The short answer
Meta and Shopify will never match, because they're measuring different things. Meta counts every sale it can connect to someone who saw or interacted with an ad. Shopify gives each sale to the last place the customer came from.
So one sale can have two owners. Neither number is the truth. Your bank balance is.
The gap isn't a fault to fix. It's a signal to read. This guide explains where it comes from, what changed in 2026, and what to track instead of chasing a match.
How each platform decides who made the sale
Attribution is just a set of rules for handing out credit. Meta and Shopify use different rules.
| Header 1 | Meta Ads Manager | Shopify |
|---|---|---|
| Who gets the credit | The ad, if someone clicked, engaged or viewed it before buying | The last channel the customer arrived from |
| Time it allows | Up to 7 days after a click, 1 day after a view or engagement (on standard settings) | The visit that led to the purchase |
| Sees other channels? | No. It only knows about Meta | Yes, but only the final step |
| Bias | Tends to over-claim | Tends to under-credit Meta |
Here's a typical customer journey. Someone watches your video on Instagram on Monday. They search your brand on Google on Wednesday. They open your email on Friday and buy.
Meta says the video made the sale. Shopify says the email did. Both are partly right, and both are counting the same £60 order.
Why Meta usually shows more sales than Shopify
Meta is marking its own homework. When we audit an account with a big gap, it almost always comes down to three things.
Audience segments set up badly. If Meta can't tell new customers from existing or engaged ones, catalogue ads and heavy retargeting end up selling to people who were going to buy anyway. Meta takes the credit for sales it didn't create.
Attribution settings that let Meta claim sales your email marketing made. Someone sees an ad, then buys from your Friday newsletter. Both platforms count it.
View-through conversions. Someone scrolls past an ad, buys days later through another route, and Meta still counts it as an ad sale.
None of this means Meta is useless. It means Meta's dashboard is built to help Meta's algorithm find buyers, not to keep your books.
What changed in March 2026, and why it matters
In March 2026, Meta changed what counts as a click. Only link clicks now count as click-through. Likes, comments, saves, shares and 5-second video views moved into a separate bucket called engage-through.
The catch is the window. Click-through gets up to 7 days. Engage-through only gets 1 day.
That creates two problems in your data reviews.
Your numbers aren't comparable year on year. If your account uses a click-only setting, this year's results look lower than last year's, even if nothing actually changed. Compare them side by side and you'll think performance dropped.
Your budget drifts to the bottom of the funnel. A video that people watch, save and come back to buy from later loses its credit under click-only. Offer statics and retargeting get clicked, so they keep all the credit. Meta moves spend towards them, the video gets less budget, looks worse, and gets cut. Then someone decides video doesn't work for the brand.
It probably does. It just stopped being counted.
Before you kill anything, check your attribution setting and look at engage-through results in your columns. Then check whether your new customer revenue moved.
A real example: £5 per new customer that was really £30
The founder of an eight-figure eCommerce brand showed me their ad account.
| Header 1 | Figure |
|---|---|
| Average order value | £60 |
| Target cost per new customer (nCAC) | £15 |
| nCAC in Meta's dashboard | £5 |
| nCAC in their tracking software | around £30 |
On Meta's numbers, they were winning by a mile. Most people would have doubled the budget on the spot. Her verdict was "that just can't be true". She was right. The real cost was double what the business could afford.
The gap came from the three usual suspects above: segments that couldn't separate new from existing customers, email sales landing in Meta's column, and view-through conversions.
Here's where it went wrong next. The team had spent hours trying to match sales to ads line by line, to explain the gap between the two dashboards. That's time nobody gets back, and it can't work. Nobody clicks one ad and buys. There's a messy middle of scrolling, searching, emails and second thoughts, and no tool sees all of it.
What to track instead
Stop asking which dashboard is right. Ask whether the ads made you money you wouldn't have made anyway. That's incrementality, and it's the only question that matters to your P&L.
As an agency, we track four things.
Money in versus money out. Total revenue against total ad spend across every channel, plus new customer revenue on its own. If spend goes up and neither moves, the ads aren't working, whatever the dashboard says.
Whether something changed, and whether the final number moved. A budget increase, a new offer, a seasonal peak. Did profit move with it? That's the test.
Tests that move one lever at a time. Scale Meta up, or Google down, and keep everything else steady. Then watch what happens to total revenue and new customers. Where you can, run a holdout: switch ads off for one region or group and compare it with the rest.
Time. Judge every change against how long your customers actually take to buy. For one of our outdoor living clients, customers took three months from first seeing an ad to buying. So we judge every change over three months, not three days.
Use Meta's dashboard for steering the campaigns day to day. Use these four for deciding where the money goes.
Frequently asked questions
-
Meta counts any sale it can link to someone who clicked, engaged with or viewed an ad. Shopify gives the sale to the last channel the customer came from. The same order often appears in both, so Meta's total usually looks higher.
-
Neither on its own. Use Meta's numbers to steer campaigns and your total revenue, new customer revenue and profit to make budget decisions.
-
New customer acquisition cost. It's what you pay in ad spend to win one first-time customer. It's more useful than ROAS for growing brands because it ignores sales from people who were already buying.
-
From March 2026, Meta only counts link clicks as click-through. Likes, saves, shares, comments and 5-second video views now count as engage-through, which has a 1-day window instead of 7.
-
If your account uses a click-only attribution setting, sales from people who engaged with an ad but didn't click a link are no longer counted. Your reported results can fall even when real sales haven't.
-
Watch total revenue and new customer revenue when you change one thing at a time, such as ad budget. If possible, run a holdout test. Judge results over your real buying cycle, not just the last few days.
About the author
Aggie Meroni is the founder of White Bee Digital, a Meta Ads agency for 7-figure eCommerce brands. She spent over 10 years in investment management before moving into paid social, is one of the UK's certified Meta Lead Trainers, and wrote the bestselling Meta Ads book Crack the Code.
Read next: ROAS vs ROI vs contribution margin.
Not sure your dashboard is telling you the truth? Let's find out what your ads are really bringing in.
