Short answer: to tell better tracking from changed attribution, compare three numbers day by day, with the same attribution setting throughout: the orders Shopify recorded, the purchases the platform received, and the purchases the platform credited to ads. Better tracking moves the second number closer to the first. Changed attribution moves only the third.
You change your tracking setup, and a week later Meta or Google Ads reports more conversions. Good news, or a trick of the numbers? An agency we work with asked us exactly this, and it's the right question to ask.
Why can conversions rise after a tracking change?
There are two very different reasons:
- More purchases reach the platform. Sales that used to get lost, for example blocked by the browser, now arrive. That's better tracking.
- The platform credits more purchases to ads. The same purchases arrive, but more of them are attributed to your campaigns, for example because the attribution setting or window changed. That isn't better tracking. It's different crediting.
Both show up as "more conversions" in your ad reports. Only the first means your data got better.
Which three numbers should you compare?
| Number | Where to find it | What it tells you |
|---|---|---|
| Orders recorded | Shopify | What really happened |
| Purchases received | The platform's event view, for example Meta Events Manager | How many sales the platform was told about |
| Purchases credited | The platform's ad reports, for example Meta Ads Manager | How many of those the platform attributes to your ads |
Better tracking shows as received purchases moving closer to Shopify's orders. Changed attribution shows as credited purchases moving while received purchases stay the same.
How do you run the comparison?
- Keep the attribution setting fixed. Use the same setting in the ad reports for the whole period. Changing it changes the credited number on its own.
- Compare day by day. Daily figures show when the change happened; weekly totals hide it.
- Use enough days. At least two weeks before and two weeks after the change.
- Allow for consent. Orders from shoppers who rejected cookies won't reach the platform, so received purchases will stay below Shopify's orders. That gap is expected.
- Write down what else changed. New campaigns, budgets or a sale can move the numbers too.
How does GTM Assistant help?
GTM Assistant sends each sale server-side, directly to your platforms, and each sale is counted once. That's the received number: when tracking improves, it moves closer to what Shopify recorded. It's built on Shopify's Customer Privacy API, so every customer's consent choice is respected before any event is sent.
Install GTM Assistant from the Shopify App Store
Frequently asked questions
Why did conversions go up after we changed tracking?
Either more purchases now reach the platform, which is better tracking, or the platform credits more of them to ads, which is a change in attribution. Comparing three numbers tells you which.
What is the difference between received and credited conversions?
Received is every purchase the platform was told about, for example in Meta Events Manager. Credited is the share the platform attributes to your ads under its attribution setting, for example in Ads Manager.
Why must the attribution setting stay the same?
Because changing it changes the credited number on its own, even if tracking didn't change at all.
How long should I compare for?
At least two weeks before and two weeks after the change, day by day, so normal swings even out.
Sources
- Shopify Help Center: Analyzing the success of your online marketing campaigns
- Jon Loomer Digital (independent guide): Meta Ads Attribution Setting: A Complete Guide
- A question from an agency we work with, October 2026 (not named)
Related: Why Shopify, GA4 and Meta never show the same number · Shopify says Direct. Most of it isn't
Want a second pair of eyes on your numbers?
Our free tracking audit compares the orders and revenue Shopify recorded with what GA4 and Google Ads received, and flags duplicate purchases. We reply within 24 hours. The report takes up to one week.