Campaign sales totals differ from store order reports when they count different orders, dates, amounts or attribution signals. Reconcile them at order level: agree on the reporting rules, match store order IDs, then explain each remaining difference. A campaign total that includes shipping cannot match a merchandise-only total without an adjustment.
Use the commerce records for order amounts and status. Keep the campaign system's evidence for creator attribution. Neither report should silently redefine the other.
Write the comparison rules first
Before exporting, put these rules at the top of the reconciliation sheet:
- Order scope. Which store, sales channels and creator codes or links qualify? Are you measuring entire orders or only promoted products?
- Status and dates. Which payment and cancellation states qualify? Does the period use order creation, payment or fulfillment time? Name the time zone and export cutoff.
- Money and credit. Which currency, discounts, taxes, shipping, reversals and attribution rules apply? How will you count an order associated with two creators?
These are reporting choices. They are not universal platform defaults. Shopify's sales-report definitions include pending, unpaid and canceled orders, while excluding test and deleted orders. A paid-order campaign measure therefore needs an explicit filter before comparison.
Keep payment status separate from fulfillment status. If your chosen measure counts paid orders, an unshipped paid order should not disappear merely because another report filters for fulfilled orders. Check both filters.
For a merchandise measure, agree to compare product value after discounts and relevant merchandise reversals. Keep taxes, duties, shipping and fees in separate columns. Shopify distinguishes net sales from total sales, which adds those other components. Also confirm that both exports use the same currency basis before doing arithmetic.
For recurring reports, use a campaign data dictionary to preserve these field definitions between reporting owners.
A worked reconciliation with six discrepancy types
The following orders, dates and USD amounts are hypothetical. They illustrate a custom campaign measure, not the default output of a Shopify report.
The agreed measure is merchandise value after discounts and merchandise refunds for eligible orders created September 1 through September 28, 2026, in America/New_York. Orders must have been paid; partially refunded orders remain eligible. Updates are included through September 30 at 09:00 in that time zone. Taxes and shipping are excluded. A qualifying creator code or tracked link admits an order, but each order counts once across the campaign. Canceled and test orders are excluded.
The campaign export shows $978. A back-office extract prepared under those rules shows $600. Six order-level differences explain the $378 gap. Another $220 of eligible orders already matches in both extracts.
| Difference | Order | Campaign amount | Agreed amount | Explanation |
|---|---|---|---|---|
| Order status | R101 | $120 | $0 | Payment is still pending at cutoff; the chosen measure requires paid orders. |
| Time zone | R102 | $90 | $0 | Created September 1 at 02:15 UTC, which is August 31 at 22:15 in New York. |
| Taxes | R103 | $110 | $100 | Campaign amount includes $10 tax; the comparison excludes it. |
| Shipping | R104 | $88 | $80 | Campaign amount includes an $8 shipping charge. |
| Refund timing | R105 | $150 | $100 | A $50 merchandise refund processed September 29 is missing from the campaign export. |
| Attribution overlap | R106 | $200 | $100 | One $100 order appears in both code-attributed and link-attributed rows. |
The reconciliation is:
| Adjustment | Amount |
|---|---|
| Campaign export | $978 |
| Remove pending order | -$120 |
| Remove order outside local date range | -$90 |
| Remove tax | -$10 |
| Remove shipping | -$8 |
| Apply merchandise refund | -$50 |
| Remove duplicate attribution credit | -$100 |
| Reconciled campaign merchandise value | $600 |
Do not apply these adjustments to every report automatically. If the campaign export already excludes tax, its tax adjustment is zero. If a refund includes merchandise, tax and shipping, split those components before subtracting it from a merchandise-only measure.
Discounts need the same care. In this example, every product amount already includes its discount. Subtracting that discount again would create a new error.
Separate order-date reporting from refund-date reporting
R105 exposes a reporting choice that can survive every other cleanup.
Shopify records a sale on the day it occurs and a reversal on the day it is processed, according to its sales and reversals documentation. A September 1 through September 28 sales report therefore does not include a reversal processed September 29.
The example instead groups orders by their creation dates and updates those orders through a later cutoff. That answers, "What is the current merchandise value of the orders placed during this campaign period?"
An event-date report answers a different question: "What sales and reversals occurred during this period?" It can include refunds for orders placed before the campaign began.
Keep both views if finance needs the event-date report and the campaign team needs an updated order cohort. Label the difference and preserve the order IDs behind it. Do not move a refund date in the source records to force agreement. Use reporting-window rules to decide when the campaign number remains provisional.
Sales also differ from cash received. Shopify says its sales reports do not track payment movement. Use the relevant payment records when the question concerns collected cash rather than order value.
Reconcile attribution without adding every total together
An order can have a creator code, a creator-link visit and another marketing interaction. Those signals can coexist.
The Modash measurement guide discusses codes, links and post-purchase surveys as methods with different coverage. Record which signal admitted each order into your campaign total. A survey response belongs in a labelled field, with its own inclusion rule.
Shopify's marketing reports also show why attribution totals can diverge. Different models assign credit differently. Its Any click model gives full credit to each clicked channel, so adding channel credits can exceed the number of orders.
For a campaign-wide order total, deduplicate by store and order ID. Keep additional attribution rows for analysis. If you need to choose which creator receives credit, use the separate link-and-code credit policy. That decision does not change the store's order amount.
Close the unexplained balance
Work through the exports in this order:
- Save each export's report name, filters, time zone and extraction time. Shopify identifies time-zone differences and delayed data as possible causes of discrepancies.
- Match store and order IDs. Classify unmatched rows as campaign-only or store-only before comparing values.
- Check row grain. Shopify warns that custom report columns can create multiple rows per order, and discount-code reports can repeat orders with multiple discounts. Count distinct orders; do not blindly sum repeated order totals.
- Compare matched orders component by component. Record every adjustment with its order ID, reason and evidence.
- Assign each unexplained item to an owner. Keep its amount in an unresolved column instead of changing an attribution rule to erase it.
Your handoff should state the starting campaign amount, explained adjustments, reconciled amount and unresolved balance. Attach the rules and export cutoff. This establishes the sales figure for reporting; creator commissions remain a separate calculation.
Start with the largest unmatched order in the current report. Check its status, timestamp, amount components and attribution rows before changing any dashboard setting.
