SALES REPORTS
Why Shopify and email sales totals differ
Reports can give credit for the same order in different ways. They may also use different dates or sales totals. Match those rules before you compare the numbers.
In this guide
Start with a few orders
Pick one email and a short time span that has ended. Find the orders in both reports. Write each order number and time. Note the product sales after deals and refunds. Keep the refund amount too.
Note which click or open got credit. Check the time limit for that credit. Mark each order as in Shopify, the email report or both.
Ten orders may show why totals differ. They cannot tell you how all your emails perform.
Check the rules for sales credit
Shopify has several rules. First-click gives credit to the first click. Last-click uses the last. Last non-direct click skips direct visits. Any-click can give more than one channel credit. Linear attribution shares the credit.
Your email app may use another rule. It might count an open within a set number of days. Write each rule down. Different rules can give different totals for the same order.
Use the same sales and dates
Decide what “sales” means. One choice is product sales after deals and refunds, with no tax or shipping. Use the same currency, dates and time zone in both reports.
Check when refunds count. Check whether the report uses the email date or order date. These can change the total.
Shopify says reports can take time to update. Wait for the stated update time before you compare.
Check the tags in your links
Google Analytics uses link tags to show where a visit came from. These include source, medium and campaign. Keep the names clear and use the same style each time.
For example: source=newsletter, medium=email, campaign=welcome_guide. Check that the final link keeps these tags. Do not put names or email addresses in them.
A tag helps trace a visit. It does not prove the email caused a sale. Check missing tags apart from the rules for sales credit.
Did the email cause an extra sale?
Attribution means giving credit under a rule. It does not prove the email made the sale happen. Someone who planned to buy may still click your email.
To look for extra sales, compare a group that gets the message with one that does not. Google calls this type of ad test Conversion Lift.
An email test needs its own plan for your store. Google Ads does not measure your email results for you.
Plan a group that will not get the test
With enough shoppers and the right tools, split people into two groups at random. One gets the email. The other does not. The second group is called a holdout.
Choose what to count and the test dates first. Keep people in their assigned groups. Keep normal order updates on. Do not send a new deal only to the group missing the test email.
Compare everyone in each group, not just those who clicked or opened. This is a test plan. It is not a built-in myuser.ai test tool.
Count costs and allow for doubt
The samples below show the math. They do not prove an email worked. A real test needs enough people to spot the change you care about. It also needs a range for how sure the result is.
Do not stop at the first good result. It may be chance. Say when the result is too close to judge.
Count item costs, deals and other costs per order. Add app and sending fees. More sales do not always mean more profit.
Choose one next step
Note which gaps came from dates, rules or missing data. Keep three totals apart: store sales, sales credited to email and your best estimate of extra sales.
Choose one fix. You could fix link tags, cut repeat sends or plan a test.
The myuser.ai cost calculator shows how many extra orders could cover a plan. It does not measure how many extra orders your emails made.
When to send, and when to wait
Each example explains what happened and why a message may help—or why it should wait.
One order counted by two reports
Fictional order: $80 of product revenue after discounts.
Monday: the shopper clicks a newsletter. Tuesday: they click a paid ad and buy.
The email tool credits the order within its email-click window. A last-click report credits the ad.
Store revenue remains $80. Adding $80 of email credit and $80 of ad credit would not create $160 of sales.
Check the order and each report’s rules. These are example rules, not a description of myuser.ai’s own reports.
Compare a group that gets the email with one that does not
Example groups chosen at random: 1,000 subscribers get the email. Another 1,000 do not get the test email. The second group is called a holdout.
During the same fixed time period, 70 people buy in the email group and 60 buy in the group without it.
That is 7% compared with 6%, a gap of 1 percentage point. At these rates, it means about 10 more buyers per 1,000 people.
If net product sales are $5,600 and $4,800, that is $5.60 and $4.80 per person in each group. The gap is $0.80.
These example numbers show the math. They do not prove an effect or predict your sales. A real test needs enough people and a careful check of uncertainty.
A sheet to compare your weekly reports
Email or campaign name: Reporting dates and timezone: Currency and revenue definition:
Store order total: Orders credited by email report: Orders credited by store report: Duplicate order IDs:
Credit rules and time limits: Refunds and delays between reports: Missing tracking or unmatched orders:
Evidence of extra sales: none / planned test / completed test What is still unclear: Next step and who will do it:
Write “none” if you have no test showing extra sales. A report giving an email credit is not the same as proof that the email caused a sale.