Foundation · 2 min
What you need to know
How do I know whether ecommerce analytics data is accurate?
Reconcile sampled transaction IDs and amounts among orders, payments and analytics; test event sequences; review diagnostics; and confirm definitions, filters, time zones and consent effects. Accuracy is demonstrated, not assumed.
Why do two analytics tools show different numbers?
Tools may differ in identity, scope, attribution, time zone, processing delay, consent, filters and transaction treatment. Compare a common transaction set before comparing aggregates.
Why doesn’t analytics revenue match store revenue?
The order system is the transaction record, while analytics may miss consented-out users, use different dates, include duplicates or handle refunds differently. Build a gross-to-net reconciliation by transaction ID.
Why are purchases being counted twice?
Purchases are often duplicated by confirmation-page reloads, client-and-server double sending, retries without deduplication or reused transaction IDs. Enforce one stable order ID and test repeated loads.
Why are some purchases missing from analytics?
Purchases may be missing because tracking was blocked, navigation ended before the event, server calls failed, consent was absent, parameters were invalid or filters excluded them. Compare missing IDs by device and payment path.
How do refunds affect analytics accuracy?
Refund data is accurate only when reversal events link to the original transaction, amount and items without deleting the purchase. Reconcile partial and full refunds separately.
How do test orders affect reporting?
Mark test orders with a stable environment, account or order attribute and exclude them consistently from business reports while retaining an auditable test record. Do not rely only on employee IP addresses.
How should bots and internal traffic be handled?
Identify verified bots and internal activity before reporting, using tool-supported filters and documented rules. Preserve a raw or testable view so an overly broad exclusion can be detected.
A quick store check
Match a sample of transaction IDs across store, payment and analytics systems and classify every difference.
Foundation · 2 min
Frequently asked questions
Should analytics exactly equal the order system?
Usually not, because scope, timing, consent and processing differ; material gaps still need explanation and reconciliation.
Can test traffic be deleted later?
Sometimes tools allow filters or exclusions, but prevention through a documented test marker is safer and more auditable.