Which numbers should I trust: Meta, Google, GA4 or Shopify?
What does each tool actually count?
Each tool answers a different question and calls the answer "sales". That is why the four numbers never match, and why none of them is simply wrong.
| Source | Counts | Cannot see | Usually leans |
|---|---|---|---|
| Shopify | Every order, discount and refund | Which ad brought the buyer, beyond the visits it recorded | Neutral on totals |
| Meta Ads Manager | Purchases within its click and view windows, some of them modeled | Other channels, buyers it cannot match | High (median 1.12x matched sales) |
| Google Ads | Conversions within 30 days of a Google ad click | Meta, email, sales its tag never received | Low in our data (median 0.81x), high on brand search |
| GA4 | Purchases from sessions it could follow | Declined consent, blocked trackers, other devices | Low (median 76% of orders) |
Sources: Meta Marketing API (default window "7d_click, 1d_view"), Google Ads Help, conversion windows, Google Ads Help, reporting by click date, Google Analytics Help (GA4 uses data-driven attribution by default). The ratios cover stores with at least about $1,000 of spend on the channel in the 4 weeks; stores where one side saw under half of the other are left out as broken tracking.
Which number should I use for which decision?
| Decision | Use | Do not use |
|---|---|---|
| How much did we sell this month? | Shopify net sales or total sales | Any ad platform |
| Do ads pay off for the business overall? | Shopify sales divided by all ad spend (MER) | The sum of each platform's ROAS |
| Which channel gets more budget next month? | Sales matched to real orders, one credit per order | Platform ROAS side by side |
| Which ad or campaign should I scale or pause? | Matched sales per ad, with the platform's number next to it | Ads Manager alone |
| Is the ad delivering? Cost per 1,000 views, frequency, click rate | Meta Ads Manager and Google Ads | Shopify |
| Where do visitors drop off on the site? | GA4 | Ad platforms |
| What does the bidding algorithm learn from? | The platform's own conversions, kept complete | Turning events off to make reports look smaller |
MER (sales divided by total ad spend) is the one number that cannot be inflated by attribution, because both parts come from your own books. It cannot tell you which ad worked. See when MER is better than ROAS.
Why Shopify is the base for totals
Shopify records each order once. Its sales report defines net sales as gross sales minus discounts and sales reversals, and total sales as net sales plus taxes, duties, shipping and fees. A refund shows up as a negative amount on the day it is processed, not on the day of the original order (Shopify Help Center).
Pick one of those definitions and use it every month. On WooCommerce the same rule applies: your order list is the base, the ad platforms are claims on top of it.
Why the platform numbers still matter
Meta and Google learn from the conversions they receive. If you send them fewer or worse purchase events, their bidding gets worse, even if you never look at their reports. So keep their tracking complete (pixel plus Conversions API on Meta, enhanced conversions on Google), and keep using their reports for what only they can show: delivery, cost per view, frequency and creative fatigue.
What you should not do is add the platforms' sales together, or rank channels by each platform's own ROAS. Each one grades its own work with its own rules.
How far apart should the numbers be?
Some gap is normal. In our data, Meta reported more than we matched in 68% of 91 stores, and Google Ads reported less than we matched in most of 65 stores. Median ROAS was 1.61 by Meta's count and 1.40 on matched sales; for Google Ads it was 2.46 by Google's count and 2.91 on matched sales.
A gap of 2x or more in either direction is different. In our filter we treat it as broken tracking, and you should too: check for duplicate purchases, missing server events or a pixel that stopped firing before you read anything into the numbers. The usual causes on the Meta side are in why Meta shows more sales than Shopify.
What about Shopify's own attribution reports?
Shopify's marketing reports split sales by channel using last non-direct click by default, and you can switch to first click, last click, any click or linear (Shopify Help Center). They are free and built on your real orders, so they make a good cross-check. They only know the visits Shopify recorded, so an ad someone saw but never clicked does not show up there at all.
A tool that matches each order to every ad touch, like TrueROAS, does the same job with click IDs and server-side records, and shows the platform's own number next to it. Whichever you use, check the numbers once a month with the routine in how to reconcile GA4, Meta and Shopify each month.

