Does Meta overreport conversions?

Usually yes, if you mean sales the ads caused. Meta counts every purchase inside its window after someone clicked, saw or engaged with an ad, including buyers who were coming anyway. The sales usually happened. The credit is what runs high, mostly on past customers and quick buyers. A doubled pixel can inflate the count itself.

The usual answer is "yes, always". That is half right. Overreporting needs a yardstick. Against your Shopify orders, Meta's count can land above or below. Against the sales your ads caused, it usually lands above, and the excess is not spread evenly.

What one store's data shows

You can't see an overcount from inside Ads Manager. One store's export shows where it would hide.

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.

Start with Direct. On the Channels sheet it holds 57.7% of revenue in last click, first click and touched alike. More than half of this store's revenue came through visits GA4 could not tie to a source. People who type your address or open a bookmark usually know you already. They are the easiest sales for any ad platform to claim. If one of them saw a Meta ad earlier that day, Meta can count the purchase, and GA4 has nothing to say against it.

Then speed. Journeys with 1 touch hold 79.5% of revenue on the Journeys sheet and take 0.5 days to buy. Meta's view-through setting counts a purchase made within a day of an impression. A shopper who scrolled past an ad over breakfast and ordered by lunch fits easily. Meta books a sale, and GA4 books one visit from somewhere else. Did the ad do the work? Neither count can say.

The slow end runs the other way. Journeys of 4 to 9 touches hold 5.4% of revenue on the Journeys sheet and take 16.9 days to buy. If a Meta link click opened such a journey, and days to buy count from that click, 7-day click shuts long before the sale. Meta then misses the purchase, unless a later click restarts the clock.

What the export cannot show: Meta's own count, ad spend, orders, or whether this store ran Meta ads at all. It shows where an overcount would hide, not how big one is. It is one store's record, not a benchmark for yours.

Where does Meta's extra count come from?

Overreporting covers three different problems. Each has its own check.

Buyers who were coming anyway. Meta's rule asks whether a buyer met an ad inside the window. The standard rule never asks whether they needed the ad. Past customers, cart abandoners and people who already searched your name pass that test with ease. Retargeting is the purest case. Meta lets you size this slice: its audience segment reporting splits results into new audiences, engaged audiences and existing customers. You define the last two in your ad account.

Credit without a visit. Meta's default usually counts 7-day click and 1-day view, and its Results help page adds 1-day engagement. Ad sets can differ, so read the Attribution setting column before you add anything up. A view-through purchase needs no click and no visit, so GA4 never sees the ad behind it. The default, part by part, is in What is the default Meta attribution setting?

The same order, twice. Meta drops a duplicate only when it can match the copies, by a shared event ID or a browser ID. Copies sent from the browser alone are the weak spot. Its developer guide says so plainly: "If you send us two consecutive browser events with the same information, we do not discard either". On Shopify, pixels come two ways: app pixels from apps and sales channels, and custom pixels a developer adds by hand. If both send Meta a purchase, one order can land twice.

Only the third is a counting error. The first two are rules doing what they were built to do.

Can Meta undercount at the same time?

Yes. That is the half the usual answer leaves out. Meta's own help says its pixel may not fire when a shopper uses an ad blocker. Shopify says the same of its Standard data-sharing level, which runs the pixel alone. And slow buyers, like the 16.9-day journeys in the export, can outlast the click window. One report can run high on warm, quick buyers and low on cold, slow ones.

Some of Meta's lead over GA4 is fair, too. Meta follows people across devices through their accounts, which cookie-based tools often can't.

The number that settles it comes from people who could not see the ads. Meta says so itself. Its March 2026 note tells advertisers to ask what an ad caused "that would not have happened otherwise". It calls tests like Conversion Lift the gold standard. Its Conversion Lift page says the method helps ensure Meta is "not overstating the impact of ads". Note which way the worry runs.

What to do this week

1. Split last month's purchases by audience segment. In Ads Manager, open the navigation on the left, select All tools, then Advertising settings, then Audience segments below Creating ads. Choose custom audiences for Engaged audience and Existing customers, then click Confirm. Back in Ads Manager, click Breakdown and choose Audience segments under By demographics. Pass: a prospecting campaign earns most of its purchases from the new audience. Fail: existing customers carry most of them, so treat that slice as the softest part of the claim. Meta warns the feature may not reach every account yet.
2. Run the ceiling test against Shopify. In your Shopify admin, go to Analytics, then Reports, set the Category filter to Customers and open New vs returning customers for the same month. Pass: Meta's new-audience purchases sit well below your first-time customers, which leaves room for search, email and word of mouth. Fail: Meta claims nearly all of them, or more. Then refresh the list behind your Existing customers segment and look for doubled events.
3. Check that one pixel sends Meta your purchases. In your Shopify admin, go to Settings, then Customer events, and list every app pixel and custom pixel. Pass: only the Facebook & Instagram channel sends events to Meta. Fail: a custom pixel also sends a Meta purchase, so remove the extra copy unless a developer gave both the same event ID.

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

Sources, 1 October 2026: Create audience segments in Meta Ads Manager (Meta Business Help Center); Use audience segments to enable breakdown reporting in Meta Ads Manager (Meta Business Help Center); Results (Meta Business Help Center); About attribution models and attribution settings (Meta Business Help Center); Handling Duplicate Pixel and Conversions API Events (Meta for Developers); Pixels overview (Shopify Help Center); About conversion count differences between Meta Ads Reporting and third-party reporting tools (Meta Business Help Center); Facebook data sharing (Shopify Help Center); Simplifying Ad Measurement for a Social-First World (Meta, 3 March 2026); About Conversion Lift (Meta Business Help Center); Customers reports (Shopify Help Center); App pixels (Shopify Help Center).

Related answers

• All Meta ads questions
• How to check if Meta overreports conversions, step by step
• Should I trust Meta ads conversions?
• How do I measure Meta ads incrementality?
• View-through conversion, defined