Is your Meta pixel tracking purchases correctly? 8 checks for Shopify stores
Eight checks you can run in Meta Events Manager and Shopify, each with what I found on Australian stores.
By Mark Tran · Reviewed
The short answer
Your Meta pixel is tracking purchases correctly when the purchases Meta Events Manager receives from your store's browser match the number of orders in Shopify for the same dates, each order arrives once, and the store sends those events to the same pixel your ads use. If those three hold, a gap you see in Ads Manager is attribution, not broken tracking.
The eight checks below are the ones I ran on Australian ecommerce stores between August and November 2026, mostly on Shopify, with the brands' names removed. You can run all of them yourself without touching your ads.
1. Meta's purchase events match your Shopify orders
Where to look: in Meta Events Manager, open the Purchase event for the last 7 or 28 days and read how many arrived from the browser. The total adds browser and server events together, so it runs higher. In Shopify, count all orders for the same days. Use total orders, not orders from Meta: the pixel sees every buyer, whether they came from an ad or not.
What good looks like: the two numbers match, or nearly match.
What I found: a kids' clothing brand's pixel recorded 15 purchases from the browser against 15 Shopify orders over four weeks in August and September. A recheck in October matched again, 13 against 13.
2. Ads Manager never claims more sales than the store took
Where to look: the Purchases column in Ads Manager against total Shopify orders for the same dates.
What bad looks like: Meta reporting more purchases than the store had orders. Apart from date and time-zone edges, that means over-counting.
What I found: a women's fashion store's ads were credited with 1,121 purchases against 1,231 total Shopify orders over four weeks, 91% of every order the store took, including email and repeat buyers. Nothing showed double counting, but Meta's numbers alone can't tell you what the ads added. Read why Meta, Google and Shopify revenue differs for how to judge it.
3. The store sends events to the pixel your ads use
Where to look: open your store in a browser, view the page source and search for pixel_id. Compare that ID with the dataset your ad account and ad sets use. In Shopify, check the Facebook and Instagram app's data sharing is set to Maximum.
What bad looks like: two different IDs. Sales still happen and the pixel still fires, but Ads Manager can't see them.
What I found:
- A kids' clothing brand's store switched to a second pixel in the same business for about 28 hours in November. That pixel logged 5 purchases, and Ads Manager showed none across two days, while Shopify's own report put 2 of those orders on the Meta ads. The app's data sharing had also dropped from Maximum to Enhanced. Nobody recorded who changed either setting.
- A womenswear boutique's ad account had one pixel, and it had last fired in 2019. The live pixel on the site was recording 43 purchases a week with strong match quality, but the ads couldn't use it until the owner shared it to the ad account.
4. No Meta pixel you don't recognise is installed
Where to look: Shopify Settings, then Customer events. List every pixel there.
What bad looks like: a Meta pixel nobody can account for, such as a custom pixel named "Meta pixel (migrated)", which Shopify can create when it converts an older pixel install.
What I found: a menswear retailer had one of these sending every page view, add to cart and purchase, with order values, to a second pixel. No active ad used it, and its owner is still unknown.
5. Each purchase arrives once, from browser and server
Where to look: in Events Manager, open the Purchase event and compare events received from the browser with those from the server. Meta treats a browser and a server event as the same purchase when they share an event ID and name, so each order should count once.
What good looks like: browser and server close to one to one on Purchase, counted as one.
What I found:
- On the kids' clothing brand, once the store was back on the right pixel, all 4 purchases over the next four days arrived as a browser and server pair. Four is a start, not proof, so the count keeps getting checked.
- A homewares store on WooCommerce went from matched pairs in September to 26 server purchases against 11 from the browser over the next three weeks. A gap that wide means one path is dropping or double-sending purchases, and only the store's order list can tell which.
- A swimwear store's first purchase event arrived from the server only, on a PayPal order, and the cause is unknown. Separately, Meta flagged that the pixel's allow list in Events Manager left out the Shop Pay checkout domain, so I added it. No Shop Pay order has been checked since, so that fix is unproven.
6. A real test order fires the purchase event
Where to look: place one small real order, then refund it, with Events Manager open. Meta's Test Events tool shows events as they arrive.
What good looks like: one Purchase event with an order value in Australian dollars. On Shopify, server events may not appear in Test Events, so check the Purchase count again an hour later.
What I found: a formal-dress boutique on Wix had never recorded a purchase event. One test order made Purchase fire for the first time, from browser and server in the same hour.
7. Match quality is judged on Purchase
Where to look: Events Manager shows an Event Match Quality score for each event once it has enough events. Read it on Purchase. In the dataset settings, check Automatic Advanced Matching is on.
What good looks like: Meta scores match quality from 1 to 10 and recommends 6.0 or higher.
What I found:
- A swimwear store scored 5.6 to 6.1 while it had no orders. After its next order, Purchase scored 8.0, with an email and phone number on that purchase. Expect a lower score until real checkouts happen.
- A jewellery brand had Automatic Advanced Matching switched off, with every field off. It is a single setting in Events Manager.
8. Sudden jumps are read by day and by source
Where to look: in Events Manager, view each event by day and split it by browser and server.
What bad looks like: more checkouts than add to carts, or one event jumping on a single day from one source. That is usually setup, testing or bots, not shoppers.
What I found:
- A swimwear store's pixel showed more checkouts than add to carts, 73 against 65. 53 of the 73 checkouts landed in a two-day burst, almost all from the server. The store had no orders in that month.
- A bridal accessories store logged 394 add to carts in five hours, 370 of them from the server and 16 from the browser, all on the home page, a pattern that points to crawlers following add-to-cart links.
- When a menswear retailer's orders dipped, its daily purchase counts in Google Analytics and Meta's server events matched almost exactly over eight days, so the dip was in sales, not tracking.
How to score it
Give your store a yes or no on each check.
- All 8: trust your purchase count. Judge the ads on total sales against total ad spend, not on Meta's claimed share.
- Checks 1, 3 or 5 fail: fix tracking before changing any ad or budget, because the numbers you'd judge them on are wrong.
- Only 7 or 8 fail: the sales count is sound. Improve match quality and read the odd days with care.
Want a second pair of eyes?
If you run an Australian ecommerce store and aren't sure your Meta tracking is right, I go through it on a free 15-minute Meta ads audit. Your live campaigns on screen, the one fix to make first. No access needed, no obligation.
Method
Every "What I found" comes from an Australian ecommerce store I audited or managed between August and November 2026: mostly Shopify, plus two WooCommerce stores and one Wix store, in fashion, swimwear, jewellery, homewares and bridal accessories. Names are removed. Pixel figures are as Meta Events Manager reported them; order counts are from the store. Where a cause or fix is not yet confirmed, the text says so. These are examples of what to check, not benchmarks for every store.
