Measuring e-commerce transactions means transferring order data – identifier, value, items, shipping and discount – into an analytics tool. It serves to attribute revenue to channels and campaigns. A difference from the actual orders in the online shop system is common and arises from several causes: measurement consent declined, blockers, interrupted loading of the thank-you page, transactions sent twice on page refresh, and orders placed by phone. A deviation of up to a few percent is acceptable; a larger one requires checking. Evaluation should also include returned and unpaid orders, which analytics does not deduct by default – without this, campaign performance is systematically overstated, especially in sectors with a high share of returns. A practical solution is a regular monthly comparison of revenue in analytics against the accounting records, with the deviation found recorded.
See also: Offline conversion import, Data quality in analytics, Returns and their cost.
