Meta Ads and GA4 Not Matching? Which Number to Trust

Why Meta Ads and GA4 Never Match (and Which Number to Trust)

Meta Ads and GA4 not matching is normal, not a tracking bug. Meta credits any purchase within 7 days of a link click, 1 day of an engagement or 1 day of an ad view, and files it on the ad's date. GA4 only sees site visits, splits credit across channels, and files the sale on the purchase date.

The two tools are answering different questions, so the goal isn't to make them agree. The goal is to know which number to use for which decision, and to catch the gaps that are real errors.

Key takeaways

Meta counts people who saw or engaged with an ad. GA4 counts sessions that reached your site. The same sale can be a full conversion in one and invisible in the other. Since March 2026, Meta's click-through column counts link clicks only. Likes, saves and shares moved to a separate engage-through column. Many accounts saw reported conversions fall with no real change in sales. Use Meta's numbers to optimise inside Meta, GA4 to compare channels, and your backend revenue plus blended efficiency (MER) to set budgets. A stable gap is fine. A gap that suddenly changes is usually a tagging, UTM or deduplication problem worth fixing that week. Why are Meta Ads and GA4 not matching?

They measure different things from different vantage points. Five causes explain almost every gap we see, and only the last one is an actual error.

Meta counts views and engagements; GA4 can't

GA4 only knows about someone once they land on your website. If a person watches your Reel, never clicks, and buys two days later by typing your brand into Google, GA4 credits Google or Direct. View-through conversions only show up in Ads Manager.

Meta's default setting credits that purchase to the ad if it happened within the view window. Neither tool is lying. Meta is answering "did someone who saw my ad buy?" and GA4 is answering "which visit led to the purchase?"

Meta sees people; GA4 sees browsers

Meta recognises a user through their logged-in account, while GA4 stitches journeys together with a first-party cookie. Someone who taps your ad on their phone at lunch and buys on their laptop at night is one customer to Meta. To GA4, they can look like two unrelated users.

Meta also collects data through two routes. It receives conversions from both the browser pixel and the Conversions API, whereas GA4 relies only on what its own tag records. Ad blockers and privacy settings hit GA4 harder.

GA4 shares credit; Meta keeps all of it

When a customer touches Meta, Google Ads and an email before buying, every ad platform tends to claim the sale. Meta and Google Ads can each count the same purchase as their own, while GA4 divides the credit across both channels.

This is why adding up "conversions" from every ad platform usually gives you more orders than you actually received.

Meta and GA4 file the sale on different days

Meta assigns a conversion to the date of the ad impression. GA4 records it on the date the purchase actually happened. A click on 29 September that converts on 3 October sits in September for Meta and October for GA4.

This matters most at month-end and around sale events. During festive weeks, when purchase lags stretch, daily comparisons become almost meaningless.

Tracking gaps: the part that is actually broken

The four causes above are by design. This one isn't:

Missing or messy UTMs. Meta traffic without UTM tags often lands in GA4 as referral or organic social instead of paid. A rising share of "Unassigned" traffic in GA4 usually points to inconsistent UTMs, missing referral exclusions or consent gaps.

Duplicate events. If both the pixel and the Conversions API send a purchase without a shared ID, Meta can count one order twice. Blocked tags. If something stops your Google Tag Manager container from loading, none of the tags inside it fire, and those users disappear from GA4 and Meta alike.

What changed in Meta attribution in 2026?

Meta narrowed what counts as a click. From March 3, 2026, click-through attribution for website and in-store conversions only counts genuine link clicks. Likes, shares, saves and comments moved into a new engage-through category.

The default setting in Ads Manager is now 7-day click, 1-day engage-through and 1-day view. Billing did not change; only how conversions are labelled and distributed did.

Two practical effects follow:

Your click-through column shrank. Conversions that used to sit under "click" now appear under engage-through. If you only report the click column, it looks like performance dropped. Some conversions disappeared entirely. A purchase made two to seven days after a like or save no longer qualifies as click-through (there was no link click) or engage-through (it falls outside the one-day window). Retargeting campaigns, where engagement-driven conversions were common, feel this most.

The upside is that Meta's click-through number now sits closer to what GA4 can see. The gap narrows but never closes.

A worked example: one ₹5 lakh month, four different answers

This is an illustration built to show the mechanics, not client data.

A D2C skincare brand has an average order value of ₹1,500. In October it spends ₹5,00,000 on Meta and ₹2,00,000 on Google Ads. Here is what each tool reports:

Source	What it reports	Purchases	Revenue	Implied ROAS / MER
Meta Ads Manager (default setting)	Click-through 430 + engage-through 60 + view-through 130	620	₹9,30,000	1.86x Meta ROAS
Meta Ads Manager (click-through only)	Link clicks within 7 days	430	₹6,45,000	1.29x
GA4 (data-driven, Meta source/medium)	Fractional credit for Meta sessions	290	₹4,35,000	0.87x
Shopify (orders with Meta UTM, last click)	Orders whose last tagged visit came from Meta	310	₹4,65,000	0.93x
Shopify, all orders	Every order from every channel	1,400	₹21,00,000	3.0x blended MER (₹21L ÷ ₹7L total ad spend)

Every row is correct by its own rules:

Meta's 620 includes 130 people who never clicked. It also books some late-September clicks that converted in October into September. GA4's 290 gives Meta partial credit when Google or email also touched the journey. It also misses the phone-to-laptop buyers it couldn't connect. Shopify's 310 is last-click only, so any Meta-influenced buyer who came back through Google is credited to Google. The blended 3.0x MER is the only number here that can't double-count. It is also the only one that can't tell you which campaign to scale.

