count needs the full week. Rank the same campaigns under Meta's incremental model. Then count how many buyers had ordered before. If it holds up, use it to compare ads; if not, read it as timing.

Pick one closed week, so late purchases have had time to arrive. Then open Ads Manager, GA4 and your Shopify admin side by side. Steps 1 to 4 happen in Meta and step 5 in Shopify. Steps 6 and 7 happen in GA4, and step 8 books the test that settles cause.

Step by step

Find the ad sets that count on 7-day click. Meta sets the attribution setting per ad set, so one campaign can mix a 7-day rule with a 1-day rule. Add the Attribution setting column, then click into the campaign to see every ad set. Menu path: Ads Manager > Columns dropdown next to View Setup > Customize columns > search Attribution setting > Apply.
Count first purchases and repeats separately. Open Compare attribution settings, pick 7-day click and set the conversion count to First conversion. Note the number, then run it again with All conversions. First conversion keeps only the first purchase after a click, so the gap is people buying again on the same click. Menu path: Ads Manager > Columns: Performance > Performance > Compare attribution settings > 7-day click > First conversion or All conversions > Apply.
See how much of the count needs the full week. Tick 1-day click in the same panel, still on First conversion. Whatever 7-day click shows above 1-day click came 2 to 7 days after the link click. That slice exists only because the window is a week long. Menu path: the same Compare attribution settings panel, with 1-day click and 7-day click both ticked.
Rank your campaigns under incremental attribution. Meta's incremental column keeps only the conversions its model counts as caused by the ad. It shows nothing for dates before 1 April 2025. Meta recommends comparing within one model, so rank campaigns in each column instead of subtracting one from the other. Menu path: Ads Manager > Columns: Performance > Compare attribution models > Incremental > Apply.
Split last week's buyers into new and returning. Shopify calls a customer returning when their order history already holds an order. Meta's window never asks that question. If 7-day click claims more purchases than you had first-time customers, part of its credit went to people who had bought before. Menu path: Shopify admin > Analytics > Reports > Category filter > Customers > New vs returning customers, grouped by week.
Check whether GA4 can follow a buyer between devices. GA4 joins devices only through a User-ID you send for signed-in shoppers, and otherwise works from each device's ID. Meta measures the person who clicked. A phone click and a laptop purchase can therefore be a Meta sale and a GA4 Direct visit at once. Menu path: GA4 > Admin > Data display > Reporting identity.
Read how long your Meta buyers take in GA4. Open Key event attribution paths, select purchase and switch the table to Campaign. Download it and read Days to key event on the paths that include your Meta campaigns. Where those paths run past a week, a Meta click that came early and alone fell outside the window. Menu path: GA4 > Advertising > Key events > Key event attribution paths > Share this report. Google's attribution guide reaches the same report as Advertising > Attribution > Attribution paths.
Book the test that settles cause. A Conversion Lift test splits people into a group that can see your ads and a control group that cannot. Meta says lift results are not meant to be compared with campaign results in Ads Manager, so judge the campaign by the test. If no campaign qualifies, plan a regional holdout instead. Menu path: create the test in Experiments, then read it under Experiments > Learn > your test > View report.

A worked example

Here are two made-up campaigns over one closed week, in round numbers. For illustration, each spent €2,000 and each purchase is worth €75.

For illustration: one closed week	Retargeting	Prospecting
7-day click, All conversions	150	80
7-day click, First conversion	120	76
1-day click, First conversion	108	46
Rank under incremental attribution	2nd	1st

For illustration, retargeting looks like the star, with 150 purchases on 7-day click. For illustration, First conversion trims that to 120, so 30 were repeat purchases by people who had already bought after the same click. For illustration, 1-day click still keeps 108 of those 120, so retargeting's buyers mostly pay within a day.

For illustration, prospecting keeps 76 first purchases on 7-day click but only 46 on 1-day click. Its credit leans on days 2 to 7. Meta's incremental column, in this made-up week, ranks prospecting above retargeting.

Now put money on it. One store's anonymised GA4 export, 1 January 2024 to 21 August 2026, carries a Break-even sheet. At a 40% margin, the Break-even sheet gives break-even ROAS of 2.5x, because 1 divided by 0.40 is 2.5. That is arithmetic for any store at that margin, not that store's ROAS.

For illustration, prospecting's 76 first purchases bring €5,700 on €2,000 of spend, a ROAS of 2.85x, above the line. For illustration, its 46 one-day purchases bring €3,450, about 1.7x, below the line. For illustration, retargeting clears 2.5x on every window: 4.5x on its 120 first purchases and about 4x on its 108 one-day ones.

So prospecting pays only if you believe the full week. Retargeting pays only if its buyers needed the ad, and the incremental ranking doubts that. No window can settle either question, which is why step 8 exists.

What to check when the report looks wrong

First conversion beats your first-time customers. They count different things. Meta's First conversion is the first purchase after an ad click or view; a Shopify first-time customer has no earlier order at all.
The incremental column is empty. Your date range likely begins before 1 April 2025, where the column starts. Move the start date later and look again.
7-day click shifted in spring 2026 with nothing changed on your side. In March 2026, Meta moved shares, saves and likes out of click-through, so only link clicks count. Accounts received the change at different points while it rolled out.
A dashboard dates Meta purchases differently from Ads Manager. Meta's Insights API can report a purchase on the impression day or the conversion day, set by action_report_time. Ask which one your dashboard uses before you match days with Shopify.
The lift test and Ads Manager disagree. That is expected. A lift test counts all conversions in its test and holdout groups between its start and end dates. Ads Manager counts purchases inside attribution windows.
GA4 shows far fewer Meta sales than 7-day click. Device switches from step 6 explain part of it, and missing UTM tags can explain the rest. Check the tags with how to track Meta ads in GA4.

What to do this week

Run step 2 on your biggest campaign. Compare First conversion with All conversions for last closed week. Pass: the two counts sit close together. Fail: All runs well ahead, so repeat orders are padding that campaign's results.
Turn steps 2 and 3 into money. Multiply each count by your average order value and divide by spend. Hold both results against your break-even ROAS, 1 divided by your margin. Pass: both clear it. Fail: only the 7-day figure clears it, so that campaign's profit depends on the window.
Book a test for the campaign whose profit depends on the window. Create a Conversion Lift test in Experiments, or a regional holdout if none qualifies. Pass: a test with dates and a control group is booked before the month is out. Fail: nothing is booked, so the window keeps the final word on your 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

What is the difference between All conversions and First conversion on Meta?
All conversions counts every purchase that happened after an ad click or view inside the window. First conversion counts only the first one. A big gap between them means people are buying more than once after the same click.
Can I compare a Conversion Lift result with my 7-day click numbers?
Meta advises against it. A lift test counts all conversions in its test and holdout groups between the test's start and end dates. Ads Manager counts purchases inside attribution windows. Use the lift result to judge the campaign, not to correct the window.
Does 7-day click include purchases from people who already bought from me?
Yes. Meta's window does not check whether the buyer is new to your store. Shopify's New vs returning customers report shows how many of last week's buyers had ordered before, so you can set that number next to Meta's.

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