Why do three dashboards show three totals?
Imagine your affiliate network shows 100 conversions, your online store shows 96 orders, and your payment processor shows 94 successful charges. It can feel as though someone has lost six sales. But the systems may be measuring different things: an attributed conversion, an order placed, and a payment successfully captured. Until we compare the underlying records, none of those totals tells the whole story.
A conversion, an order, and a successful charge are different events.
Start with an apples-to-apples comparison
Before searching for missing sales, check that the reports cover the same dates and time zone. Confirm whether each includes test orders, canceled orders, upsells, subscription rebills, refunds, or pending payments. Check whether amounts use the same currency and whether a report counts orders or individual items. A midnight cutoff or a different definition of “sale” can create a discrepancy even when records are accurate.
Walk through a realistic mismatch
Here is one fictional investigation. The affiliate network lists 100 conversion records. The store lists 96 order records. We discover that four network conversions have no confirmed matching store order. Among the 96 store orders, 94 have successful captured payments and two are still awaiting payment. This explains how the three reported totals can coexist—but it does not yet prove why those four network records have no matching order.
What do we do with the four unmatched conversions?
We keep those four conversion records rather than deleting them or automatically labeling them fraudulent. We review their original transaction references, status changes, reporting times, and available merchant order IDs. Perhaps a conversion was duplicated, delayed, or never completed as an order. Until the evidence confirms the reason, the records remain unmatched and available for investigation.
Four unmatched conversions are an open question—not four confirmed lost sales.
Why exact references matter
A $129 conversion and a $129 order that happened a few minutes apart might be related, but that is not enough to prove they belong together. A shared click reference, provider transaction ID, merchant order ID, or other known relationship is stronger evidence. TraceKit's approach is to preserve those original references and keep the marketing attribution separate from the commercial transaction.
Matching references provide stronger evidence than similar amounts or timestamps.
Other reasons numbers can change
Reports may also disagree because a customer received a refund, an upsell was counted as a second conversion, a subscription renewed, a payment failed, or an affiliate platform received a reversal later. Some integrations deliver events more than once or after a delay. Those cases should be examined as distinct source events rather than silently merged into the original purchase.
Different event types and statuses should remain visible.
A checklist your team can use
When two reports disagree, start by recording the exact totals and the reporting period. Then check definitions, statuses, currencies, and source freshness. Compare the individual IDs and identify which records match, which have different statuses, and which remain unmatched. Document the explanation for each group and revisit open exceptions when more evidence arrives.
What does TraceKit help you understand?
The goal is not to force every platform to display the same number. It is to explain why the numbers differ using available evidence. That can help teams investigate affiliate payouts, checkout problems, payment issues, and campaign reporting with greater confidence. Exact capabilities depend on the integrations and records available for your account.
A trustworthy report explains the differences instead of hiding them.
Explore the connected picture.
See how TraceKit connects source evidence across the customer journey.