A founder looking at the GA4 row alone would cut Meta, since it's below break-even. Looking at the Meta default row alone, they would triple the budget. Both would be wrong.

Which number should you trust?

Trust each number for the decision it was built for. None of them works for every decision.

Decision	Number to use	Why
Which ad, ad set or creative to scale inside Meta	Meta Ads Manager, with click-through and engage-through as separate columns	It's the signal Meta's delivery system optimises on, and it's consistent within the platform
How Meta compares with Google, email or organic	GA4 (data-driven model, same date range)	One rulebook applied to every channel
How much total budget the business can afford	Backend revenue and blended MER	Can't double-count, and ties to actual money in the bank
Whether Meta is truly driving extra sales	Incrementality: Meta's incremental attribution column or a holdout/lift test	The only method that estimates what would have happened without the ads

On that last row: Meta's incremental attribution setting optimises delivery using models that predict whether an ad caused a conversion, and choosing it removes your ability to edit attribution settings. You don't need to switch campaigns over to read it. You can add incremental conversion columns next to your standard data through the Compare Attribution Settings option in Ads Manager. Expect a lower number. That's the point. Jon Loomer Stellaheystella

How do you reconcile Meta and GA4? A weekly checklist

Reconciling doesn't mean forcing the numbers to match. It means explaining the gap and noticing when the gap moves.

One-time setup

Tag every Meta ad with UTMs using Meta's dynamic parameters (campaign, ad set and ad names), so GA4 reports Meta as paid social and not referral. Set up the Conversions API with deduplication. The pixel's event ID must match the server event's ID. Meta suggests an order number or transaction ID, so two real purchases stay two and never become four.

Check Event Match Quality for your purchase event in Events Manager and fix weak customer data parameters. Confirm your GA4 attribution settings. Note the model (data-driven by default) and lookback window, and write them down so everyone compares like with like. Add Meta's column breakdowns. Show click-through, engage-through and view-through as separate columns instead of one blended results column.

Every week (about 30 minutes)

Pull four numbers for the same 7-day range: Meta click-through, Meta total, GA4 Meta-attributed purchases, and total backend orders. Compute the ratio of Meta click-through to GA4 and compare it with your running baseline. If the ratio jumps sharply in one week, check for broken UTMs, a new landing page without tags, a GTM change, or duplicate purchase events before blaming the campaign. Check GA4's Unassigned channel share. If it's rising, find out why. Track blended MER week over week. This is the number the business actually runs on.

Set your own variance band. No universal "correct" gap exists; it depends on your category, purchase lag, share of mobile traffic and how much you spend on video. Record the Meta-to-GA4 ratio for 4–8 stable weeks, and treat that range as normal. Investigate only when you move outside it.

Frequently asked questions Why does Meta show more conversions than Google Analytics?

Meta credits purchases from people who viewed or engaged with an ad without clicking, and it can connect the same person across devices through their login. GA4 only sees visits to your website and shares credit across every channel in the journey. Meta's number is almost always higher as a result, and that is expected.

Should I trust Meta Ads Manager or GA4?

Trust each for a different job. Use Meta Ads Manager to decide which ads and audiences to scale inside Meta, GA4 to compare Meta against other channels on equal rules, and backend revenue with blended MER to decide total budget. Neither tool alone tells you whether ads caused the sales. That takes an incrementality test.

What is engage-through attribution in Meta Ads?

Engage-through attribution counts conversions that happen within one day after someone interacts with an ad without clicking a link, for example by liking, saving, sharing or commenting. Meta introduced it in March 2026 when it limited click-through attribution to genuine link clicks. The default setting is now 7-day click, 1-day engage-through and 1-day view.

What is a normal gap between Meta and GA4 conversions?

No universal benchmark exists, because the gap depends on your category, purchase lag, device mix and how much of your spend goes to video and awareness. Track the ratio between Meta's click-through conversions and GA4's Meta-attributed conversions for four to eight stable weeks. Treat that range as your normal, and investigate only when a week falls outside it.

Does the Conversions API fix the Meta vs GA4 mismatch?

No. The Conversions API helps Meta recover purchases that the browser pixel misses because of ad blockers or privacy settings, which usually makes Meta's numbers more complete, not closer to GA4. It must also be deduplicated with a shared event ID, or Meta can count the same order twice and widen the gap.

Why do Meta and GA4 show different numbers for the same dates?

Meta assigns a conversion to the date of the ad impression, while GA4 records it on the date the purchase happened. A click on the last day of one month that converts in the next month appears in different months in each tool. Compare weekly or monthly totals, not single days, especially around sales.

Should I switch my Meta campaigns to incremental attribution?

Look before you switch. Add the incremental attribution columns through Compare Attribution Settings in Ads Manager and see how far they sit from your standard numbers. Expect a lower count, since the model estimates only conversions your ads likely caused, and note that choosing it for delivery locks your attribution settings.

The bottom line

Meta and GA4 will never match, and chasing a match wastes weeks. Fix the gaps that are real errors (UTMs, deduplication, blocked tags), then give each number its job: Meta for in-platform optimisation, GA4 for channel comparison, and backend revenue for budget.

If your numbers are drifting and nobody on the team can explain why, our analytics and attribution audit traces your Meta, GA4 and Shopify setup end to end and shows where the gap comes from. It's a sensible first step before you move budget.